8-18-97 Electronic Filings

8-18-97 Electronic Filings on Internet Domain Names

###

Number: 327

From: "Jeff Griswold" <jgrizzz@msn.com>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 1:24am
Subject: Domain Name Service

I just wanted to throw my two cents into the hat. I think that we
should seriously consider a regional approach to name resolve. If the
address's took on a regional identifier it might help ease some of the log
jamming the internet experiences. I would look at at least a state level
code for the US.

"www.business.fl.com" or "www.business.ny.com"

Just a thought
J. Griswold

###

Number: 328

From: <Nlordi@aol.com>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 12:47pm
Subject: Filing of Comments on Internet Domain Names

RECOMMENDATION

The issue of Internet domain name registration and administration is but
a subset of a much broader issue with far reaching consequences, ie.,
the future structure(s) and mechanisms for global Internet governance.
We need to fully appreciate that whatever path we pursue we are not only
addressing management of Internet domain names, but are in fact establishing
a precedent for a private self governance system for the Internet.

Therefore, the US government should work within international circles
to discuss the establishment of a multi-national panel with a compensated,
full time staff with the necessary resources to focus on and
address the issue of global Internet governance. Recommendations should be
provided to the Internet community within 6 to 9 months.

The US government should also consider the upcoming 1 - 3 October 1997
conference (http://www.ispo.cec.be/standards/conf97/welcome.html)
"Building the Global Information Society for the 21st Century"
sponsored by the International Telecommunication Union (ITU), International
Organization for Standards (ISO), and International Electotechnical
Commission (IEC), as a possible venue to initiate discussions
on the topic global Internet governance.

DISCUSSION

Internet domain names is but the first of many issues to surface.
A partial list of other issues which may effect the openness and
universal connectivity of the Internet, and may need to be addressed in
an appropriate forum in the very near future include:

- intellectual property rights regarding the content one posts to
the Internet,
- settlement fees between Internet Service Providers,
- the establishment of rules for peering and exchanging traffic between
Internet Service Providers,
- regulation of gambling over the Internet,
- regulation of pornography over the Internet,
- regulation of taxation,
- the use of "cookies" and ones rights to privacy,
- the establishment of legal jurisdictions and boundaries
regarding the Internet
- management and assignment of IP address space,
- governance and structure of the RIPE, APNIC and ARIN registries,
- ownership of IP address space,
- regulation of email spam,
- intellectual property rights surrounding the use of third party cache
and replication servers,
- governance of Internet telephony, and
- a concern that non-US Internet users pay an inordinate cost for
international connections between the US and other continents.

Any one of these above issues could effect the global stability and
connectivity of the Internet just as easily as the domain names issue.

We must therefore ask ourselves "given the existing set of proposals
regarding Internet domain names, are we establishing a mechanism and
precedent for Internet governance under which new Internet governance
issues will likely be addressed ?"
Our answer to this question is yes, as new Internet governance issues
surface, common sense would dictate to try to address these issues in
an already established framework.

Consequently, we must also ask ourselves, "are the processes and set of
participants in the IAHC / gTLD-MoU sufficient enough to address new
and emerging issues which effect the connectivity of the Internet
and global Internet governance ?".
Our answer to this question is no, the IAHC / gTLD-MoU processes
were not designed to support and are not sufficient to support
global Internet governance.

In fact, in the charter of the IAHC (http://www.iahc.org/iahc-charter.html)
no mention or mechanism is made of how to
address new emerging issues of Internet governance.

Therefore, the US government should consider sponsoring a session to
discuss the establishment of a multi-national panel with a compensated,
full time staff with the necessary resources to address the issue of
Internet governance, with recommendations provided to the
community within 6 months.
The US government should also consider the upcoming 1 - 3 October 1997
conference (http://www.ispo.cec.be/standards/conf97/welcome.html)
"Building the Global Information Society for the 21st Century"
sponsored by the International Telecommunication Union (ITU), International
Organization for Standards (ISO), and International Electotechnical
Commission (IEC), as a possible venue to initiate discussions
on global Internet governance.

ADDITIONAL RECOMMENDATIONS REGARDING THE ISSUE OF DOMAIN NAME MANAGEMENT

First, the system is stable, and contrary to the popular press there
does not appear, upon examining the Securities and Exchange Commission
Form S-1, filed July 3, 1997 for Network Solutions Inc.,
any exorbitant profits. In fact, it appears after reading the S-1 that
NSI is investing a majority of its fees in hardware, software, and
third party vendors in support of back office operations to deliver
its services.

The US government, specifically the National Science Foundation,
should be commended for actively managing the NSF/NSI cooperative agreement
for the betterment of the entire Internet community.
The real winner of this cooperative agreement is the Internet community.
The initiative of NSF in establishing an Intellectual Infrastructure Fund
with millions of dollars on account will eventually benefit the entire
Internet community.

The system established under NSI has provided a stable system for a fair
price. Where else can one request a service (the issuance of a domain name)
and have it "delivered" without first paying for the service, and without
any prior credit history ? In fact, the net fee of $35, besides supporting
the assignment of an Internet domain name, also supported, without cost to
the end user, the management and assignment of IP address space.

There are still many unanswered questions regarding the IAHC / gTLD-MoU plan.
For example, will the processes defined by the IAHC / gTLD-MoU provide a
comparable level of stability ? What fee(s) will end users have to
pay under the IAHC / gTLD-MoU plan for domain names ? Are 7 iTLDs enough ?
Where is the full time staff to provide a transition to the
processes defined in the gTLD-MoU ?

We are concerned that the IAHC / gTLD-MoU is moving forward
too quickly and fails to address several issues, such as those
outlined in Digital Equipment Corporation's Attachment to the gTLD-MoU
(http://www.itu.int/net-itu/gtld-mou/decadd.htm), ie.,
a concern for consensus building, stability of existing top level domains,
stability of name registration, and preserving intellectual property rights.
Furthermore, we are particularly concerned with the failure to address
the recommendations raised by the IAHC regarding the .US top level domain.

THE .US DOMAIN

The US Domain is an official top-level domain in the DNS of the Internet
community. It is administered by the US Domain Registry at the
Information Sciences Institute of the University of Southern California
(ISI), under the Internet Assigned Numbers Authority (IANA).
(above per http://www.isi.edu/in-notes/usdnr/usdom-overview.html)

The "Final Report of the International Ad Hoc Committee: Recommendations
for Administration and Management of gTLDSs" (available at
http://www.iahc.org/draft-iahc-recommend-00.html) states that:

"Current difficulties with gTLDs are highly exacerbated by inadequate
use of the .us TLD and the IAHC requests that scaleable and functional
Second-Level Domains (SLDs) for .us be defined and used, in the spirit
of those that already exist for many other ISO 3166-based TLDs."

Furthermore, the IAHC report states:

"The IAHC suggests that the domain administrator for .us undertake
further delegations along the lines of the functional SLD structure,
suitably modified to account for large-scale delegation and management,
and suggests this type of functional structure to other ISO 3166
national name space administrators as a widely understood structure."

Why has this issue not been discussed further either by the
US Domain Registry and IANA (both at the Information Sciences Institute
of the University of Southern California) and/or gTLD-MoU Policy
Oversight Committee ?
Would dealing with the .US top level domain provide the community
with additional time to deal with the issues of domain name
management under the broader umbrella of global Internet governance ?

This is but one example that we may be moving too fast without fully
discussing previous recommendations.

###

Number: 329

From: Gordon Irlam <gordoni@base.com>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 8:24am
Subject: Comments of Gordon Irlam

Document available at:
http://www.base.com/gordoni/thoughts/ntia-dns-comments.html

Before the
U.S. DEPARTMENT OF COMMERCE
National Telecommunications and Information Administration
Washington, DC 20230

In the Matter of ) ) REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01 INTERNET DOMAIN NAMES )

Comments of Gordon Irlam.

gordoni@base.com

Gordon Irlam
2310 Rock St. #10
Mountain View, CA 94043

18 August 1997

Gordon Irlam respectfully submits comments in this proceeding. Gordon Irlam
is employed as a software developer. He is a domain name registrant and
operator of a DNS server. He has been using the Internet for 12 years,
participates in the IETF, is a member of the W3C Advisory Council, the
Association for Computing Machinery, and the American Economic Association.

----------------------------------------------------------------------------

Table of Contents

A. Appropriate Principles

* Additional principles

B. General/Organizational Framework Issues

* Questions 1-8

C. Creation of New gTLDs

* Questions 10-13

D. Policies for Registries

* Questions 15, 16, 19, 20

----------------------------------------------------------------------------

A. Appropriate Principles

The Government seeks comment on the principles by which it should
evaluate proposals for the registration and administration of
Internet domain names. [...]

The following additional principle is suggested:

Domain name registrants should acrue the wealth they have created.
Domain name registrars should merely be able reap the value of the
registration services they perform.

It is the domain name registrant that performs the vast majority of
investment resulting in the creation of wealth associated with a particular
domain name. It is important that domain name registrants be able to reap
the value of this wealth. Any policy that enables domain name registrars to
reap the wealth associated with investments made by domain name registrants
would create economic inefficiencies. These economic inefficiencies would
manifest themselves as under-investment in the creation of domain name
content, and an over-investment in domain name registration services.

Two implications of this principle are:

* For the duration of use of a particular domain name, the domain name
should essentially be viewed as the property of the domain name
registrant.
* Domain name owners should not be held captive to particular registrars.
Registrants should be able to switch registrars at will. This is
necessary to ensure the registrant and not the registrar can captures
the wealth created by the registrant.

----------------------------------------------------------------------------

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain
name registration systems?

The registration system for the .COM, .ORG, and .NET domains have the
advantage of being part of a single system, with a single common set of
procedures for the registration of domains, payment of bills, and querying
the status of registrations.

The registration system for the .US domain is highly fragmented rendering it
effectively unusable. Different sub-domains are administered by different
authorities. These different authorities have different policies and
procedures. There is no central database that can be searched to access
information on registrations within the US domain. For many subdomains it is
not even possible to determine which domain names are already allocated. It
is even difficult to determine the appropriate procedures and authority to
contact to register within a particular domain.

The registration systems for most of the other ISO country code domains
besides .US also have various problems. However each of these registration
systems is ultimately under the control of the respective national
governments, and it considered important that national governments are
offered the opportunity to play a role in the provision of domain
registration services.

All the present registration systems have the disadvantage of a lack of
competition in the provision of registration services for the underlying
domain names. This lack of competition results in registration services
being provided inefficiently. A second problem resulting from the lack of
competition is prices need not being set based on the cost of the underlying
services provided. Today this is not as serious problem. Most of the domain
registration systems are under some form of regulations that limit the
prices that can be set. However, were such regulations to be removed,
without provisions for competitive registration within these domains,
serious problems would result. Domain name registrars for particular TLDs
would be able to set almost arbitrary prices for the continuance of
registration within particular TLDs. The structure of the World Wide Web
makes it impossible for a registrant to switch to a different domain name
without loss of almost the entire value built up around a preexisting
domain. Consequently, registrants would have little alternative but to pay
what would essentially be arbitrary monopoly fees set by domain name
registrars. This situation can only be prevented through either the
continued regulation of prices, or the introduction of competition in the
provision of registration services within TLDs.

2. How might current domain name systems be improved?

The provision of registration services could be improved by having multiple
competitive registries for the provision of registration services within
existing domains. This would increase the efficiency of the registration
process. More importantly, it would ensure registrants, rather than
registrars, are able to capture the value of the wealth created by the
registrants investment in a particular domain. This is especially important
if the existing regulatory limits on the fees charged for domain name
registration were to be removed.

The operation of the root name servers could be improved through a
competitive biding process to determine the operators of the root name
servers. An example scenario follows. Registrars could be required to pay an
annual fee based on the number of domains for which they are currently the
registration agent. This fee would be passed on to domain name registrants,
and should be of the order of $1-2 per domain per year. The fees collected
would be used to pay the costs associated with the operation of the root
name servers. Contracts to operate a subset of the root name servers for a
three year period could be competitively bid on a yearly basis. This way,
only a third of the root name servers would undergo operational transitions
each year, ensuring high operational stability of the DNS infrastructure.
For operational reasons, IP addresses associated with the root name servers
need to pass from one contractor to the next, and must not be considered the
property of the incumbent contractor.

3. By what entity, entities, or types of entities should current
domain name systems be administered? What should the makeup of
such an entity be?

We don't feel any existing entity is presently ideally suited. However,
given the overwhelming importance of operational stability, and the urgency
of the present matter, we feel the framework provided by the gTLD MOU that
was created by the IAHC represent by far the best of the available options.

Desirable aspects of the gTLD MOU include the provision of competitive
commercial services acting under a framework of public trust, effective self
governance by bodies including IANA with a proven track record in operating
the Internet infrastructure, and ultimately this governance overseen by
appropriate international bodies.

4. Are there decision-making processes that can serve as models
for deciding on domain name registration systems (e.g., network
numbering plan, standard-setting processes, spectrum allocation)?
Are there private/public sector administered models or regimes
that can be used for domain name registration (e.g., network
numbering plan, standard setting processes, or spectrum allocation
processes)? What is the proper role of national or international
governmental/non-governmental organizations, if any, in national
and international domain name registration systems?

The role of the FCC in regulating the provision of long distance phone
service serves as a reasonable model of the necessary role for governmental
or non-governmental organizations in creating the framework for the
provision of DNS services. Desirable features of the regulation of long
distance phone service include:

* Long distance service is provided by commercial organizations within a
competitive framework established by the government and administered by
the FCC. Similarly, DNS registration should be provided competively,
but under a framework established and regulated by appropriate
governmental or non-governmental organizations.
* Phone subscribers have the ability to change long distance providers at
will. Analogously, domain name registrants should have the ability to
change registrars at will.
* Long distance companies do not have monopoly ownership of the right to
provide long distance service to particular blocks of phone numbers.
Similarly, registrars should not have a monopoly on the provision of
registration services to those domains within a particular TLD.
* It is possible to keep the same phone number when changing long
distance providers. Analogously, domain name registrants should be able
to keep the same domain name when changing registrars. A phone numbers
are treated as belonging to the subscriber, not to the long distance
company.
* The competitive provision of long distance phone service is not all
encompassing. It interoperates with the regulated local phone services,
and the phone services regulated by foreign governments. Similarly, the
competitive provision of services for the .COM, .ORG, and .NET domains,
should be able to coexist with regulated provision of the .EDU, and
.GOV domains, and the various different ISO country code domains.

5. Should generic top level domains (gTLDs), (e.g., .com), be
retired from circulation? Should geographic or country codes
(e.g., .US) be required? If so, what should happen to the .com
registry? Are gTLD management issues separable from questions
about International Standards Organization (ISO) country code
domains?

The generic top level domains should be preserved. The regulatory and
administrative issues at hand have nothing to do with whether domain names
exist within a gTLD or an ISO country code domain. The present issues are
most significant for the gTLD domains simply because the gTLD domains are by
far the largest and most heavily commercialized of the TLDs. Therefore, it
is in the gTLD domains that the pressures for transition to competitive
market based registration services, as well as the risks of monopoly
control, are the greatest.

On a long term basis, as ISO country code domains grow, they too would
likely benefit from a competitive management structure. Consequently, it
would be best to develop a general purpose system for the competitive
registration of domain names. Initially this system would only contain the
.COM, .ORG, and .NET domains. However, in the future we should imagine the
possibility of national governments wishing to delegate registrations for
their ISO country codes through this same system. The system should be
designed to allow for this. Doing so does not pose any technical problems.

6. Are there any technological solutions to current domain name
registration issues? Are there any issues concerning the
relationship of registrars and gTLDs with root servers?

A clear distinction needs to be made between administrative and operational
activities. The provision of registration services is an administrative
function. The operation of the root name services is an operational
function. There is a serious danger of monopoly control of the Internet when
a single party is free to perform both administrative and operational
activities.

Domain name registrars should be prohibited from operating root name
servers, and vice-versa.

Fundamentally, it is the operators of the root servers that have control of
the Internet name space. To ensure that they don't abuse this power it is
necessary to ensure that the operators of the root name servers treat the
data provided by each registrar equally. If the root name servers only used
data provided by a particular registrar, that registrar would be able to set
monopoly prices for domain name registration services.

7. How can we ensure the scalability of the domain name system
name and address spaces as well as ensure that root servers
continue to interoperate and coordinate?

To ensure the scalability of DNS, continued investment in the development of
the BIND DNS software should occur. Additional root name servers should also
be deployed. These investments could might be funded from the NSF
Infrastructure Fund established under the cooperative agreement.

Rather than occurring on an ad hoc voluntary basis, operation of the root
name servers could be performed under competitive contract, with contractual
requirement to ensure root name servers continue to interoperate and
coordinate.

8. How should the transition to any new systems be accomplished?

The NSF needs to act in a timely fashion to ensure the complete contents of
the existing domain name databases are publicly available on the expiration
of the cooperative agreement. Article 10(E) of the cooperative agreement
provides the NSF can request the necessary information from Network
Solutions as part of the final report to be submitted by Network Solutions
to the NSF. It is important that the NSF request this information because it
will be required by any successor organization.

The initial registrars for each existing domain name should be allocated
evenly and randomly between the initial set of qualified competitive
registries.

This entire transition process needs to be performed very quickly, in a
period of no more than 5-10 days. Consequently it is vital that operational
planning for the transition be occurring between all relevant parties today.

Under the terms of the existing cooperative agreement Network Solutions is
permitted to collect a fee of $50 per year payable in advance for each
registered domain, with an initial two year payment being required when a
domain is first registered. This fee is intended to cover the cost of
providing registration services for the following year, or two years.
Network Solutions is presently collecting the full amount of this fee
despite the cooperative agreement being scheduled to expire in 6 months
time. The unearned portion of these fees should either be returned to the
domain name registrars, or passed on to the successor registration body.
With over 1 million domains registered, the amount of money at issue is
substantial.

----------------------------------------------------------------------------

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations
that constrain the total number of different gTLDs that can be
created?

No hard restrictions on the number of gTLDs exist. However, great care and
prudence should be exercised. Once created, it is very difficult to change
or remove a gTLD. Consequently, gTLDs should only be created if there is a
pressing need.

11. Should additional gTLDs be created?

It is our perception that at the present time no new domains need be
created. This is not however a strongly held opinion.

Right now the primary focus should be on the management of the existing
gTLDs, and their transition to a competitive registration system. The
creation of new gTLDs is likely to only serve as a distraction to this
effort.

12. Are there technical, business, and/or policy issues about
guaranteeing the scalability of the name space associated with
increasing the number of gTLDs?

So long as all of the new gTLDs are administered within a single
registration system there are no problems of scalability associated with the
creation of new gTLDs.

If different gTLDs are administered under different administrative regimes,
then the number of different regimes should be kept to a minimum. End users
of the registration system can't be expected to know about and understand
all the different regimes, policies, and procedures. Creating multiple
regimes would likely result in problems similar to the present confusion
that exists regarding registration in the .US domain.

13. Are gTLD management issues separable from questions about ISO
country code domains?

Yes, gTLD management issues can be separated from the management of ISO
country code domains. Similar problems to those facing the gTLDs exist for
the ISO country code domains, but they are presently far less severe, and do
not require solving simultaneously. At a future date, the same mechanisms as
used to solve the problems associated with the registration of gTLD domains,
might be applicable to the ISO country code domains.

----------------------------------------------------------------------------

D. Policies for Registries

15. Should a gTLD registrar have exclusive control over a
particular gTLD? Are there any technical limitations on using
shared registries for some or all gTLDs? Can exclusive and
non-exclusive gTLDs coexist?

None of the gTLDs should be under the exclusive control of a single
registrar. Unless subject to government regulation of the fees charged for
registration services, the registrar would be in a position to abuse this
situation by imposing monopoly fees on domain name registrants. Such fees
would result in economic inefficiencies.

There are no technical problems to using shared registries for all gTLDs.
There are minor organizational, managerial, and deployment issues that still
need to be solved, but these issues are readily solvable.

Exclusive and non-exclusive domains could coexist, but the creation of
exclusive domains would be bad public policy. If for some reason exclusive
domains should continue to exist in the future, transition plans for their
eventually maturing and becoming shared should be drawn up, and regulations
should also be imposed to limit fees in the interim.

16. Should there be threshold requirements for domain name
registrars, and what responsibilities should such registrars have?
Who will determine these and how?

There should be a set of requirements on domain name registrars. These
requirements should be focussed on ensuring the stability of the domain name
registration system. Registrants should be contractually required to be able
to guarantee a set of operational requirements, and have appropriate funding
to underwrite this guarantee.

It is reasonable for the bodies established under the gTLD MOU, the PAB,
POC, and CORE, to determine such requirements.

19. Should there be a limit on the number of different gTLDs a
given registrar can administer? Does this depend on whether the
registrar has exclusive or non-exclusive rights to the gTLD?

All gTLDs domains should be a part of a single shared database. Registrars
should be able to administer all domains within all of the gTLD domains.
Allowing all registrars to act as registrars for all gTLDs will maximize
competition in the provision of registration services within each gTLD.

Exclusive domains should be phased out.

20. Are there any other issues that should be addressed in this
area?

It is important that appropriate contractual language exist to require
registration information be shareable between registrars. Registrars need to
disclaim all rights to the underlying registration information they have
collected, including all selection and compilation based copyrights, and any
future rights under the recently proposed database treaty. Registrars must
agree to provide the complete registration information for a particular
domain to a competitor in response to an authentic request from the domain
name owner to switch registrars.

----------------------------------------------------------------------------

###

Number: 330

From: "Lawrence W. Hillman" <sofar@compuserve.com>
To: "'Temp'" <dns@ntia.doc.gov>
Date: 8/18/97 7:12am
Subject: Domain Name Changes

I think the number of domain name "extensions" (.com, .gov, etc) should be
kept to a minium. Otherwise, we'll end up with many very similar names and
this will make them difficult to remember and reduce the identity value of
"trademark" names.

Larry Hillman

LHillman@HillcoRealty.Com

###

Number: 331

From: "Dave A" <daveanin@netlabs.net>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 9:43am
Subject: Keep with Network Solutions

I believe the Internet's naming should stay under one company, for
simplicity's sake...

Think about it, if I had 10 companies to choose from each offering different
services for DNS, such as .games, .home, .biz, .rec, etc... it would get way
out of hand... It should stay with .com, .net, .org, etc... it's easy, it's
established, and it works. Let's face it, other companies are just getting
greedy they can't enjoy profits from this system.. The Internet wasn't
designed for profits in its root structure anyway.

Leave NSI alone... Just because a lot of the good .com's are taken, doesn't
mean new businesses can't think of something creative for a domain name.

Dave Aninowsky
BNL Enterprises Inc.
31 Washington St.
Matawan, NJ , 07747
webmaster@pcwares.com

###

Number: 332

From: <Marktrade@aol.com>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 12:25pm
Subject: internet issues

FROSS ZELNICK LEHRMAN & ZISSU, P.C.
633 Third Avenue
New York, New York 10017

Telephone: (212) 953-9090
Facsimile: (212) 953-1037

August 18, 1997

VIA E-Mail
dns@ntia.doc.gov.

Office of Public Affairs
National Telecommunications and
Information Administration (NTIA)
Room 4898
14th Street and Constitution Avenue, NW
Washington, D.C. 20230

Attention: Ms. Patrice Washington

Reference: Request for Comments on Internet Issues

Dear Sir:

I am writing on behalf of the Intellectual Property
Law Section of the American Bar Association.

The Section will not be submitting written comments
in response to the request for comments published on July 2,
1997. However, the Section strongly believes that the issues
raised are of the utmost importance for intellectual property
owners and members of the intellectual property bar. Accord-
ingly, and in view of the summer season during which many
interested persons are away for periods of time, the Section
respectfully requests that public hearings be conducted to
insure full public review and analysis of these issues.

Respectfully submitted,

MDE/gf Mark D. Engelmann
Chair - ABA IPL Section
Committee 203
PTO Affairs - Trademarks

###

Number: 333

From: "proteus" <proteus@tpa.net>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 12:25am
Subject: TLD FILTERING SOLVES YOUR DILEMNA

Subject: TLD FILTERING AND WHAT IT DOES FOR ALL OF US

To: The Internet Community and every human on the planet

Cc: vice.president@whitehouse.gov, president@whitehouse.gov,
first.lady@whitehouse.gov

TLD FILTERING AND HOW IT WORKS FOR ALL
OF US TO PROTECT OUR CHILDREN ANDPROTECT OUR RIGHTS.A NEW IDEA FOR AMERICA
AND
THE INTERNET COMMUNITY

TLD filtering is a simple and effective answer to a dilemma facing our
society.

The two conflicting issues are:

1. Should parents have the right to determine what content reaches their
children during their formative years.

2. Do adult Americans have the right to freedom of expression and
access
to ideas and content of their choice, as guaranteed by the First
and Fourth
Amendments of the Constitution?

With the unanimous ruling by the Supreme Court that the Communications
Decency
Act was unconstitutional and unenforceable in any event, there have
emerged a number of proposals to deal with this conflict.

These haveincluded rating systems, watchmen organizations, and
possible Federal involvement with some other aspects under their
jurisdiction.

At the same time, the internet community itself has responded to these
concerns with a desire to "police themselves" without Uncle Sam getting
involved
in censorship. Other countries are also not satisfied with recent United
States
gestures toward control of TLD (top level domain) and naming conventions,
that
they feel are questions that should be put to the world community instead of
being pre-empted by us.

The way TLD filtering works is simple.
ISP's can develop methods of offering their customers a menu of TLD's that
they
want to see or that theydo not want to see. Business corporations might want
to
filter out any TLD involved with sexual content. Adults wanting access to
sexual
materials could select FULL TLD * i.e., any and all TLD's that the ISP
might have. The advantages are obvious. Everyone wins. Parents can control
what
their kids see, adults can say and do what they wish.

The BBS community *ancestors of the Internet * always controlled access to
adult content by requiring proof of age. Many *. But not all .. ISP's do
this now.
Certainly one of the leaders in this area is MCS (Macro Computer
Solutions,http://www.mcs.net), lead by Karl Denninger who is
one of the proponents of an extended EDNS (Enhanced Doman Name System) .
Currrently not all ISP's carry the alternate domains, i.e., those presently
carried by the InterNIC. (Those very recognizable TLD's are .COM, .ORG,.EDU,
and
.NET. The other "official" TLD's are MIL and GOV, under Federal
control). But there are other registration authorities and other countries'
telecommunications agencies who oppose the InterNIC and its recently
proposed
expansion plans. The Internet community has largely rejected these
recommendations, and the issue is really up in the air. Two websites that
discuss this are http://www.edns.net andhttp://www.alternic.net and there
are others
as well. There are also a number of well-intentioned public service-oriented
organizations who are trying to address the problem with ratings and even
chip-controlled access. But these are not standardized, and one of the jobs
that
any kid has had has been finding a way to get around parental controls.

With this method, TLD filtering, it is impossible for any child to access
a TLD not on the list provided by his ISP. With this method, an adult could
also
filter out some groups, let in others, based on his ownsense of what is
right
and wrong being on the web.

TLD filtering is a way for us to have our cake and eat it too.

It is a win win for everyone.

Larry BuckmasterPresidentConsulTech of Chicago
Proteus@tpa.net
312/902-2776

As of July 29, 1997, at 3:55 pm CST, the phrase "TLD Filtering" did not
exist
on the AltaVista Search Engine in either the Web or Usenet. This document, I
believe, is the first discussion of it, and I am sending it to various
correspondents
to catalog it as a message with a timestamp.

No claim or credit or the concept is claimed; only a desire to chronicle
what
may be a simple idea with vast consequence to many people.

###

Number: 334

From: Geoffrey Atkin <lmga@millcomm.com>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 7:08am
Subject: comments on Internet Domain Names

My comments are in answer to section B, questions 1, 2, and 5, section
D, question 15, and section E, questions 22, 24, and 28.

Thank you for the opportunity to voice these opinions.
Geoffrey Atkin

The IAHC memorandum is not perfect, but it is the best solution anyone
has been able to come up with. I agree with its philosophy that generic
top-level domains should be administered by multiple registrars on the
public's behalf. I might quibble with their choices for new top-level
domains, but I understand that there was considerable debate on this
point. I agree that both Internet and international standards bodies
should be involved in the administration of generic top-level domains.

It is vital that the current NSI monopoly be ended, and the expiration
of their contract in March 1998 may be the only opportunity to do so.
Replacement registrars must be in place well in advance to ensure a
smooth transfer of power.

The major problems with the current system are (1) many people are
dissatisfied with the service provided, fees charged, and/or policies
set by NSI, and there is little recourse since NSI has a monopoly on the
most popular top-level domains; and (2) the explosive commercial growth
of the Internet has led to contention for "good names" in .com. The
latter is because domain names are absolutely unique world-wide, while
company names are not, even when trademarked. Also, NSI assigns names
without regard to trademark status; when disputes arise, a common
outcome is for the name to become unavailable to anyone.

I think the IAHC plan can solve the first problem, but it will not solve
the second. The rest of my comments are on the subject of trademarks.

Some have argued that trademark law does not apply to domain names. I'm
no lawyer, but I believe this view is totally without merit.

Domain names are intended to be mnemonic identifiers for a group of
computers owned or administered by a specific entity or organization.
Thus, the main practical use of domain names is to help Internet users
identify to whom they are sending information and from whom they are
receiving it. This is also one of the purposes of trademarks (as
applied to company names) and this is why domain names come into
conflict with trademarks.

I think the .com domain needs to be overhauled. In its place we need a
name space which is strictly for names of corporate entities, available
only to holders of national trademarks for those names, subdivided by
business and by country. For example,

chiquita.ag.us

It is necessary to separate by country because two unrelated companies
in different countries can have the same name, and can hold trademarks
to the name in their respective countries. The existing two-letter
national domains are an obvious starting point. We then need to
subdivide the .us domain by industry or business, because two companies
in the U.S. can hold national trademarks to the same name so long as
they are in different lines of business.

Here is one possible set of replacements for .com, based on Standard
Industrial Classifications (SIC) headings:

.ag.us Agriculture, forestry, and fishing (SIC 0xxx)
.cs.us Mining and construction (SIC 1xxx)
.mf.us Manufacturing (SIC 2xxx and 3xxx)
.tr.us Transportation (SIC 4xxx)
.pu.us Public utilities (SIC 4xxx)
.wh.us Wholesale trade (SIC 50xx and 51xx)
.rt.us Retail trade (SIC 52xx to 59xx)
.fn.us Finance, insurance and real estate (SIC 6xxx)
.sv.us Services (SIC 7xxx and 8xxx)

These subdivisions could also be applied at the state level, for
companies that do not hold a national trademark on their names. For
example, if Chiquita was not a national trademark, but merely a local
banana grower in North Carolina, it might be chiquita.ag.nc.us.
Ideally, the same registrars should be able to assign names in either
ag.nc.us or ag.us.

Whatever second-level names are chosen, I think the list needs to be
short enough to be understandable and long enough that most companies
with the same name wind up in different domains. (And of course, it has
to avoid conflicting with two-letter state abbreviations.)

We don't need to worry about other countries; they have their own
registries. As a point of interest, most have second-level domains such
as .co.uk and .co.jp, which serve as national versions of .com.

If such an overhaul is made, it would be sensible to encourage greater
use of the two-letter national domains, and it might be reasonable to
close .com to new registrations. But there is no reason to eliminate it
altogether, and it would be highly disruptive to do so.

The Department of Commerce could certainly play a helpful role. For
example, it could provide domain registrars with the data necessary to
check for national trademark conflicts before assigning a domain name.
This would help prevent trademark disputes, which would benefit everyone
except possibly the trademark lawyers.

Geoffrey Atkin

###

Number: 335

From: "Christopher Hadden" <webmaster@jasc.com>
To: "'dns@ntia.doc.gov'" <dns@ntia.doc.gov>
Date: 8/18/97 11:12am
Subject: Comments on Domain Names

To whom it may concern:

Inter-network communications is a very open thing by design. I do not believe that Network Solutions, Inc. should retain it's monopoly as the domain name registrar for top level domains such as .com and .net.

I welcome some of the ideas contained in the International Ad Hoc Committe's proposal including the addition of top level domains. I also believe that problem resolution of domain name disputes should be handled at an international level.

If I may be of any help in this discussion process, please let me know.

Sincerely,

Christopher Hadden

Webmaster, Jasc Inc.
webmaster@jasc.com
http://www.jasc.com

###

Number: 336

From: <BDarcy@lga.att.com>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 1:34pm
Subject: AT&T's Response to NOI on Domain Names

Attached is AT&T's response to NOI on Domain Names

CC: NTIADC40.SMTP40("rosenblu@lga.uucp","kleinman@lga....

Before the

DEPARTMENT OF COMMERCE

NATIONAL TELECOMMUNICATIONS

AND INFORMATION ADMINISTRATION

Washington, D.C.

Request for Comments on the                  )
Registration and Administration              )       Docket No. 970613137-7137-01
of Internet Domain Names                     )

COMMENTS OF AT&T CORP.

Pursuant to the Public Notice released July 2, 1997, AT&T Corp. ("AT&T") respectfully submits its Comments to the National Telecommunications and Information Administration ("NTIA") on the questions presented concerning the registration and administration of Internet domain names. AT&T commends NTIA's leadership on the these issues and welcomes the opportunity to provide its comments. AT&T also reaffirms its willingness to work actively with the U.S. Government, the International Ad Hoc Committee("IAHC"), Policy Oversight Committee ("POC"), Network Solutions, Inc. ("NSI"), and the industry at large to help define an approach that can achieve broad domestic and international consensus, and be implemented in a practical and timely way.

AT&T responds below to each of the questions posed in the Public Notice, as enumerated in that Notice.

A. Appropriate Principles

a. Competition in and expansion of the domain name registration system should be encouraged. Conflicting domains, systems, and registries should not be permitted to jeopardize the interoperation of the Internet, however. The addressing scheme should not prevent any user from connecting to any other site.

AT&T supports this principle.

b. The private sector, with input from governments, should develop stable, consensus-based self-governing mechanisms for domain name registration and management that adequately defines responsibilities and maintains accountability.

AT&T supports this principle.

c. These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.

AT&T supports this principle. In addition, any new mechanism must be scalable and thus be able to grow with the Internet.

d. The overall framework for accommodating competition should be open, robust, efficient, and fair.

AT&T supports this principle. AT&T recognizes the importance of clearly defined transition plans and operations plans in evolving to a new framework.

e. The overall policy framework as well as name allocation and management mechanisms should promote prompt, fair, and efficient resolution of conflicts, including conflicts over proprietary rights.

AT&T supports this principle.

f. A framework should be adopted as quickly as prudent consideration of these issues permits.

AT&T supports this objective. AT&T urges the U.S. Government to ensure that appropriate protections for trademark owners are thoroughly considered and included in the framework.

B. General/Organizational Framework Issues

B.1. What are the advantages and disadvantages of current domain name registration systems?"

The main advantage is that it works. There have been complaints about the current registrar, and some of the complaints have merit. On the other hand, the current registrar has grown the system from a small start to a system that has now registered well over one million names in ".COM," and it has built up expertise and support systems to handle the current load of new requests. Any replacement must be able to handle the current load as well as the expected rapid growth without major disruption.

The main disadvantage is the monopoly position of the current provider. Once the current NSF Cooperative Agreement ends, there are no checks on the registrar. The ".COM" domain is very well known to Internet users and it is thus at this time the "place to be" for anyone wanting to do business on the net. If the registrar changed the rules to exclude arbitrarily some businesses or greatly increased the rates, most name holders would have no alternative but to live with the change.

B.2. How might current domain name systems be improved?

Because of the potential for abuse of NSI's monopoly position to the detriment of Internet service providers and their customers worldwide, as noted above, AT&T recommends evolving from the current monopoly system to a structure that accommodates multiple domain name registrars. However, there are a number of operational and trademark issues that arise under any new proposal for multiple domain name registrars distributed throughout the world. AT&T thus recommends significant further dialogue with key stakeholders prior to the implementation of any specific proposal.

There are several other improvements that AT&T recommends be undertaken within the current or any new structure. For instance, one of the major improvements would be increasing the security of the system. The recent situation where the AlterNIC redirected requests so that people who wanted to access the InterNIC were sent to the AlterNIC instead shows the vulnerabilities of the current system. The IETF has done work on Secure DNS; these features should be implemented on root servers and other DNS operators should be encouraged to use them as well. The code exists, is exportable, and is free of restrictions on its use. Use of this technology is vital to prevent further hijackings.

B.3. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

AT&T believes there is merit in the concept contained in IAHC's proposal for multiple domain name registrars throughout the world. Any organization effecting its implementation has to be broadly based. Such organization must include representatives of the user community including trademark owners, the providers, and the server sites. Furthermore, it must be international. The current DNS works because most sites agree that it should. If too many ISPs disagree, there won't be a single DNS, which would be catastrophic to the Internet. The technology differs in a major way from the telephone numbering system, where the numbering plan is implemented in an inherently centralized fashion.

B.4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

There are lessons to be learned from certain aspects of telephone number administration process, though we must be conscious of significant differences as noted above. Examples of processes from which we can gather some insights are US 800/888 number assignment, assignment of new NPAs in the US, the current local number portability processes being discussed at NANC (North American Numbering Council), assignment of telephony country codes and assignment of two- and three-letter country codes by ISO. Any learning from these processes can not be simply replicated for the Internet. New processes that benefit from the learning should then be tailored to the different environment of Internet domain names.

B.5. Should generic top level domains (gTLDs), (e.g., .COM), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the .COM registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

While a proposal that all domain names be under a specific country code has appeal, such a system would pose numerous problems. The main one is that it may be too late. There are over 1.3 million names in ".COM" already, and approximately 60,000 more in each of ".NET" and ".ORG." Many of these names are well known to Internet users, and in many cases businesses have spent significant amounts of money promoting recognition of their name. To change all these names at this point would cause massive confusion among users and inhibit the growth of commerce on the Internet. In all events, eliminating ".COM" is not a practical or realistic alternative. Even if it were possible, many of the same technical and administrative issues would arise in the ".US" domain, because ".COM" consists overwhelmingly of U.S. entities today.

Furthermore, there are multinational companies that are not easily associated with any one country code. For example, it is not obvious that Nestle would be associated with ".CH" (Switzerland), or Shell Oil (Royal Dutch Shell) with ".NL" (Netherlands). These companies would have to register in almost every country code if they were permitted to do so by the registrar for that country code domain. Such a process would be unnecessarily costly and inefficient.

B.6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

The registration issues are not primarily technical. The root name servers can handle millions of TLDs if necessary (the ".COM" domain is currently held by the root name servers and has over a million entries). The technology to support shared registries exists; many commercial relational database management systems have the required features. Work would have to be done to implement the system, but it is technically feasible.

The relationship of registrars and gTLDs is not constrained by technical limitations. The formal relationship between multiple registrars and multiple root name server operators would have to be worked out in several dimensions. Significant substantive work needs to be done on many related legal and organizational issues.

B.7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and coordinate?

There are technical limits to the number of possible servers for any single domain, including the root and ".COM." These arise from packet size constraints in the current DNS, and limitations on the bandwidth necessary to download backup copies of the zone files.

These Comments specifically apply to addresses as well as domain names. There are some fundamental technical concerns about the current IP address space. First, it's limited, and will remain limited until IPv6 becomes widespread. This is already posing problems for many companies. The only long-term solution is deployment of IPv6.

A second issue is "provider-based addressing." This is more properly known as "topology-based addressing" -- your IP addresses must be given to you by your upstream provider. Although there are unfortunate consequences of this scheme -- in particular, it's hard for a site or ISP to connect to multiple providers, or to change upstream providers if necessary -- the requirement is due to limits on the memory sizes of today's routers, and the amount of CPU time necessary to calculate paths through ever larger routing tables than are used today. The industry requires automatic renumbering technology.

B.8. How should the transition to any new systems be accomplished?

Any proposed new system should have an associated transition plan that spells out how the migration from the current system will be carried out. The plan should include schedules, milestones, contingencies if dates cannot be met, test plans (including capacity tests), and criteria for meeting the milestones. The plan should specifically address the scale of the current registry and show how all current domain name registrations will be handled. The plan must indicate what is expected of the current registrar during the transition. Because the current registrars have indicated that they feel they have ownership rights to the current data, the plan must indicate the terms and conditions pursuant to which the new registrar will be able to obtain and use the data. As part of the plan an ongoing test system should be maintained to provide rigorous testing of new registrars and new versions of software that may affect the central database of TLDs.

The plan should be made available to the Internet community and substantive comments should be addressed.

B.9. Are there any other issues that should be addressed in this area?

Other issues will certainly arise; this series of questions is adequately comprehensive at this time.

C. Creation of New gTLDs

C.10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

There are no significant technical limits; i.e., the technology can probably handle any number that is operationally feasible. There are clearly trademark-related implications as new gTLDs are created (for example, ".COM" and ".STORE" for the same company name).

C.11. Should additional gTLDs be created?

Yes. To support the growth of the Internet and electronic commerce, the industry must ensure that domain names capacity is increased.

C.12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

There is no significant technical limit. The business and policy issues require further exploration.

C.13. Are gTLD management issues separable from questions about ISO country code domains?

Yes, and they should be separated.

C. 14. Are there any other issues that should be addressed in this area?

The recent NSI outage shows the dangers of a single point of failure. Research and engineering should be done to add redundancy to the generation of all gTLDs, including ".COM," ".NET" and others, and eliminate single points of failure wherever possible. The planning effort should address both software and operational processes.

D. Policies for Registries

D.15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

All gTLDs must be served by multiple registrars, to promote competition. There are no inherent technical limitations on using shared registries, but implementation options need to be well tested.

D.16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have? Who will determine these and how?

AT&T suggests certain threshold requirements applicable to domain name registrars. For example, all registrars should be bound by a common set of rules, submit to the jurisdiction of a single, pre-determined court and post a bond to satisfy judgments. Measures should be taken to ensure the smooth operation and transition to other registrars, particularly where a registrar has become bankrupt or has ceased to do business. Underlying these obligations would be full disclosure of the registrar's credentials and compliance with these requirements, so that the holders of domain names have the information needed to make a reasonable choice of registrar.

D.17. Are there technical limitations on the possible number of domain name registrars?

Any reasonable number of registrars can be supported technically, subject to limitations noted in the answer to B.7. The operational and policy issues will dictate limits before the technical considerations will.

D.18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

Trademark issues are a major concern as the number of registrars increases. See AT&T's response to the questions in Section E. Operational impacts of increasing the number of registrars must be well understood and resolved prior to implementation.

D.19. Should there be a limit on the number of different gTLDs a given registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

At this time, AT&T does not foresee the need for a limitation on the number of gTLDs that a given registrar can administer. AT&T does anticipate a need to revisit and examine the issue based on the registrar qualification process and the robustness of operational and trademark protection processes that need to be in place for effective sharing of registration for multiple gTLDs.

D.20. Are there any other issues that should be addressed in this area?

Other issues will certainly arise; this series of questions is adequately comprehensive at this time.

E. Trademark Issues

E.21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-a-vis domain names?

All trademark rights, including service and certification mark rights, trade dress rights and company name rights, should be protected on the Internet vis-a-vis domain names. Moreover, sufficient measures should be taken to avoid public confusion.

E.22. Should some process of preliminary review of an application for registration of a domain name be required, before allocation, to determine if it conflicts with a trademark, a trade name, a geographic indication, etc.? If so, what standards should be used? Who should conduct the preliminary review? If a conflict is found, what should be done, e.g., domain name applicant and/or trademark owner notified of the conflict? Automatic referral to dispute settlement?

There should not be a requirement of preliminary review of an application for registration of a domain name other than review by an entity such as the proposed Administrative Challenge Panels (ACP).

E.23. Aside from a preliminary review process, how should trademark rights be protected on the Internet vis-a-vis domain names? What entity(ies), if any, should resolve disputes? Are national courts the only appropriate forum for such disputes? Specifically, is there a role for national/international governmental/nongovernmental organizations?

AT&T has developed a continuum of concepts and proposals. AT&T is prepared to discuss these (and other parties') proposals in future fora before the U.S. Government and industry groups.

An entity such as the Administrative Challenge Panels would decide whether the domain name should be registered. The standard which should be applied by the panels is "identical or confusingly similar to or dilutive of any demonstrable trademark right". The principles of stare decisis, res judicata and collateral estoppel should apply. Moreover, the losing party should be allowed to file a de novo action locally. In addition, a reasonably long period of time (five years) should be set aside in which any party that believed it was being damaged could petition to cancel a domain name.

E.24. How can conflicts over trademarks best be prevented? What information resources (e.g., databases of registered domain names, registered trademarks, trade names) could help reduce potential conflicts? If there should be a database(s), who should create the database(s) ? How should such a database(s) be used?

It is unrealistic to believe that conflicts over trademarks are preventable. However, if there is a legitimate conflict, AT&T supports the individual parties being offered the opportunity to craft a workable solution. For example, if a trademark registrant of ACME for food services applies to register a domain name of ACME and another trademark registrant or owner of demonstrable trademark rights of ACME for toys opposes, the parties could use the Dispute Mediation Procedures so that the first party registers "ACME-FOOD.COM" and the second party registers "ACME-TOYS.COM." This is consistent with the current settlement of trademark disputes parties limit their use of a trademark to their respective goods or services by agreement. However, where there exists a well-known, strong mark such as AT&T, no third party other than the trademark owner should be able to use that mark. Principles of dilution and unfair competition should be used to determine the scope and breadth of a mark's protection.

It is AT&T's view that the creation and maintenance of trademark databases would be overly burdensome, costly and difficult to administer.

E.25. Should domain name applicants be required to demonstrate that they have a basis for requesting a particular domain name? If so, what information should be supplied? Who should evaluate the information? On the basis of what criteria?

No. Domain name applicants should not be required to have a basis for requesting a domain name. However, the number and nature of domain name registrations owned by an individual or entity should be considered and should be a strong factor in determining the good faith of a potential domain name applicant. If, for example, an applicant holds scores of other domain names which are identical or confusingly similar to the names or marks of various third parties, that factor alone should create a strong presumption, in any litigation or dispute resolution procedure, that the applicant is attempting to traffic in trademarks, is not acting in good faith and thus not entitled to domain name registrations.

E.26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

As the number of different gTLDs increases, so too does the risk of having to challenge a domain name applicant. The number of disputes and costs will increase substantially as the number of gTLDs and registrars increase. Moreover, trademark owners will face the daunting and expensive task of monitoring domain name registrations for each gTLD. Also, registrars might require that challenges be conducted in a country across the globe. This will certainly increase costs to a domain name challenger and will be particularly burdensome to owners of well-known marks and to small companies that are really primarily national or regional in scope of business, but to whom their brand is of significant value.

E.27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

Directory servers can help people find Web sites in ways other than guessing at "www.company-name.COM." But that solution is not workable for other uses of domain names, such as e-mail. In other words, the industry can ameliorate the contentiousness of the trademark battle by deploying directory servers, but the industry cannot eliminate it -- uses of the DNS by programs (as opposed to humans typing the names) are not compatible with any sort of vague, interactive lookup process.

Requiring the use of geographic and/or subject area qualifiers in domain names, to parallel the legal separations, is unworkable. Apart from the fact that most people can't and won't remember such information, the resulting domain names would be far too long to be used.

E.28. Are there any other issues that should be addressed in this area?

Yes. AT&T discusses below specific issues with IAHC/MOU proposals:

1. The major problem confronted by trademark owners is "cybersquatting", or the extortion and piracy of domain names that are identical or confusingly similar or dilutive of a prior valid trademark right. (For example, a third party other than AT&T registers "1800CALLATT.COM" and offers to sell the domain name to AT&T.) Another major problem is the use of domain names that are identical or confusingly similar or dilutive of a trademark. (For example, a third party other than AT&T registers "AT-T.COM" or "AT-TWEB.COM." The current IAHC/MOU proposals do not adequately address these problems. It has been stated that the ACPs have a limited function. However, litigating in court is extremely expensive, especially with a proliferation of cybersquatters. This problem needs to be addressed by IPOC and a compromise solution found that is acceptable to all interested parties.

2. Disputes would be determined by Administrative Challenge Panels international experts in the field of intellectual property law and Internet domain names. It is not clear who would serve on these panels, how panelists would be compensated or what their qualifications would be. There is a need to address this issue in greater detail.

3. The IPOC Guidelines indicate that to be "internationally known" a mark must be registered in a least 35 or more countries in at least four geographic regions for the same goods or services; to be "globally known," the mark must be registered in 75 or more countries for the same goods and services. These are extremely high thresholds, and will not protect a majority of trademark owners. Enormous amounts of money are invested in promoting trademarks that do not meet these thresholds. Such trademarks may have significant commercial appeal, and thus their owners would enjoy brand equity in such names. However, there is nothing in the current proposals to protect the interests of these trademark owners. For example, a medium size business has one trademark in the U.S. that is well-known in its industry and in the U.S. and all European Community countries. (This business also owns a U.S. trademark registration that is incontestable and trademark registrations throughout the European Community.) The current proposals do not afford protection to this trademark owner.

4. As proposed, the Administrative Challenge Panels will not have jurisdiction to decide disputes beyond the narrow question of preventing in appropriate circumstances, the registration of "a second-level domain name in any of the CORE gTLDs which is identical or closely similar to an alphanumeric string, that for purposes of this policy is deemed to be internationally known, and for which a demonstrable intellectual property right exists . . ." (Articles 2 and 3 of the MOU, emphasis supplied). Industry has revised the question of deciding disputes respecting all valid trademark rights. There is sentiment that all entities who have rights should be allowed to challenge, not just the owners of internationally known marks. Also, significantly, some domain names could be considered an infringement of a trademark and still not be considered identical or closely similar. The litmus test should be the internationally accepted standard: likelihood of confusion.

5. Difficult jurisdictional issues arise. The domain name registrant is located in Country A, the trademark owner in Country B, the registrar in Country C. The problem is the domain name registrant may "forum shop." To alleviate this problem, domain name applicants should agree to be subject to jurisdiction in a specified choice of venues.

6. The IPOC Guidelines indicate the Administrative Challenge Panel process is subject to be reopened by any person, and unknown parties may thus have the right to intervene. If an infringer is dissatisfied with the outcome of the Administrative Challenge Panel, it would initiate an action in a country whose laws may be favorably disposed to such party and force trademark owners to engage in exposure for such litigation. This and other issues related to the finality of arbitration rulings require further discussion.

7. The 60-day publication period is not mandatory and there is no procedure for opposing a domain name. There should be a 60-day mandatory publication period in which any party who believes it would be damaged by the domain name could oppose. This would also be favorable to the domain name registrant, because obstacles to use of the domain name would be identified prior to the domain name holder investing time and money in such domain name and corresponding website. AT&T recognizes that the mandatory waiting period was removed from the Memorandum of Understanding due to opposition, but believes the concept has value.

8. Sui generis protection for domain names, essentially creating a wholly new form of intellectual property protection to protect domain names, is unnecessary and will prejudice existing trademark owners. It has been proposed the use of a domain name for a two-year period on the Internet will create a right in the domain name. This is unacceptable. Trademark rights are only created by substantive law.

9. There should not be an opt-out clause in the domain name application that permits domain name holders to opt-out of binding arbitration. This would effect a clear and significant prejudice to trademark owners.

10. The Administrative Challenge Panel may, on its own initiative, decide to modify an TLD in lieu of excluding it. This may weaken a trademark owner rights in a mark and should not be permitted.

11. It is AT&T's view that the IPOC, which governs policy issues in the domain name process, has insufficient representation by trademark owners. While other organizations (like ISOC, CORE and IAB) are represented by two members, INTA and WIPO, ostensibly the only representatives of the trademark community, are allotted one member each. This must be remedied in order to result in balanced representation.

CONCLUSION

AT&T appreciates the opportunity to provide its views on this important subject, and looks forward to continuing participation in this and other fora dealing with the critical issues involving proliferation of new technology-based services in the U.S. and worldwide.

Respectfully submitted,

AT&T CORP.

By /s/ Ava B. Kleinman

Mark C. Rosenblum
Ava B. Kleinman

Its Attorneys

Mahal Mohan

295 North Maple Avenue
Room 3252J1
Basking Ridge, New Jersey 07920
(908) 221-8312

August 18, 1997

###

Number: 337

From: Glenn B. Manishin <glenn@technologylaw.com>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 9:17am
Subject: CPSR Comments on Internet Domain Names

On behalf of the Computer Professionals for Social Responsibility,
attached in Microsoft Word 6.0 format is CPSR's electronic submission in
response to NTIA's request for public comment on registration and
administration of Internet domain names. A hard copy of the same is
being delivered today to Patrice Washington of NTIA's Public Affairs
office.

Please contact me at this e-mail address or by telephone at 202.955.6300
if you have any questions in this regard

Sincerely,

Glenn B. Manishin

Counsel for CPSR

+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
   Glenn B. Manishin             glenn@technologylaw.com    
   Tel/Vms 202.955.6300                    gbm@clark.net
   Fax 202.955.6460              gmanishin@earthlink.net
   -------------    Washington, DC, USA    -------------
   Firm:     http://www.technologylaw.com/
   Personal: http://www.clark.net/pub/gbm/
   F1:       http://www.clark.net/pub/gbm/F1/f1intro.html
+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+

CC: "Glenn Manishin" <glenn@technologylaw.com>

Before the

UNITED STATE DEPARTMENT OF COMMERCE

NATIONAL TELECOMMUNICATIONS AND

INFORMATION ADMINISTRATION

Washington, D.C. 20230

In the Matter of                     )
                                     )
Request for Comment on Registration  )       Docket No. 970613137-7137-01
and Administration of Internet       )
Domain Names                         )

COMMENTS OF THE

COMPUTER PROFESSIONALS FOR SOCIAL RESPONSIBILITY

Aki Namioka, President Glenn B. Manishin
Harry Hochheiser Michael D. Specht
Andy Oram Christine A. Mailloux
Computer Professionals for Social Blumenfeld & Cohen - Technology Law Group
Responsibility <http://www.technologylaw.com>
<http://www.cpsr.org/home.html> 1615 M Street, N.W., Suite 700
P.O. Box 717 Washington, D.C. 20036
Palo Alto, CA 94302 202.955.6300
415.322.3778
Counsel for CPSR

Dated: August 18, 1997

The Computer Professionals for Social Responsibility (CPSR) <http://www.cpsr.org/ home.html>, by their attorneys, submit these comments in response to the Notice <http://www.ntia.doc.gov/ntiahome/domainname/dn5notic.htm> released by the National Telecommunications and Information Administration (NTIA) <http://www.ntia.doc.gov> soliciting public input on the present and future systems for registration and administration of Internet domain names.

SUMMARY

The current controversy over the Internet's Domain Name System (DNS) raises important questions about how the Internet, as a decentralized, global medium, should be administered and governed. While much of the Internet was invented and originally funded by the US Department of Defense and the National Science Foundation (NSF) <http://www.nsf. gov/>, and although its governing institutions were initially established and sanctioned by the American government, the Internet's technical standards and basic policies have in fact been set by a number of ad hoc, consensus-based consortia comprised of Internet service providers, engineers and users. This system worked for a long time because of the shared goals and technical sophistication of the Internet's original academic-based user community. See And How Shall the Net Be Governed?, by David R. Johnson & David G. Post <http://www.cli.org/ emdraft.html>.

But increasing commercialization and explosive growth have begun to strain the consensus-driven process of Internet administration. The strong and widely publicized reactions of many providers and users (and foreign governments) against the Memorandum of Understanding (MOU) developed by the Internet Ad Hoc Committee (IAHC) <http://www.iahc.org> for revisions to Internet domain name registration and administration <http://www.gtld-mou.org/> show that the stakes are high and that more open, considered and perhaps formal mechanisms for Internet self-governance -- and for evaluating alternative DNS proposals -- are called for in the present environment. In these comments, CPSR focuses on proposals for reforming DNS with a view toward maintaining open Internet self-government, introducing competition into Internet domain name administration, separating DNS management from trademark protection, and supporting the continued growth of the Internet itself. (CPSR addresses each section of the NTIA Notice, and as requested our comments in Sections II-VI follow the organization of the Notice.)

As discussed more fully below, CPSR -- a public interest alliance of information technology professionals and others concerned about the impact of computer technology on society, founded in 1981, with over 1,400 members and 22 chapters nationwide -- believes that DNS is too important to the structure of the Internet for DNS "reform" to proceed in a hasty or ill-conceived manner, particularly without adequate input from consumers and other users of the Internet. Whatever its merits, the IAHC process was closed, rushed and unbalanced, leading to a proposal that should not be endorsed by the US government. CPSR commends NTIA for commencing this open, thorough public airing of DNS issues, and for its express acknowledgment that DNS reform, like other aspects of Internet governance, should remain a matter for the Internet community itself, not national or international government agencies.

CPSR proposes that changes to the current DNS model must reflect the twin goals of maintaining Internet self-governance, thus minimizing government's substantive role in Internet administration, while avoiding the continuation of de facto DNS monopolies in the increasingly commercialized Internet. We also emphasize, however, that there is no present "crisis" in DNS administration that requires hasty implementation of any system for DNS reform, including those proposed by IAHC, Network Solutions, Inc. (NSI) <http://www.netsol.com/ papers/internet.html> and others. Unlike IAHC or NSI, CPSR believes that the twin aims of competitive Internet services and non-governmental Internet administration can and should be applied to the DNS system. A sensible plan for DNS reform combines the better elements of both the IAHC and NSI proposals, while jettisoning others. In this light, CPSR proposes the following principles for reform of the Internet DNS system:

A. The Internet domain name registration process should be opened to competition for all existing and newly created generic top-level domains (gTLDs).

1. Shared gTLDs should be administered by competing registrars, with restrictions imposed only based on any technical limitations.

2. No registrar (NSI or others) should enjoy a proprietary interest or commercial "ownership" of any gTLD, including ".com".

B. Domain registration should be separated from trademark issues.

Registrars should not be involved in trademark dispute resolution, but rather should refer all trademark issues to appropriate national and international judicial bodies.

C. The Internet's "root" server administration responsibilities should be coordinated and centralized in order to assure reliability and scaleability of the Internet.

D. The DNS reform process and ongoing DNS administration should be handled in an open, balanced and non-governmental manner, with full participation by consumers and small commercial entities, in addition to trademark owners.

1. International quasi-governmental organizations (ITU, WIPO, OECD, etc.) should have no formal role in Internet governance or domain name registration. The extensive new bureaucracy for domain name management and oversight proposed by IAHC, including a Swiss-based Council of Registrars (CORE), a Policy Advisory Board (PAB) <http://www.pab.gtld-mou.org> and a higher level interim Policy Oversight Committee (iPOC), is unnecessary and counterproductive.

2. National governments (Commerce, DOD, etc.) have no necessary role in DNS administration except for ISO 3166 TLDs (e.g., ".us," ".de," etc.) and maintaining fair, open and competitively neutral Internet self-governance organizations.

3. The IAHC process was inconsistent with open Internet self-governance and biased towards trademark owners. With encouragement from NTIA, the Internet Society (ISOC) should be required to open up the process to permit full debate by the global Internet community on DNS practices. The absence of any "crisis" in domain name resources allows for thoughtful and deliberate consideration of DNS issues.

4. Hasty implementation of the IAHC approach will continue to splinter the Internet community and would unnecessarily involve international quasi-governmental organizations in Internet governance. The DNS reform process should be slowed in order to permit achievement of a consensus approach that all interest groups (including Internet users/consumers) can support. No "rush to reform" is necessary.

5. The US government should not endorse, and should actively oppose, intervention by ITU and WIPO in the DNS administration process. The government should not attempt to unilaterally dictate any specific domain name registration process for gTLDs, which are global Internet resources.

I. INTRODUCTION

A. Background

1. The domain name system (DNS) provides a key service in the global Internet. In translating Internet 'host' names into Internet Protocol (IP) numbers needed for transmission of information, DNS shields network users and administrators from the complexities of using IP numbers to identify computers, routers and networks on the Internet. Without a reliable, efficient and robust DNS system -- including registration of domain names and management of DNS servers -- the Internet as currently configured could not function in the user-friendly manner available for such current Internet applications as electronic mail and World Wide Web (WWW) browsing. DNS administration also plays an important role by maintaining 'root' servers that operate as the highest-level Internet databases for translating domain names into IP addresses, without which domain name-based Internet communications would not be possible.

2. Many governmental, quasi-governmental and private groups, along with users and Internet service providers, are currently examining the limitations of the current DNS and domain name registration models, including questions related to introducing competition to domain registration, oversight mechanisms for DNS administration, opening of new generic top-level domains (gTLDs), and the role of registrars in resolving trademark disputes regarding domains.

3. One of these groups, the so-called Internet Ad Hoc Committee (IAHC), has proposed and is implementing a plan for establishing new gTLDs, with significant input from and the endorsement of international multilateral organizations, such as the World International Property Organization (WIPO) <http://www.wipo.int/> and the International Telecommunication Union (ITU) <http://www.itu.ch>, as well as major trademark holders. After signing of its MOU in Geneva, the IAHC was dissolved on May 1, 1997 and replaced by the iPOC <http://www.gtld-mou.org/docs/ipoc-members.html>.

4. The IAHC proposal has drawn opposition and criticism from parties ranging from Network Solutions, Inc. (NSI) <http://www.netsol.com>, the current US government-selected contractor responsible for registrations in the '.com,' '.gov,' '.org,' and '.edu' domains, to the Domain Name Rights Coalition <http://www.domain-name.org/dnrc-comments.html>, the World Internetworking Alliance <http://www.wia.org/pub/dns-brief.html>, the Organization for Economic Cooperation and Development (OECD) <http://www.oecd.org/>, Directorate General XIII (telecommunications) of the European Union <http://europa.eu.int/en/comm/dg13/13home.htm>, and major US-based Internet service providers. Alternate DNS models have also been developed by other providers of domain name services, such as eDNS <http://www.edns.net/> and AlterNIC <http://www. alternic.net>, and in draft form by the Internet Engineering Task Force <http://www.iahc.org/contrib/draft-iahc-higgs-tld-cat-03.txt>. At least one entity has filed suit against the IAHC plan in the United States <http://www.wia.org/dns-law/pub/stef-io-declaration.html>.

B. CPSR's Concerns

1. CPSR is a public interest alliance of information technology professionals and others concerned about the impact of computer technology on society, founded in 1981, with over 1,400 members and 22 chapters nationwide. CPSR believes that the Internet is a potentially revolutionary medium that has flourished, and grown exponentially over the past several years, in large part due to marketplace competition and the absence of restrictive government involvement. We believe these same objectives should be applied to efforts to 'reform' DNS administration. Given its central role in the development of the Internet, the US government has judged well when to lead the process of Internet management and when to cede authority to the Internet community and private organizations. In the case of DNS administration, it is time for more decentralization, competition and private initiative, although governments can and should serve as 'stewards' of the Internet to ensure open and inclusive processes, prevent anticompetitive practices, and complete the final step in the transition from a government-managed to a private sector, community-administered Internet.

2. While CPSR agrees that monopoly domain registrars are a poor substitute for a competitive DNS model, we caution that DNS is too important to the structure of the Internet for DNS 'reform' to proceed in a hasty or ill-conceived manner, particularly without adequate input from consumers and other users of the Internet. Whatever its merits, the IAHC process was inconsistent with consensus-based Internet policymaking and is biased towards the rights of trademark owners. The IAHC plan should therefore not be endorsed by the US government. CPSR applauds NTIA for commencing this open, thorough public airing of DNS issues, and for its express acknowledgment that DNS reform, like other aspects of Internet governance, should remain a matter for the Internet community itself, not national or international government agencies.

3. The current proposals for DNS reform have been advanced by groups with commercial and institutional objectives that do not necessarily reflect the broader interest in a competitive, non-governmental Internet structure. ISOC is a respected organization, with an important role to play in the Internet community, that to its credit, has correctly recognized that problems exist with domain name administration. However, the outcry provoked by IAHC's gTLD-MOU shows that its process was not open enough. By unilaterally implementing its proposal through quasi-governmental international organizations (ITU and WIPO), IAHC has sought to legitimize a role in Internet governance it simply does not enjoy, at a time when its controversial proposal was still undergoing critical examination by the Internet community at large. Substantively, although it proposes competitive provision of domain registration in a series of new gTLDs, the IAHC plan is geared largely to the creation of international rules (and mandatory, non-judicial arbitration) for trademark protection of domain names, including a new bureaucracy within WIPO for the resolution of domain disputes. The NSI response, in contrast, proposes to protect its private commercial interests by extending its de facto monopoly on '.com' registrations into a competitive DNS environment, without sharing of TLDs, and has advocated that national and/or international governments should be involved in 'oversight' of certain 'administrative' DNS functions <http://www.netsol.com/papers/ internet.html>.

4. Unlike IAHC or NSI, CPSR believes that the twin goals of competitive Internet services and non-governmental Internet administration can and should be applied to the DNS system. To do so, however, requires recognition that (1) as NSI has explained, there are some functions, known as 'root' server management, which need to remain centralized and coordinated in order to promote reliable and scaleable Internet routing, and (2) 'rationalization' of trademark law (the pursuit of what IAHC terms 'efficient' dispute mechanisms as an alternative to court litigation for trademark protection) for DNS purposes is inconsistent with the traditional, and successful, non-governmental approach to Internet administration. (While the US government for years subsidized certain Internet facilities and still contracts for DNS and IP-number allocation functions, Internet policies, standards and operating arrangements have been set with little if no government involvement since the 'commercialization' of the NSFNet backbone in 1994.)

C. CPSR's Proposal

A sensible plan for DNS reform combines the better elements of both the IAHC and NSI proposals, while jettisoning others. In this light, CPSR proposes the following principles for reform of the Internet DNS system:

1. The domain name registration process should be opened to competition for all existing and newly created gTLDs.

a. Shared TLDs should be administered by competing registrars, with restrictions imposed only based on any technical limitations.

b. No registrar (NSI or others) should enjoy a proprietary interest or commercial "ownership" of any gTLD, including ".com".

2. Domain registration should be separated from trademark issues.

Registrars should not be involved in trademark dispute resolution, but rather should refer all trademark issues to appropriate national and international judicial bodies.

3. The Internet's "root" server administration responsibilities should be coordinated and centralized in order to assure reliability and scaleability of the Internet.

4. The DNS reform process and ongoing DNS administration should be handled in an open, balanced and non-governmental manner, with full participation by consumers and small commercial entities, in addition to trademark owners.

a. International quasi-governmental organizations (ITU, WIPO, OECD, etc.) should have no formal role in Internet governance or domain name registration. The extensive new bureaucracy for domain name management and oversight proposed by IAHC, including a Swiss-based Council of Registrars (CORE), a Policy Advisory Board (PAB) <http://www.pab.gtld-mou.org> and a higher level Interim Policy Oversight Committee (iPOC), is unnecessary and counterproductive.

b. National governments (Commerce, DOD, etc.) have no necessary role in DNS administration except for ISO 3166 TLDs (e.g., ".us," ".de," etc.) and maintaining fair, open and competitively neutral Internet self-governance organizations.

c. The IAHC process was inconsistent with open Internet self-governance and biased towards trademark owners. With encouragement from NTIA, ISOC should be required to open up the process to permit full debate by the global Internet community on DNS practices. The absence of any "crisis" in domain name resources allows for thoughtful and deliberate consideration of DNS issues.

d. Hasty implementation of the IAHC approach will continue to splinter the Internet community and would unnecessarily involve international quasi-governmental organizations in Internet governance. The DNS reform process should be slowed in order to permit achievement of consensus approach that all interest groups (including Internet users/consumers) can support. No "rush to reform" is necessary.

e. The US government should not endorse, and should actively oppose, intervention by ITU and WIPO in the DNS administration process. The government should not attempt to unilaterally dictate any specific domain name registration process for gTLDs, which are global Internet resources.

II. PRINCIPLES

NTIA has sought comment on a set of principles by which the Administration can "evaluate proposals for the registration and administration of Internet domain names." The proposed NTIA principles are:

"(1) Competition in and expansion of the domain name registration system should be encouraged. Conflicting domains, systems, and registries should not be permitted to jeopardize the interoperability of the Internet, however. The addressing scheme should not prevent any user from connecting to any other site.

(2) The private sector, with input from governments, should develop stable, consensus-based self-governing mechanisms for domain name registration and management that adequately defines responsibilities and maintains accountability.

(3) These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.

(4) The overall framework for accommodating competition should be open, robust, efficient, and fair.

(5) The overall policy framework as well as name allocation and management mechanisms should promote prompt, fair, and efficient resolution of conflicts, including conflicts over proprietary rights.

(6) A framework should be adopted as quickly as prudent consideration of these issues permits."

CPSR Comments

A. The basic thrust of the NTIA principles are appropriate in light of the needs for competition in gTLD domain registration and for Internet self-governance.

B. The international role of gTLDs is a crucial aspect of the principles. Neither the US nor any other national government has authority to dictate unilaterally policies or rules for gTLD creation or administration. While the vast majority of '.com' domain registrations certainly remain in the United States, the US government can best promote the global expansion of Internet access -- to the resulting commercial benefit of American-based Internet software, hardware and content companies -- by ensuring that it is not viewed internationally as attempting to secure a favored position for US-centric domain holders in the gTLD scheme.

C. The IAHC and NSI proposals raise administrative, legal, financial and business issues far beyond the technical focus and scope of the ISOC. ISOC and the other members of IAHC (such as ITU and IANA), have no clear charter to dictate gTLD policy and little if any experience in legal procedure or public policy. A need exists, therefore, for the establishment of new 'consensus-based self-governing mechanisms' that allow balanced, fair and open participation on these Internet policy issues by all segments of the Internet community. This includes Internet users, who have essentially been excluded from the IAHC process. Given the crucial nature of end-user participation, the Internet community deserves legitimate public interest, non-profit input into decision making processes. The needs of the Internet community should not be subordinated to profits or control issues. The US government can thus act as a catalyst in assisting the creation of the new self-governance organizations (open and balanced consortia of Internet professionals, providers and users) that will be necessary to complete the transition to a fully non-governmentally administered Internet.

D. Although the US government has fostered the growth of the Internet and laid the framework for its current expansion and commercialization, its future role should not include oversight of DNS administration or the selection of registrars and other DNS-related entities (InterNIC, IANA, etc.). Rather, the US and other national governments should encourage open, consensus-based Internet self-governance, intervening (as in this proceeding) only to assure public debate and to prevent any single segment of the Internet community from asserting its special interests above those of all Internet users. As in its oversight of the North American Numbering Plan Administrator (NANPA), the US government can also act to guarantee that vital Internet communications resources are managed impartially by the affected industries and to prevent disruptive efforts by other governments and inter-governmental organizations to assert authority over Internet administration.

E. Conflict resolution for domain name registration need not and should not include resolution of disputes over trademarks or other proprietary rights. DNS policies should be based on technical efficiency, reliability and competitive considerations, without regard to the relative legal rights of trademark owners and domain holders. As discussed in Section VI, 'efficient' trademark dispute resolution is a code word for supplanting judicial decision-making with DNS registrar-imposed trademark rules. Particularly in the IAHC proposal -- which provides the basis for WIPO to establish new international common law of trademarks applicable solely to DNS and without resort to national courts -- combining DNS administration and trademark dispute resolution is poor public policy.

III. GENERAL/ORGANIZATIONAL FRAMEWORK ISSUES

In this Section, NTIA asks for comment on a number of issues related to the operational model for domain name registration and the technical basis for alternatives to the current model, including retirement of gTLDs from circulation, the relationship between root servers and gTLDs, and interoperability among registrars.

CPSR believes that the current model for DNS administration -- monopoly registrars for both gTLDs and ISO (country code) domains -- is anticompetitive and will, in the long run, present substantial risks to the scaleability of the Internet. We believe that the disadvantages of the monopoly model clearly warrant a shared registration approach to gTLDs, but that certain functions, namely root server administration, should remain centralized and coordinated in order to preserve the basic reliability and 'transparent' functionality of the Internet itself.

A. Existing gTLDs should not be retired. In addition to the transaction costs and confusion surrounding such an approach, the existence of 'supra-national' domains is inherent in the global nature of the Internet, helping to maintain the Internet as a geographical and politically indifferent medium. Instead of eliminating the .com and other gTLD registries, competition should be introduced into domain name registration.

B. Shared gTLDs are a workable and efficient approach to introducing competition to the DNS system. Much as in the US system for administration of '800' telephone numbers, common standards and technical interfaces can be created -- based on non-governmental industry standards processes -- to allow shared use of the .com resource. Competing registrars would assign available domains in all gTLDs on a first-come, first-served basis, with real-time querying of the appropriate gTLD database. Internet users would in this way be accorded, for the first time, real choice of domain name registrars, with the consequent competitive pressures leading to reduced prices, improved service and innovative registration options.

C. The IAHC proposal for the establishment of new gTLDs is a good approach, in that it permits use of identical domain names in different gTLDs for different purposes (e.g., mcdonalds.com and mcdonalds.nom). The specific new gTLDs selected by IAHC, however, raise questions about global transparency and potential duplication and user confusion that merit additional study. IAHC's proposal, now rescinded, to limit the number of registries to 28 (four per WTO region) is unnecessary, as there should be no limits on the creation and operation of domain registries except as may be required for technical reasons. Rather than mandating financial and other registrar qualifications, 'quality control' of domain registration should be a function of the marketplace. So long as a registrar follows any technical standards established for assignment of a domain name from a shared TLD, there appears to be no reason why prior "authorization" from any governmental or Internet-governance entity should be required for entry into the domain name registration market.

D. Root server administration (and related "population" of zone server databases) is a key DNS function that should remain centralized, under the control of a single entity worldwide, and coordinated through a technically reliable and integrated process. Root servers must maintain interoperable, updated databases in order to ensure that all IP networks can communicate across the Internet. As NSI has correctly noted, the IAHC proposal is deficient because it does not reflect this need for root server (what NSI terms "the .dot") centralization and "risks the fragile stability of the Internet" by "increasing the likelihood of failure of the administrative functions and services of the Internet." Competition in root server administration would therefore -- absent some new technical approach to Internet architecture -- appear to represent an exception to the general principle of competition for DNS. (Of course, competition could still exist as part of the root server administrator selection process.)

E. The US government has traditionally been the "first-mover" on issues related to Internet architecture and administration, but in this instance should refrain from selecting the entity responsible for root server administration. Backbone Internet carriers and other entities, commercial and governmental, responsible for IP transport have the technical background to select a responsible root server administrator and (as in ARIN <http://www.arin.net/>, RIPE <http://www. ripe.net/>, etc.) establish compensation mechanisms under a non-governmental structure.

1. End users and domain holders have no substantial interest in root server administration since root servers are "transparent" to them. NSI's proposal that the US government assume "interim authority" over root server administration -- and then transfer such authority to an international legal entity -- is unnecessary. Absent some fundamental breakdown in the Internet's technical standards development process, the shared mutual interest of backbone and major Internet transport providers in maintaining functional, high-quality root server management (and ensuring their ability to route domain-based messages to all other IP networks) is sufficient to guarantee a responsible root server administration.

2. Root server centralization does not guarantee error-free administration or coordination (synchronization), as the events of the past month (NSI's snafu) have illustrated. Nonetheless, with the increasing reliance of individuals and businesses worldwide on the Internet, any benefits from competitive provisioning of root server management would appear to be outweighed by the risks of network collapse (and overhead requirements) arising from multiple root server administrators.

F. Oversight of shared gTLD administration should remain a non-governmental function. It should not be placed under the auspices of quasi-governmental organizations like WIPO or the ITU, as IAHC has proposed. So long as registrars comply with consensus-based technical requirements for domain name registration, there is no need for the creation of CORE, iPOC and the related new bureaucratic bodies IAHC has proposed. Thus, as it has already indicated, the US government should not endorse the IAHC Memorandum of Understanding (MOU) and should oppose the use of ITU and WIPO as oversight bodies for the DNS system.

G. Administration of gTLDs and ISO country code domains is completely separable. Although the same principles of competition and self-governance should also apply to ISO-domain registrations, these resources have traditionally been considered to be subject to the rules and policies of national governments. There is no technical reason why competitive, shared gTLD administration cannot co-exist with the many monopoly ISO-domain registrars. Indeed, the creation of a competitive system for efficient, worldwide administration of gTLDs would have the commensurate benefit of creating marketplace pressures that would encourage national governments to open their own ISO-domain registries to competitive entry. As the OECD has recommended to EU-member nations for their own country code domains, the US government should consider opening the ".us" domain to competing registries and should actively promote similar procompetitive policies by other national governments for ISO country code domains.

H. The "transition to any new systems" for domain name registration should be accomplished with the full, open and considered participation of all stakeholders in the Internet. The IAHC proposal, in contrast, raises serious questions about the authority of ISOC and the other members of IAHC to impose their chosen solution for DNS reform and the essential openness of Internet self-governance.

(1) IAHC does not have sufficient input from the Internet community; the Internet Society is not a member oriented group, and its decisions cannot legitimately claim to be based on Internet community "consensus."

(2) IAHC's aggressive timelines are unjustified, especially given the lack of urgent technical problems. As recently as August 1996, it was deemed unlikely that the growth of the .com domain would pose any technical problems.

(3) IAHC's comment period was very short (2 weeks), and its extremely rapid implementation timeline does not allow for sufficient debate and generation of proposals for registrars.

(4) According to a brief filed by the World Internetworking Alliance <http://www.wia.org/pub/dns-brief.html>, the participation of international groups in the IAHC raises concerns regarding the legal nature of the IAHC proposal as an intergovernmental agreement.

(5) Despite the participation of ISOC, IAHC did not include substantial input from the community of Internet end-users. Furthermore, the IAHC's proposed bureaucracy is similarly lacking in community representation.

(6) IAHC is proceeding as if its proposal was thoroughly and completely accepted, including a Geneva MOU signing ceremony in early May 1997 and the continuation of plans to implement CORE-managed shared gTLDs in October 1997. However, this purported Internet consensus is not all clear. For instance, the Commercial Internet Exchange (CiX) <http://www.cix.org>, a consortium including members such as AT&T, CompuServe and MCI, has expressed reservations regarding the IAHC plan, and a new organization known as the Open Internet Congress (OIC) <http://www.interactivehq./org/oic/> has similarly raised objections to the closed, hasty and "top-down" nature of the IAHC process.

IV. CREATION OF NEW gTLDs

In this Section, NTIA specifically seeks comment on whether new gTLDs should be established and whether there are any technical, practical and/or policy considerations that constrain the number of different gTLDs or the management of gTLDs.

CPSR believes, as discussed in Section III above, that IAHC has developed a reasonable approach to the creation of new gTLDs, with different gTLDs signifying different uses of domains. Although this approach was driven more by the demands of existing registrars and trademark owners for TLD-based alternatives to domain-name disputes in the .com registry, in order to reduce pressures for litigation over .com domains, the concept of creating "special-purpose" gTLDs is an attractive one from the perspective of both Internet users and commercial businesses. (Whether or not the specific gTLDs developed by IAHC are appropriate or optimal is open to question, however.)

A. Creation of new gTLDs is not a necessary predicate to the introduction of competition into the DNS system. Existing gTLDs registries could be administered on a shared basis. There is no technical reason to create new gTLDs, but there is also no technical constraint (except the capacity of root servers and IP space) to the number of gTLDs that the Internet can support. Unlike the situation in the North American Numbering Plan for telephone numbers, IP resources, although finite, do not appear to be as anywhere near in such short supply as telephone numbers and area codes. The creation of new gTLDs should not, at least in the short run, create any undue stress on available IP resources.

B. Competitive provisioning of DNS is not sustainable if only new gTLDs are opened for shared administration. As discussed in Section V, the .com and other NSI-administered TLD registries are the product of the historic, US-government monopoly over gTLDs and NSI's status as the "sole-source" government contractor for domain registrations in gTLDs. Extending this monopoly into an era of competitive DNS would allow one private, commercial entity to obtain the most currently valuable gTLDs. Competition in DNS is accordingly inconsistent with exclusive control of any gTLD, whether .com or a newly created gTLD.

V. POLICIES FOR REGISTRIES

In this Section, NTIA asks whether registrars should have exclusive control of a particular gTLD, whether shared and "exclusive" gTLDs can co-exist, whether there should be any threshold requirements for domain registrars, and whether there are any technical, business and/or policy constraints in light of IP number space or other issues on the number of domain registrars and how many domains any registrar can administer. NTIA also asks whether a registrar should be permitted to administer both exclusive and shared gTLDs.

CPSR believes that monopoly control of any gTLD is antithetical to the interests of Internet users and to the long-run objectives for competitive, non-governmental administration of the Internet. NSI's proposal for the retention of proprietary, so-called "brand" TLDs is a naked attempt to convert the public .com resource to its private, commercial property. There is no legal or policy justification for such an approach either in NSI's existing NSF contractual rights or in the legitimate needs of Internet DNS administration. Whether or not new gTLDs can or should be created, all TLDs should be opened to competitive administration, and no TLD registrar should be entitled to maintain any proprietary interest in an "exclusive" TLD.

A. As a technical matter, exclusive and shared TLDs can co-exist, provided that all TLD administrators comply with industry standards for interoperability with the Internet root servers.

B. TLDs should not be subject to the exclusive control of any single registrar. NSI's approach, under which new registrars would establish exclusive domains marketed commercially on a "branded" basis, ignores the proper status of domain names as globally unique identifiers for all users, both commercial and non-commercial (much like NANP telephone numbering resources). While both domain names and telephone numbers can and should be privately held, the higher-level resources (gTLDs and areas codes, respectively) are not the private property of any provider or entity.

C. Given commercialization of (and competition in) the Internet backbone, transport and access functions, DNS is the last bastion of government-funded, sole-source administration of the Internet. Neither NSI nor any other private entity can lay claim to convert the .com registry, which until recently was administered on a "cost-plus" contractual basis, into a private monopoly. NSI's position that "[l]egitimate corporations will not invest time, stockholder capital and other resources in Ôshared' brands," ignores the monopoly roots of the .com registry and is inconsistent with market realities. Already, many organizations and commercial ventures exist, in America and worldwide, that compete for registration services for .com domains to end users. Given the ubiquity and market recognition of the .com domain -- a product of its monopoly status rather than any special technical or business expertise of NSI -- NSI should not be permitted to transfer its NSF-contract role into a huge (as much as $35 million per year) financial windfall and commercial marketplace "head start" over competing TLD registrars.

D. NSI is correct, however, in critiquing IAHC's initial proposal to artificially limit the number of gTLD registries to 28 and its continued efforts to impose "qualification" requirements (financial and otherwise) on applicants for registrar status. These restrictions may raise potentially serious antitrust concerns. In fact, there is no underlying need for any "awarding" of registration rights, since the only technical requirement for gTLD administration in a shared TLD environment is that each registrar follow industry specifications for reservation and assignment of domains and interoperate with the Internet root servers. Moreover, if root server administration is centralized, the consequences arising from the failure of any individual registrar to follow these standards (so long as any one registrar's data could not contaminate the entire root server database) would be limited and localized -- its domain name holders would be unable receive Internet transactions (Web HTTP "hits" and electronic mail messages) from the larger Internet, but the interoperability and functionality of the Internet itself would remain uncompromised. Therefore, marketplace forces should be adequate to ensure that gTLD registrars operate in a technically efficient manner (although national consumer protection laws may need to be invoked against "rogue" or "sham" registries that engage in deceptive or fraudulent registration practices).

VI. TRADEMARK ISSUES

In this section, NTIA seeks comments on intellectual property issues (trademarks, etc.) on the Internet, including whether trademark rights, if any, should be protected on the Internet, whether domain registration should require any preliminary review of an application for potential trademark conflicts, and how trademark disputes regarding domain names should be resolved.

CPSR strongly believes that the infusion of trademark issues, and trademark dispute resolution processes, into domain registration is a serious problem that should be corrected. Much of the impetus for the IAHC plan stems from differences between trademark owners and others as to the relative rights, and economic leverage, that NSI's own (and unilaterally imposed) dispute resolution policies have engendered over the past several years. Thus, given the role of WIPO and significant commercial trademark owners in IAHC, IAHC's proposal quite naturally seeks to provide a single, internationally based procedure for arbitration of trademark disputes, coupled with mandatory waiting periods, that would substantially increase the ability of large corporate entities to chill smaller businesses and users from registering potentially infringing domain names. At the same time, much of the litigation against NSI has arisen not because of any inherent conflict between the domain registration process and trademark rights, but rather because NSI chose to insert itself as the mediator of disputes between trademark owners and domain holders by cutting off or threatening to terminate domain registrations on receipt of a complaint from a trademark holder. Finally, relative to the millions of domain names registered, there has been a trivial number of disputes, which should not lead to the imposition of a cumbersome, international and mandatory dispute resolution approach that will undoubtedly slow down the domain registration process for all Internet users.

A. DNS administration should be separated from trademark rights and intellectual property dispute resolution. gTLD registrars should not intervene in disputes between trademark owners and domain holders, but rather should refer disputes to the applicable national courts. The US government can assist this sensible result by making clear its view that because registration of a domain name is not the "use" of a trademark, the domain registration process cannot constitute "contributory infringement" under US trademark law such that registries would be exposed to legal liability for ignoring a request to "take down" a potentially infringing registration. (CPSR notes that there is a significant line of domestic US trademark cases involving proprietary claims to telephone numbers, but that in none of these cases was the telephone company or other number assigning entity either sued or found liable for contributory infringement.)

B. There is no need for any international body of trademark law or supra-national forum for decision or arbitration of Internet-related trademark issue. Intellectual property rights have always varied on a nation-by-nation basis, and multinational corporations are experienced in dealing with the multiplicity of different trademark laws world-wide. Conversely, jurisdictional limits on the authority of national courts protect smaller registries and domain holders from being subject to suit in other nations over US-based domain activities. Simply put, in the absence of a multilateral treaty, trademark issues on the Internet, like other intellectual property rights, should be subject to the domestic laws of each affected nation.

C. Domain trademark conflicts can be prevented by use of trademark databases prior to registration of new domains. Although the information sources and operation of such databases are beyond CPSR's expertise, we note that similar databases already exist in the US for both patents and trademarks and are commonly used by firms specializing in dealings with the PTO. While domain applications should continue to be processed on a first-come, first-served basis -- and no "entitlement" to a domain name should be required to be demonstrated by an applicant -- the cost and burden of changing domains in the face of a potentially infringing registration should be sufficient to persuade a reputable applicant to select a different domain in the first instance. So-called "pirate" applicants, whose registrations are an effort to extract financial concessions from trademark owners, can be dealt with adequately in the courts or, if the trademark owner finds it economically advantageous, through private settlements.

D. CPSR believes that the recent fascination of many parties with trademark rights to Internet domains is a short-run issue only. The desire of many commercial entities to obtain easily recognizable ".com" domains is a function of the inexperience of many Internet users -- novices typically type URLs into their browser -- and the relative unsophistication of WWW search engines. As user experience with the Internet increases, domains become progressively less important to navigation. Indeed, technology is already developing that may supplant DNS with other Internet "directory" services, whether intelligent agent-based search engines, "push" and "channel" technologies by which information is sent to users (instead of retrieved by them), IP-address based directories (e.g., LDAP), and the like. The important and competitively crucial matter of introducing competition to DNS administration should therefore not be held hostage to a quixotic desire to create a new, international law of Internet trademark rights or to perfect an "efficient" trademark dispute mechanism that displaces national courts, since the importance of literal or pneumonic domain names will likely decrease substantially as the Internet evolves and matures.

CONCLUSION

NTIA and the entire Executive Branch have appropriately intiated a forum for public debate on the controversial issues of domain name registration and Internet administration, while making clear that such matters are to be resolved principally under the auspices of open, consensus-based Internet self-governance organizations. Having developed and for years funded the Internet, the US government should assist the final steps in the transition to a fully non-governmental model of Internet administration. The DNS process should be opened to competition through shared gTLDs and rescission of NSI's de facto monopoly of the ".com" registry, but with continued centralization and coordination of root server administration. Trademark rights and trademark dispute resolution should be separated from the DNS process, with domain name intellectual property issues referred to national courts or subject to multilateral international treaties. The IAHC gTLD-MOU, given both its procedural infirmities and its improper resort to multinational quasi-governmental Internet oversight, should not be endorsed, and should actively be opposed, by the US government.

Respectfully submitted,

COMPUTER PROFESSIONALS FOR

SOCIAL RESPONSIBILITY

By: Glenn B. Manishin /s/__________
Aki Namioka, President Glenn B. Manishin
Harry Hochheiser Michael D. Specht
Andy Oram Christine A. Mailloux
Computer Professionals for Social Blumenfeld & Cohen - Technology Law Group
Responsibility <http://www.technologylaw.com>
<http://www.cpsr.org/home.html> 1615 M Street, N.W., Suite 700
P.O. Box 717 Washington, D.C. 20036
Palo Alto, CA 94302 202.955.6300
415.322.3778

Counsel for CPSR

Dated: August 18, 1997

###

Number: 338

From: Dan Steinberg <dstein@travel-net.com>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 12:40pm
Subject: Docket No. 970613137-7137-01

Re: Request for Comments on the Registration and Administration of
Internet Domain Names

My comments are attached as a Word 6/95 file.

Thank you for the opportunity to present my views.

Dan Steinberg
MBA LLB BSc

SYNTHESIS:Law & Technology
Box 532, RR1 phone: (613) 794-5356
Chelsea, Quebec fax: (819) 827-4398
J0X 1N0 e-mail:dstein@travel-net.com

RESPONSE TO NTIA RFC

[Docket No. 970613137-7137-01]

PREAMBLE

I welcome the opportunity to comment on "Request for Comments on the Registration and Administration of Internet Domain Names" [Docket No. 970613137-7137-01], as published in the July 2, 1997 Federal Register. I feel that the RFC process is extremely important, as any changes to the existing structures have the potential for greatly influencing the direction of the internet.

This response was deliberately submitted after many other submissions, in an attempt to avoid repeating the obvious and in an attempt to synthesize/analyze some of the common threads. One of the most common threads is a suggestion that a hierarchical structure be imposed on the internet (similar to the existing structure under the .US and .CA domains). The presumed advantages cited for this structure include:

decrease or elimination of domain/trademark disputes

increase in the quantity of possible URLs or mail addresses

But there are structural problems inherent with such a framework. With a hierarchical structure similar to the current .US framework, an organization that physically moved would have to change internet addresses. For a physical move to require a corresponding cyberspace transition is ludicrous. It defeats one of the most elegant features of the net.

I have provided responses to all questions in each section. The views expressed herein are not necessarily the views of any of my clients or business partners. Thank you for the opportunity to share my thoughts on these important issues.

Dan Steinberg

SYNTHESIS: Law & Technology

35, du Ravin

Chelsea, Quebec

Canada J0X 1N0

DETAILED RESPONSE

A. Appropriate Principles

The Government seeks comment on the principles by which it should evaluate proposals for the registration and administration of Internet domain names. Are the following principles appropriate? Are they complete? If not, how should they be revised? How might such principles best be fostered?

The current 'competition' between competing registry schemes is an indication that competition/expansion should be encouraged. Despite dire predictions, the net has not been destabilized by this competition. So far everyone has cooperated to ensure that no domain confusion occurs. At last count, there were over 20 different Root server cooperatives and over a hundred new registries actively doing business.

In order for the private sector to develop self-governing mechanisms, it is necessary to arrange for a seamless transfer of authority. Implicit in the transfer of authority is the recognition that:

someone else currently has authority

that 'someone else' i.e. the US govt. has decided 'who' in the private sector should be in charge of providing a framework for self-governing mechanisms. This is a non-trivial step.

c. These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.

Agreed.

d. The overall framework for accommodating competition should be open, robust, efficient, and fair.

Agreed. But this should not take precedence over internet stability and the ability meet consumer demand for more domain names.

e. The overall policy framework as well as name allocation and management mechanisms should promote prompt, fair, and efficient resolution of conflicts, including conflicts over proprietary rights.

This is an admirable goal, but probably impractical without a resolution of the current problems in international trademark law and the establishment of a new body of intellectual property law covering domain names.

f. A framework should be adopted as quickly as prudent consideration of these issues permits.

There is really no hurry to establish a new framework, other than the need to decide what to do about the registry when the current contract with NSI expires next March.

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name registration systems?

It is important to distinguish between the merely administrative problems encountered by internet users when they interact with the InterNIC and actual structural issues that require an examination of public policy and/or legislation.

In general, the current system 'works'. It has been argued that there are many court cases and inequities, but this is true of any system. No matter what policies, procedures, etc. are instituted, there will always be conflicts and someone will always try to 'get away with' something.

Several have advocated that separate policy and administrative entities would better serve the internet. This is a good idea on the surface. However, the question remains: who has the right to decide on an organization that will have a mandate to make policy? And who decides/approves applications to make entries in the registry.

If the same organization (i.e. the US govt.) grants the policy rights to someone and the admin. rights to grant new registrars then nothing has changed. The various proposals to split the policy/registrars do not address the issue of 'who' owns this beast. If this function is devolved to the member nations (under an ISO TLD structure) then there is an immediate potential for conflict. So the choice is between no change and chaos.

The only disadvantage to the current domain system is in the current mechanisms/policies/agents for running the 'show'. There are known weakness that are discussed in response to later questions.

2. How might current domain name systems be improved?

It is my belief that there is not much inherently wrong with the current domain system. The only major problem is that there is a tendency within the US to consider .COM a US domain, and a tendency outside the US to consider it international. While it is possible to envisage a reorganization of the net to move away from this structure, it would most certainly not be in keeping with the current administrations' policy of not interfering with a growing successful experiment. It is impossible to make such structural change without conflict and disturbing many businesses, and that would almost certainly serve to remove some of the 'cachet' from internet commerce.

3. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

My personal recommendation is that US government accept ongoing responsibility for the internet as a public trust akin to the current role in ensuring world peace. This puts stewardship at an international diplomatic level. Objections to stewardship (and proposals to take it over) get made at the diplomatic level. This takes care of the top level issues. Once the US government accepts ongoing stewardship, it is free to decide on how to run it, just as it is free to decide how to and when to intervene in the interest of world peace. Choices include:

subcontracting the work out to a monopoly via an RFP process

subcontracting to a number of registrars with all who meet the admission criteria accepted

maintenance of the database as a public trust by the Library of Congress or some other existing government body

subcontracting the work out to a public organization such as the IETF, ISOC or similar group

It is evident that the current gTLD MoU has created a great deal of interest in the public policy issues regarding internet governance. I am personally not in favor of the gTLD-MoU, but this personal position in no way affects the underlying questions of tranfer of authority. Before the gTLD-MoU can gain validity (if ever), it is necessary for the US government to officially devolve such authority. In keeping with my four possibilities, such a devolution could take place via a subcontract to ISOC, which has sponsored the IAHC process. Again, this is not a personal recommendation, merely an observation on how the gTLD-MoU could gain validity. It is still a matter of public policy to be decided by the US government in consultation with concerned citizens.

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

My only comment on this is to suggest that existing models are not appropriate. In particular, the spectrum allocation process is most definitely not applicable. There is no current shortage of domain names. If there were, then the InterNIC would not be busy 24 hours/day. When a shortage arrives, adding 10 more TLDs gives an order of magnitude more domains. Expanding the domain name alphabet to include more than plain ascii adds several orders of magnitude and allows sites local to a foreign culture more freedom. This option is discussed more fully in a later comment.

In a recent US Supreme Court Decision (Reno vs. ACLU), an attempt was made by all parties in the case to equate the internet with "something" i.e. a park, a BBS, a city, etc. This is part of human nature to relate something new to a known entity. As the justices observed, the internet is not just like 'something'. It is a unique entity. Comparison with other existing structures is doomed to failure. Similarly, attempts to use existing models like spectrum to describe/resolve the issues will fall short.

5. Should generic top level domains (gTLDs), (e.g., COM), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the COM registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

Well, retiring .COM would solve most of the trademark issues as trademarks conflicts could then be solved by local courts. But this brings up national issues. Since the birth of domain name system, there have been numerous countries broken up by war, vote, etc. The most obvious examples are the breakup of the former U.S.S.R and Yugoslavia, but there are others. There is no guarantee that trend to split up nations will stop. Of course splitting up nations into separate states is really bad for ISO TLDs. If a country splits up, the separate components are sure to want new domains as part of their new national identity. And this is not something that can be done easily, like substituting a word in a word processing document. The internet does not necessarily know where in physical space a particular domain or sub-domain is located. So geographic boundaries become problematic.

Another possibility is that some nations/states may choose to merge. The leadership in Belarus appears ready to take their country back into the USSR (which doesn't exist anymore), but they will settle for Russia. What to do with those .BE domains? It is not as difficult as the country-split problem, but still an administrative nightmare.

All proposals to turn the internet into a hierarchical system have a certain elegance. But so does an org. chart that splits an organization on functional lines. And that's not always the best way to allocate workload. When the internet was just getting started, there was a hierarchical, structured addressing scheme available that had the blessing of governments and telecommunications organizations: X.500.

But the internet 'way' prevailed, because dstein@travel-net.com is easy to remember and much easier to paste into a message than something with separate fields that have to be filled out. People adopted the internet in part because it was so easy. There are downsides to this ease, such as difficulty in implementing global e-mail directories, but the benefits have far outweighed the disadvantages. That is why the internet has flourished, and X.500 is still struggling.

To summarize, these ideas seem like a good way to complicate and probably stifle the internet.

What is the alternative? I don't have a ready answer. In the absence of a viable alternative, the best option is the status quo. There is no sense in jumping in with solutions that could make things worse. The internet is thriving right now, and if a few people are making money by warehousing domains and a few people are in domain/trademark conflicts, it's a small price to pay. There are always winners and losers in any system and there are always a small minority that manage to 'cheat' the system. Such is life.

Domain warehousers/hijackers exist in part because they recognized the value of domains before their victims. Just because someone was asleep at the switch and failed to register their domains is not always a reason to protect them. Sure, there are inequities but this is true in any system.

The same is true for domain/trademark conflicts. In some cases, a trademark holder is being punished for not figuring out the value of the internet in time. In others, a big organization is flexing its muscles to step on a small player who happens to have a domain that the big boys

want. If these cases were more frequent, there would be cause to change the system immediately. But the vast majority of domain registrations are unchallenged.

In order to retire .COM it would be necessary to either:

move .COM to .COM.US or .COM.CA , .COM.UK, etc., or

make .COM so expensive that no one wants it

There have been proposals to significantly raise the bar on top-level names and mandate that the registration fee for XXX.COM or XXX.NET would be many thousands of dollars.

Such a proposals would be very difficult to impose. And it would be a goldmine for those hated domain speculators. For it would be difficult to get registered domains away from these organizations. To prove use, all they have to do is put up a web page, which is not very expensive. There are speculators with thousands of registrations. Imagine the difficulty in proving that they don't have a use for these domains, and imagine a system that would/could distinguish between warehousing and legitimate business needs such as a company like Nabisco protecting its intellectual property.

Besides, there is no practical reason to raise prices other than to encourage a hierarchical structure. There is nothing inherently wrong with easy names. Getting rid of easy names just to avoid trademark/domain conflicts seems like overkill. As the registered owner of several domains, I would stand to gain if the price went up, so I am not acting in self-interest when I object to raising prices. One of the main reasons that there is so much interest in the domain name issue is the underlying resentment for the $50/year charge that the InterNIC

instituted. Without discussing the need for a charge, it is easy to state that raising the charge significantly will NOT be popular. And raising the barrier artificially merely to make a more elegant structure would probably be grounds for a net revolt and construed as unwelcome interference.

6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

Current issues are legal and public policy problems. Technical solutions are not relevant to the problem. The relationship of registrars and TLDs with root servers is governed by convention and IETF RFC. These relationships have proven workable in the past.

7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and coordinate?

Scalability is only an issue if the TLD space is completely open. Unless the space is opened to permit .STEINBERG and .COMPAQ, there is no problem. To use an airline analogy, travel agents can book flights on more than 7 different airlines. The databases are shared, robust and take millions of hits/day without a problem using relatively old technology.

8. How should the transition to any new systems be accomplished?

I am not certain that any new system is required. Should the US government decide to grandfather some of the competing proposals for new gTLDs, it should be done without fragmenting the domain space. Other than that, transitions should be done in an orderly manner.

9. Are there any other issues that should be addressed in this area?

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

Creating additional TLDs is only useful if there is a mechanism to share. Otherwise, the owner of synthesis.com (not me) will want to register synthesis.firm and synthesis.web, etc. Perhaps a rule that an organization can only register a domain name under only one TLD would be in order. There would undoubtably be exceptions and conflicts raised by this rule. A possible workaround for the US only would be the limitation on using a 'famous' mark.

11. Should additional gTLDs be created?

There is nothing inherently wrong with creating more gTLDs. Variety is the spice of life. However, the creation of new gTLDs should not be considered a panacea. Some firms have been content to share the existing domain space and take on XYZ.COM, leaving XYZ.NET and XYZ.ORG to others. Other firms have aggressively registered their trademark name as domains in .COM, .NET, .ORG and most of major country domains. There is no reason to believe that these trends will not continue.

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

No system is without technical limitations. The existing infrastructure can certainly handle thousands of gTLDs. A different structure would be required for 'infinite' gTLDs. For business/policy issues, it is not recommended that the domain space be opened up immediately. It is first necessary to solve the international trademark issues now facing WIPO and then work out an integration of trademark issues with domain names. Until then, it will not be possible to decide who gets .PRINCE or .APPLE.

13. Are gTLD management issues separable from questions about ISO country code domains?

gTLD management is necessarily separate in that it cannot be managed by an organization assigned by local government. ISO country code domains are the 'domain' of the national governments or their authorized representatives. The US government cannot interfere with the sovereign affairs of a country but it can and should have a say in the management of the gTLDs that it has responsibility for, even if that is limited to saying "we don't want to be responsible anymore".

14. Are there any other issues that should be addressed in this area?

D. Policies for Registries

15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

Some of the current problems stem from the fact that a registrar has exclusive control over .COM, .NET and .ORG and this organization has operated as a quasi-judicial body without the appropriate enabling legislation. Most of the common complaints about the current registrar would be eliminated if there were choice. There is no technical reason why shared registries could not be operational immediately. It is a much more difficult task to 'register' someone for an airline flight in real time, but computer systems have been designed to do just that and access to the 'registry' or flight is shared among thousands of travel agents every day around the world.

17. Are there technical limitations on the possible number of domain name registrars?

No. As discussed above, the airline reservation system is proof that multiple registrars can conveniently share a database. It is just a matter of scaling the computer to the size of the traffic.

18. Are there technical, business and/or policy issues about the name space

raised by increasing the number of domain name registrars?

The main issue with raising the number of registrars is accountability. There must be a mechanism whereby the registrars are not responsible for deciding on the relative rights of entities contesting a particular domain. There must be a single point of contact and (more importantly), a single place for court orders to be enforced.

19. Should there be a limit on the number of different gTLDs a given registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

To continue with the airline analogy, there is usually no limit on the number of different airlines that a travel agent a book flights on. Users of a particular travel agent are normally most interested in flights that begin/end at a local airport. There are limits on the number of local flights, but these are determined by landing rights at the local airport. Thus, there should be no limit on the number of gTLDs a given registrar can administer. They should all have the right to take registrations in all gTLDs.

20.Are there any other issues that should be addressed in this area?

An improvement could be made in the area of voluntary transfer. Under the current process, the existing domain holder sends a message to the InterNIC in order to remove their registration. The transferee then sends a message with the new registration. In theory (and occasionally in practice), a third party could manage to register the domain between the transferor dropping the registration and the transferee sending the new registration. This is akin to someone else claiming a house while it is being transferred. Such a situation would be ludicrous in the physical world and it makes no more sense in cyberspace. There should be an established policy/procedure covering voluntary transfers that facilitates this process and removes the risk. Such a procedure could also be used for court-ordered transfers. This would have the additional benefit of plugging a loophole in the current legislation.

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-a-vis domain names?

It is difficult to impose protection when the extent of rights is not known. US court decisions have clearly indicated that a domain name identical to a trademark is not in itself sufficient to force the transfer of the domain. Any protection of rights should be something that can be imposed during the registration process, and this is where it is impossible. A registrar is not necessarily qualified to make even a first-cut decision on trademark infringement, so even if the registrar spends all day comparing applications to a database list, there is not guarantee that anything useful will come of it. What will the registrar do? Refuse the application and wait to be called into court? This protection would not be effective as the issue of who has rights to a contested mark must be settled in a court of law.

Giving trademark holders specific protection would encourage entities to trademark their domains, and this is unworkable. You cannot clear up the confusion by simply getting people to trademark their domain names.

Domain names are not necessarily trademarks, even though some trademarks have been infringed by domain names. You can have several firms using the same trademark, as long as there is no confusion. For example, the mark VO2Max is used by several organizations including:

K2 clothing for class 25 clothing and 18 leather goods

Mars (makers of the Mars bar) for a yet-to-be-released meal-replacement bar

As an unregistered mark by a local retailer who incidentally carries the K2 line..

Who actually has VO2Max.com? The Mars company. The local retailer in this example uses magi.com and VO2max is just a username, as in magi.com/~vo2max. K2 is either out of luck or happy to have K2.COM.

The DNS should not be merged with trademark law. There is a cost to this. There are over 1,000,000 .com domains now and a dozen lawsuits. This is not a bad ratio actually. The low cost (and we think it is too high already) of a domain has served to promote the idea of freely-available domain names. This has served to foster a culture of growth in the internet and encouraged many thriving startups. Imposing trademark space on domain space will have the following effect:

Add uncertainty to the process because it is not a simple task to do a trademark search and you never really know until it's too late if you missed something

Add cost because researching and establishing a trademark is not for the legal faint of heart. I did it for my own companies, but wouldn't recommend it as a hobby

Add time to the process because you can't just do a whois lookup. Time is money in business.

All of the three will serve to drive up the cost of a domain. Those who make money with domain registrations or do registrations as part of a full-service package will see a decrease in business, as this would constitute a new entry barrier. Those considering a new business venture would have to raise more capital and increase the timelines for their startup. It is this increase that could be disastrous. As anyone who has been through a startup will tell you, it is the time between the initial idea and the actual moneymaking transactions that is the most difficult to finance. Any increase in startup time will serve to raise the entry barrier to new business.

22. Should some process of preliminary review of an application for registration of a domain name be required, before allocation, to determine if it conflicts with a trademark, a trade name, a geographic indication, etc.? If so, what standards should be used? Who should conduct the preliminary review? If a conflict is found, what should be done, e.g., domain name applicant and/or trademark owner notified of the conflict? Automatic referral to dispute settlement?

No. In order to set up a process of review, it is necessary to establish rules and procedures and implement an administrative body that must employ procedural due process. This implies a court structure and procedures similar to trademark applications. From a public policy standpoint, such an organization would be contrary to President Clinton's stated goal of not interfering with the growth of the internet. Obtaining a trademark is already costly, time-consuming process. Obtaining a domain name is relatively inexpensive and almost instantaneous. This is one of the 'features' of the internet and should not be interfered with via the introduction of bureaucracy. Only a court is competent to decide if a domain is in conflict with a trademark, and the courts have explicitly stated that the process is not simple and limited to a comparison of names.

23. Aside from a preliminary review process, how should trademark rights be protected on the Internet vis-a-vis domain names? What entity(ies), if any, should resolve disputes? Are national courts the only appropriate forum for such disputes? Specifically, is there a role for national/international governmental/nongovernmental organizations?

Absent an international body to resolve international trademark disputes, there is no point in establishing any more protection for trademark rights. The current system protects trademark holders through established laws and court procedures.

24. How can conflicts over trademarks best be prevented? What information resources (e.g. databases of registered domain names, registered trademarks, trade names) could help reduce potential conflicts? If there should be a database(s), who should create the database(s)? How should such a database(s) be used?

The current system, which includes the courts, provides for a mechanism to resolve conflicts over tradmarks. It is doubtful that an additional database would be of any use. The current trademark system provides for organizations to share most trademarks across geographical and/or market distinctions. Some organizations like IBM operate across many geographical and classes of goods and services. No one could reasonably expect to trademark IBM shoes today. Nor would they expect to get away with registering IBM.COM. If they did, the courts would take it away from them.

For lesser known marks, it does no good for a sporting goods firm in Decatur, Georgia to know that someone in California has a trademark on peach for use in electronic computers. The firm will register peach.com, and if they call their store Peach sportsware and have established common-law rights to the mark, they will prevail in court. There is little chance that someone would confuse peach sportsware in Georgia with peach computers in California. All that is requires is a change from the current InterNIC policy of putting a domain on hold when a trademark holder makes a request. It is true that domain holders do have recourse through the courts by seeking a temporary restraining order, but all this does is favor cash-rich domain holders over poor ones. There is no need for the current :

"It should be mine. Take it away from them, InterNIC"

"OK. Hey, you. You are getting cut off"

"I object. Here's my TRO, InterNIC. Let the courts decide"

"OK, domain holder. I won't put your domain on hold or transfer it"

A simple solution is to do away with the policy. Domains only change on the agreement of both parties or on court order. The courts are well-versed in trademark law and can best decide between a domain hijack and legitimate use by a common-law mark holder. This solution also does away with the need to seek a Tunisian trademark just to defend rights to a domain. While great business for the Tunisian trademark office and a few law firms, this artefact does nothing to further the real intellectual property rights of a domain holder. There are businesses in Tunisia whose success depends on the rights conferred by their trademark registrations. The current situation demeans them and could be construed as interference in the sovereign affairs of a nation.

When two organization have rights (in one form or another) to a name, then first-come-first- served is as good a way as any to decide who gets it. Failing that, an auction or bid process would have to be arranged every time there was a conflict. Unless all parties thought of the name at the same time, this auction would punish the early adopter. Since there is no way to arrange this equitably, an auction (or any other mechanism whereby the person who thought of it first has to wait) does not provide for an equitable solution.

25. Should domain name applicants be required to demonstrate that they have a basis for requesting a particular domain name? If so, what information should be supplied? Who should evaluate the information? On the basis of what criteria?

No. I join the many others that believe that this kind of information cannot be properly validated. Companies change businesses and are sometimes sold to other organizations. Any basis/purpose could be mere speculation even if the organization has a legitimate business purpose. This would be needless bureaucratic overhead. Evaluation of such information can only be performed by a body employing due process and more particularly, procedural due process. This implies a courts structure and procedures similar to trademark applications. From a public policy standpoint, such an organization would be contrary to President Clinton's stated goal of not interfering with the growth of the internet. Obtaining a trademark is a costly, time-consuming process. Obtaining a domain name is relatively inexpensive and almost instantaneous. This is one of the 'features' of the internet and should not be interfered with via the introduction of bureaucracy.

26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

The number of different gTLDs could only affect the cost if there were rules stating explicitly that only one registration per domain in a single gTLD was permissible. There would have to be a grandfathering clause to protect the existing rights of companies like IBM (IBM.NET, IBM.COM) and myself (Y2KLAW.NET, Y2KLAW.ORG) in order to prevent chaos.

27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

This should be considered in the same light as the response to 25 above. Valid but conflicting trademark rights, along with valid but conflicting common-law or usage rights cannot be resolved in a single domain. Trademark rights and priorities differ under US and WIPO rules, therefore the trademark law cannot be used to make a choice. In some jurisdictions there is no such thing as a common law right to a mark. That would mean that someone living in France could see their domain removed but not someone in Massachusets.

28. Are there any other issues that should be addressed in this area?

National identity issues have not specifically been covered by this but they are relevant as part of the international emphasis of the internet. If the US government wishes to recognize the international nature of the internet, a good first start would be the adoption of a policy whereby the full UNICODE character set was allowable in domain names i.e. accents, cyrillic characters, kanji, etc. It would require a new version of the BIND software, but new versions are a normal occurrence on the internet. This would permit groups linked by culture to communicate without 'wasting' an English domain name. Those that wanted to be seen by the whole internet would continue to use acsii characters. This would incidentally serve to decrease the number of potential trademark/domain conflicts.

###

Number: 339

From: "Donna L. Hoffman" <hoffman@colette.ogsm.Vanderbilt.Edu>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 4:01am
Subject: Comment on the Registration of Internet Domain Names

Attached is my response to the RFC. Please acknowledge receipt.

Thanks.

DLH

CC: NTIADC40.SMTP40("hoffman@interval.com")

August 18, 1997

Patrice Washington
Office of Public Affairs
National Telecommunications and
Information Administration
Room 4898
14th Street and Constitution Avenue, N.W.

Washington, D.C. 20230

Re: Request for Comments on the Registration and Administration of
Internet Domain Names

Dear Ms. Washington:

This submission is in response to the Department of Commerce's Request for Comments on the Registration and Administration of Internet Domain Names, which appeared in the July 2, 1997 edition of the Federal Register [Docket No. 970613137-7137-01].

I am an Associate Professor of Management in the marketing division, Founder and Director of the Owen Electronic Commerce emphasis, and the Co-Director of Project 2000, all at the Owen Graduate School of Management, Vanderbilt University.

Project 2000, founded by me and Professor Thomas P. Novak in 1994, is a sponsored research program dedicated to the scholarly study of the commercialization of emerging media like the World Wide Web on the Internet. I have published widely on the topics of Internet and Web-based commerce and consumer behavior in computer-mediated environments from the strategic, policy, and measurement perspectives. In the summers, I am a Visiting Scholar at Palo Alto based Interval Research Corporation, Paul Allen's high-technology research and development laboratory, and throughout the year serve as consultant to companies ranging from Fortune 500 multinationals to Net start-ups on Internet strategy and marketing issues.

In winter and spring of 1996, I was an Expert Witness for the Plaintiffs in the joined trials of ACLU v. Reno and ALA v. Reno, challenging the constitutionality of the CDA portion of the Telecommunications Bill of 1996.

Further details regarding the Project 2000 program of research, the electronic commerce emphasis, and my professional background may be accessed at http://www2000.ogsm.vanderbilt.edu/.

The views I express here are my own and do not necessarily represent the views of Vanderbilt University or the Owen Graduate School of Management or the Interval Research Corporation.

In my comments below, I focus my remarks on Section A, Appropriate Principles.

Respectfully submitted,

Donna L. Hoffman
Associate Professor of Management and
Co-Director, Project 2000
Owen Graduate School of Management
Vanderbilt University
Nashville, TN 37203

Comments of Professor Donna L. Hoffman
Submitted August 18, 1997

A. Appropriate Principles

The Internet has changed dramatically and radically since its beginnings over two and half decades ago as a research and communication tool for academics and scientists. In the last four years, virtually coincident with important policy decisions by the National Science Foundation and the introduction of NCSA's, use of that portion of the Internet known as the World Wide Web has exploded (by any measure one would care to employ). From the perspective of the "newbies" surging onto the network of networks every day from every corner of the globe, the World Wide Web is, in effect, virtually synonymous with the Internet.

These elemental changes in user perception of the Internet mirror its definitional, structural, and compositional changes. The Internet is no longer just an open, decentralized set of networks whose users consist exclusively of mostly white, upper-educated, technically competent professional males, all UNIX proficient, engaged in academic and scientific pursuits.

Instead, the new Internet, especially in the United States, is rapidly beginning to mirror the larger general population, while the technically elite and demographically homogenous original Internet community shrinks in proportion.

For example, in the United States alone, 45 million people 16 years of age and older have used the World Wide Web at least once and among individuals using the Web for a year or less, about 45 percent are female. Interestingly, there are now twice as many newbies (30 million using for one year or less) as there are long-time users (15 million using for more than a year). Even more important is that the number of hard-core really new newbies (5.3 million hopped on the Web for the first time less than three months ago and use the Web a few times a week or more) equals the number of hard-core old-timers (5.5 million have been on the Web for two or more years and use the Web a few times a week or more).

So what do these few Internet demographic facts have to do with the current chaos over the domain name registration system? Well, as it turns out, quite a bit.

Just as the Internet user population grows and changes, so does the structure of the emergent industry. Much of the present conflict can be traced to the fact that the Internet is no longer a mere conduit or tool enabling the far-flung and open exchange of binary information, but rather a nascent computer-mediated environment whose globally located denizens are struggling to shape with their own unique identities and preferences. Given that the dominant path for this process is commercialization, is it any surprise that its consequences should be so hotly and vigorously contested?

While the DNS issue may seem, on the face of it, to be a mere technical matter best left to the insular, self-defined and long-standing "Internet community," in fact, what to name a Web server, identifying the best naming process, and determining who shall be responsible for managing this process, are critical steps on the path to a globally successful commercially mature Web.

To date, key stakeholders have largely been omitted from the proposal process, and important issues and concerns subsequently raised by the broader constituent community virtually ignored. The result has been that, despite all the effort and energy employed over the past several years by the long-standing technical Internet community, no single proposal has emerged that satisfactorily addresses the key issues facing this broader Internet community. Indeed, the IAHC proposal, for example, raises more troubling questions than it addresses.

Focusing on the process, however, rather than the deficiencies of any specific proposal, what should we do now? In my opinion, the key "appropriate principle" is to define a process which takes into account not only legal and technical considerations, but business and social ones, as well.

Several sets of new criteria to consider when developing a domain name registration system have been proposed since the IAHC proposal was released, yet none I have studied go far enough nor admit a perspective wide enough to address all constituent concerns.

Thus, assigning names on the basis of trademark or SIC codes, for example, would not solve the naming problem, as would any system devised around product categories, product classes, industry categories, or the like. The key reason is that all such systems are inconsistent with the notion of building a Web-based brand identity for consumers in a competitive market.

In essence, any serious proposal must consider explicitly in its proposed solution how it will impact: 1) the Internet user (i.e. consumer/customer); 2) business (whose interests must be defined beyond the protection of trademarks); and 3) global society. Thus, the stakeholder base must be expanded to include all Internet constituents, not just the long-standing Internet community, and the criteria by which we judge any one proposal must be enlarged to reflect a broader set of concerns.

Such considerations must be based on our current knowledge of Internet users, Web-based business, and how networked societies are addressing the presence of an open, decentralized, many-to-many medium. Such knowledge is considerable, growing every day, and can provide much needed direction to the current debate.

In this way, a solution may be achieved that will satisfy all key criteria (e.g. not just trademark protection and technical feasibility, but also consumer relevance, commercial viability, and the long-term growth and development of the Internet).

Because the Internet is evolving so rapidly, it is critical to seek a solution which is precise enough to satisfy current domain name registration needs (the best categorization schemes are mutually exclusive and exhaustive), yet flexible enough to expand elegantly as the Internet evolves (there will obviously be categories tomorrow we have not conceived of today).

At present, the broad, global Internet community does not possess enough information on which to base an intelligent proposal decision. Open public debate is an excellent first step. Now we must, through argument and consensus: 1) enumerate all key stakeholders, 2) identify their (often competing) needs, and 3) establish the criteria upon which a proposal should be developed. Only then will we have a basis for evaluating any particular proposal to determine whether it best addresses all criteria while simultaneously meeting the critical needs of all Internet constituents.

Finally, any proposal should start with a statement of objectives. It has been taken as a given that the domain name space must be expanded, and quickly, else the Internet may face grave danger. Yet no cogent analysis of this assumption exists. Why do we need to expand the domain name space and what will happen (and when), if we do not?

###

Number: 340

From: "Paul M. Kane" <paul.kane@icb.co.uk>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 12:50pm
Subject: RFC on the Registration and Administration of Internet Domain Names.

Please see the attached WORD document in response to the DoC RFC on the
Registration and
Administration of Internet Domain Names.

Sincerely,

PK.

<><><><><><><><><><><> <><><><><><><><><><><>< ><><><><><><><><><><><>

Paul M. Kane.
Director, Internet Computer Bureau Plc.
Bridge House, 181, Queen Victoria Street, London, EC4V 4DD UK

Before the:

U.S. DEPARTMENT OF COMMERCE

National Telecommunications and Information Administration

Washington, DC 20230

In the Matter of :

REGISTRATION AND ADMINISTRATION OF

INTERNET DOMAIN NAMES

Docket No. 970613137-7137-01

Comments of Paul M. Kane,

Director,

Internet Computer Bureau Plc, London.

paul.kane@icb.co.uk

I, respectfully submit the following comments in this proceeding.

I have been a Director of the above company since July 1996, having worked in the computer industry as a systems analyst/designer since 1983.

I have participated as an invited attendee in at nine public and private conferences at national and international level on this issue.

I have followed most of the discussions on the IAHC discussion list and as European Adviser to the Directory Corporation (a signatory to the gTLD MoU), have followed much of the debate on the gTLD Policy Advisory Board discussion lists.

ICB plc is not a Signatory to the gTLD inorder to remain impartial and objective to the process of Domain Registration. As a small company at the leading edge of Internet technology, we believe technology will not be constrained by arbitary rules. To build consumer confidence in the Internet as a conduit for commercial transactions requires certanty and security. Consequently we have developed numerous applications that use technology to resolve domain name conflicts and promote Electronic Commerce in a stable enviroment.

Paul M. Kane

15th August 1997.

Summary of Comments:

Internet Governance:

1. The global nature of the Internet implies that any one government should not have supra-national control of the medium. Each of the United Nations Organisations, other than WTO, represent the interests of a particular market sector, and the Internet governance needs to remain non-partisan and conscious to the dynamic technology that underpins the Internet.

1. IANA is respected as the "guardian" of TLDs and the "dot". Constitute IANA as a international treaty organisation, with direct funding from Internet constituent members, namely APNIC, ARIN, and RIPE, and ISO 3166 member countries. IANA has already developed a stable, consensus-based self-governing mechanisms. Develop this role by directly involving the private sector, with input from governments through ISO 3166 registries. Unfortunately the ROOT Servers are operated on behalf of IANA by InterNIC which also hosts the .COM, NET, ORG TLDs. Invite IANA to take the operation of the ROOT in-house and InterNIC to host their gTLDs on servers directly funded by them rather than from the NSF funding resources. As time is short, it may be necessary to extend InterNICs contract, under new conditions, to ensure smooth transition to the new management structure.

Top Level Domains.

1. The DNS architecture is monopolistic by nature and consequently the number of gTLDs should be increased, with competition amongst shared and exclusive Registries.

1. The March 1998 deadline has focused many entrepreneurs minds' as all proposals ultimately result in a single entity having exclusive power over a specific gTLD namespace.

1. InterNIC has built a market for Domain Names in one gTLD in particular, namely .COM. They have been successful and after many years of problems now operate an efficient registry. InterNIC has been a victim of its own marketing success. Companies have sought a .COM domain because it is perceived as the "most likely" domain for "international" companies to be found. Browsers automatically insert .COM if the user just types a known trademark as an address. Domain Names are not Trademarks, the DNS is not a directory system and never will be. The direct association of a Trademark, Registered or Common Law, with a Domain Name has caused conflicts to arise. Increasing the gTLD namespace may dilute this conflict if administered correctly, conversely it could exacerbate the problem if mismanaged.

1. As an organisation InterNIC should reappraise its flawed Domain Name Disputes policy although there recently seems to have been a change of policy. The Prince US v Prince UK dispute seems to signal the policy is being substantially reviewed. Their billing system was chaotic, with long standing Domain Names potentially being cut-off, although new registrations seem to be OK. The 30% Internet Structure fund payment for each Domain registration has inflated InterNICs registration fee above its true cost of $35/year. This Domain "rental" fee is not particularly excessive, when compared with say a standard telephone line rentals. As they are a monopoly with relatively static expenditure downward price pressure should be appropriate.

1. InterNIC's current monopoly position must be concluded and replace with competition but the mechanics for its replacement are yet to be evaluated, tried and tested. Stability is of paramount importance. If InterNICs involvement in the industry is to be concluded, a view I do not support, the six month "wind-down" provision in the existing contract assists the gentle tranfer to a new system.

1. The IAHC proposal merits support in my view and its introduction should not be delayed. However, I am unable to substantiate the justification for 7 new domains in one introduction as opposed to one or two at a time. If seven ( some inappropriately named) gTLDs are to be introduced they should be implemented one-by-one based on performance and effectiveness of the system, once proven. If their Domain operation is successful why limit it to seven. Once a domain has been designated it is extremely difficult if not impossible to retract it.

1. The rationale for increasing gTLD was to give Trademark Owners the opportunity to identify themselves using their mark. I remain to be convinced that existing Domain Name Owners will not be compelled to register in each parallel domain "to protect" themselves against passing-off or dilution. In which case increasing the domains in small numbers will have little effect and the proposal of say 150 gTLDs becomes more and more appropriate. There have been a number of Domain Name conflicts involving the .COM domain, until the IAHC system is tried and tested in the market there is no evidence to suggest that the additions of new gTLDs will not just compound the number of domain name disputes. Yes we need the IAHC to compete with InterNIC, but stability and public trust is of paramount importance.

1. Try ONE (or at a push two) new IAHC gTLD initially, to establish operational effectiveness and verification that domain names disputes do not dramatically increase. This incremental route has been envisaged by the gTLD MoU as IAHC have offered no guarantees as to when (if) their proposed gTLDs will be entered in the Root. In addition, Incremental introduction, gives late entrants to the Registrar business the necessary incentives and the ability to sell "choice" names in competition with "first wave" Registrars. This gives the IAHC the opportunity to obtain greater consensus / participation in the registry process.

1. I welcome WIPO making its Mediation and Arbitration service available to resolve disputes as National Courts remain a more expensive alternative option. The ACP process could be used to address disputes in the InterNIC name space as well, for consenting parties. WIPO is not a signatory to the gTLD MoU specifically and remains "neutral".

1. Evaluate the other gTLD proposals over a longer time frame, exclusive and share gTLDs can co-exist providing a firm basis on which to stimulate competition in the Domain market.

Consumer Protection

1. For Electronic Commerce to gain and retain public trust, security is the key to build public confidence. Domain Names do not offer any security because one can navigate the Internet without using a Domain Name (using IP Address) and where domain names are used, traffic can be redirected to a spoof site without consumer awareness.

1. Consumers need to know that they are communicating in the virtual world with the company known to them in the physical world. Consumers and existing Consumer Protection Authorities must be able to identify and marginalise counterfeiters, pirates and passing-off/diluting Internet sites. As a company, ICB has developed a number of technical solutions to reduce the opportunity of fraud in the world of Electronic Commerce.

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name registration systems?

ISO 3166 Registration System

Advantages:                                          Disadvantages:
Efficient                                     
Uncomplicated                                        Controlled monopoly
Regionally Accountable
     Frequently cost recovery basis
Distributed Processing

InterNIC Registration System

Advantages:                                          Disadvantages:
Efficient                                            high profit /cost ratio
Uncomplicated                                        Monopolistic
                                                     Not regionally accountable
                                                     Processing failure/delay affects 
                                                             global network

2. How might current domain name systems be improved?

DNS system

1. DNSs open architecture is fundamentally robust, resilient and efficient. Diversifying the physical locations of the Primary Root Servers would improve global interaction and enhance network efficiencies. This process has started with two Root Servers located outside the USA. Increasing the number of mirror Root Servers to nodes of high traffic activity would assist reduce bandwidth wastage, benefit redundancy factors and enhance the political diversity and international nature of the Internet. (max say 25 Root Servers)

3. By what entity, entities, or types of entities should current domain name systems be administered?

1. The Internet Assigned Numbers Authority has done an excellent job. Long may this continue. Establishing a legal basis for IANA has advantages and disadvantages. Becoming a legal entity makes it a target for litigation and claims for further gTLDs to be added to the Root Zone File are likely to be forthcoming. Due to the nature of the Internet it may be prudent to ensure IANA has protection and framework similar to an international governmental organisations formed by International Treaty.

What should the makeup of such an entity be?

The private sector, with input from governments, should be able to develop stable, consensus-based self-governing mechanisms. The framework already exists with ISO 3166 countries having designated and control over their national Internet Registration Authorities. Constitute IANA such that the national domain registration authorities, administer and pay for the activities of IANA. If and when there are gTLD registries invite them to participate and pay for their share of costs for the administration of IANA.

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)?

1. Rough Consensus seems to work effectively. Network Numbering is best co-ordinated by ARIN, RIPE and APNIC in conjunction with IANA, IETF, IESG and IAB. A Top - Down management model will not work with a dynamic medium such as the Internet. Too many techie rebels !

Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)?

What is the proper role of national or international governmental/non-governmental organisations, if any, in national and international domain name registration systems?

1. National Governments are active in promoting trading activity and co-ordinating consumer protection measures on the Internet. Consumer protection is key to the long-term commercial stability of the Internet. An internationally co-ordinated approach is vital if consumers are to become confident using the medium for transactions.

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation?

1. There are considerable advantages in removing generic Top Level Domains from an administrative perspective but the success of .COM and the significant investments made in the name space indicates that gTLDs are here to stay. The opportunity for retirement has been lost.

Should geographic or country codes (e.g., .US) be required?

1. Yes. Almost all countries other than the US, predominately use their national ISO 3166 code. It is jurisdictional and effective. More use should be made of the .US domain, but it should not be compulsory

If so, what should happen to the .com registry?

1. InterNIC should not have any proprietary rights to .COM domain. It was employed as a contractor to administer an existing asset. It has administered the Domain well and has profited from its activities. Any extension to the contract should be on a cost recovery basis and any legal claims to Intellectual Property clarified in a supplementary contract. This introduces competition between the various gTLD registry groups, without disruption.

Are gTLD management issues separable from questions about International Standards Organisation (ISO) country code domains?

1. A number of ISO 3166 registries seem to think so. Harmonisation of Domain Name application forms would help registries share information on Domain Name holders, but specific regulation is probably best left for determination within each jurisdiction.

6. Are there any technological solutions to current domain name registration issues?

1. Yes. Consumers want to directly associate a company known to them in the real world with their Internet Address. The actual arrangement of characters in the Domain Name is irrelevant, provided a direct association can be made from the Company name to the Web's IP Address. Non-discriminatory directories are the key to navigating the telephone world, the Internet is no different. We have a client who provides www.dir.org, an Internet "telephone" directory and as a company we have a working model of a DNS technical solution.

Are there any issues concerning the relationship of registrars and gTLDs with root servers?

1. Not sure what is meant here. Registrars sell the name space on behalf of Registries. Registrars do not interact with Root Servers only the Registries who have authorised them. If it is suggested that the Registries of any new gTLDs should collectively have their own distributed gTLD Root and Domain Name Servers, running in parallel and connected with IANAs established Root Servers then, there is considerable merit to this argument from practical, consumer confidence, security and political perspectives.

1. The DNS already is a distributed database. National domains work effectively as parallel systems connected to the Root with their own national DNS subsets. If more gTLDs are added later, by any qualifying party, also with their own parallel infrastructure, they will have no detrimental effect on the technical/operational stability of the Root and offers potentially numerous advantages.

1. One of the prime motivations for increasing the gTLD is to remove any one entity's monopoly control in the international TLD name space. The experience of InterNIC scenario, is that by virtue of not being based on a distributed model, it has a contractor role with significant control. When InterNIC's Root Servers were fed defective zone file information on the 16th July, all .COM, .NET, .ORG domains were inaccessible. Also on the 17th July an accidental fibre cable cut between Washington and New York affected UCL, ULCC, RAL and OXFORD in England, CNES in France, all sites in Russia and BBSR in Bermuda.

The new gTLD structure should be more robust and based on a distributed dynamic model, so multiple registry contractors can co-exist servicing the same gTLD. If one registry contractor is unsatisfactory or has down time, another can take its place without unduly straining the system. (Distributed Registry) Failure to introduce this structure could create an InterNIC mirror entity. The IAHC has not specified the detail of its technical structure for its CORE as yet, but an overview appears to be a central processing house, similar to the InterNIC model except with potentially hundreds of authorised "intermediaries" selling the name space allocated in the CORE monopoly name space. InterNIC has no authorised intermediaries but a shared repository containing over 1 million customers.

1. From a political perspective operating CORE as a distributed model enables registry contractors to be spread over a number of countries which fosters a truly international Domain.

1. The provision of an international dispute resolution procedure to reflect the multi-jurisdictional nature of the gTLD name space is a major contribution of the IAHC proposals. If both shared and monopolistic models are to co-exist it would be helpful if WIPO would be willing to make its Arbitration and Mediation services available to any party holding a gTLDs in contention.

7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and co-ordinate?

1. Multiple Sub-domains.- multiple sub-domains within each TLD, either controlled by the Central Registry (Monopoly CORE) or competing registries as a sub-set of the Central Registry, (Distributed CORE).

1. More gTLDs - both distributed and monopolistic

1. Navigation requires effective directory services to identify the site in the virtual world known in the physical world.

8. How should the transition to any new systems be accomplished?

1. Any new system should be complementary and supplementary not in substitution. The existing system works fine, just needs extension, the idea of transition to a new operating environment may introduce more problems than it solves, and should be undertaken gradually, with each element subjected to real time evaluation.

1. The IAHC propose seven new gTLDs using a new model of multiple sales outlets called Registrars. The IAHC's progressive approach will enable new Registrars to join their system of Domain registration as an evolutionary process. An organisation known as CORE is providing Registry services in a monopolistic environment. It is not clear if the CORE Registry is a monopoly or a distributed competitive process comprised of competing data processing houses servicing CORE. Either way the IAHC recognise that all seven gTLDs may not be added at once. This incremental scenario is wise.

1. An incremental approach would enable all facets of the new process to be examined and tested. It may be for example, existing Domain Name holders feel compelled to register in parallel domains, frustrating the reason for increasing the name space. It may be that the introduction of additional gTLDs leads to an exponential increase and compounding of domain name disputes. Either way it would be prudent to introduce one or two of the new gTLDs now with more later depending on the market realities. This incremental approach also enables Registrars who do not enter in the first wave to share in "choice" domains of the new gTLDs as they are introduced. Failure to introduce an incremental approach could lead to new registrars effectively being excluded as the "choice" names would have already been taken.

Are there any other issues that should be addressed in this area?

1. Increasing the gTLDs name space should be considered supplementary name space rather than an Internet governance issue.

1. IAHC seeks to build a gTLD reputation for .STORE, .WEB, .PER, .ARTS, by determining the parameters of allocation within the designated name space.

1. The original intent of the IAHC was to increase gTLD name space, not to provide a platform for Internet governance. The IAHC introduced its Policy Oversight Committee for administration of its policies but unfortunately this title has been misconstrued as has inadvertently become entangled and associated with governance of the whole Internet Domain Name space. Control of Domain Names is not a foundation for Internet Governance and control of IANA by one gTLD interest group, should be resisted.

1. The Internet is driven by private enterprise. Satisfying and protecting customer interests on the Internet, should be the role of governance.

1. We do not believe this was the IAHC's intent and certainly Internet governance by such a body (without national representation) should be rejected as it is technically impossible to enforce "governance" in the traditional sense. The multi-jurisdictional in nature of gTLDs prevents Nation States from being able to protect their citizens interests and regulate their markets in the traditional sense. Consequently any "governance" should not be directed towards fallible Domain Names but protecting consumers against fraud and deception by co-operation by consumer protection authorities at an international level.

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

1. Yes.

2. Technical: - the caching rate of the DNS System, when gTLDs are in the tens of thousands.

1. Practical - the more gTLDs the more opportunities there are for counterfeiting, even with stringent domain name allocation procedures. The IP Address controls navigation around the 'Net not the domain name.

2. During July consumers were confused when identical names in different gTLDs were used with one service provider emulating the other. Consumers were duped into paying twice the normal price.

1. Policy - The more gTLDs the higher the potential for deception. .- The larger the number of gTLDs the more demand for judicial determination and dispute resolution, by mediation, arbitration or litigation. A means of confirming that a site owner is who they claim to be is vital for consumer confidence. Verification and SSL certification offers a degree of consumer protection.

1. We support IAHC's objective to administer a dispute resolution process for their allocated domains and determining which Registrars will be eligible to sell their domains. The Internet community appears to want an enlarged gTLD name space and the IAHC is just one of many organisations potentially able to satisfy that demand.

11. Should additional gTLDs be created?

1. Yes but incrementally. Introduce one or maximum two new gTLDs as a trial. The gTLD system is non-reversible and revocation of existing gTLDs is near impossible from a consumer perspective. A large section of the market has demonstrated it wants gTLDs. Certainly there is need for the DNS structure to be enhanced to enable consumers to identify their chosen retailer on the Internet, even if they have a trading name that is similar or identical name to that of a competitor. Only one Domain Name should be allowed per trading entity. Much of the .COM name space is unused and held by "cyber-squatters". It may be that increasing the gTLDs will cause multiple parallel registrations in each new gTLD thereby frustrating real expansion in namespace. Selling Domain Name space is big business. If the .COM remains, enlargement is inevitable and should be encouraged to enable the existing monopoly to be broken.

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

1. The methodology of creating scaleable sub domains to ensure the scalability of the name space within each new gTLD was briefly discussed but was rejected in the light of pressure from the Registrar community, because Third Level Domains are a harder sell, than SLD. Rejecting the use of third level domains is a mistake if the number of gTLDs is to be kept small. The act of tiering the name space makes it scaleable, selling third level domains like "xyz.CLOTHES.web" works. If a local shop only trading in France, why not allow "xyz.VETMENTS.web".

2. The crux is to be able to identify the shop known to the consumer in the physical world in the virtual world, irrespective of its Domain Name.

3. The technology exists but is not being applied, because the payback time to recover the investment is too long for most private enterprises.

13. Are gTLD management issues separable from questions about ISO country code domains?

1. ISO country codes are jurisdictional in nature, gTLDs are potentially multi-jurisdictional. Management should be viewed separately to ensure national laws are respected and multi-jurisdictional interests are accountable..

14. Are there any other issues that should be addressed in this area?

D. Policies for Registries

15. Should a gTLD registrar have exclusive control over a particular gTLD?

1. A Registrar should not have exclusive control over a Registry offering a particular gTLD. Conversely, if a Registry offering a particular gTLD has control over a particular Registrar, the consumer has the potential for demanding enhanced service and a framework for compliance with standards.

Are there any technical limitations on using shared registries for some or all gTLDs?

1. Depends how the shared registry is configured. Technology can be liberating or constraining. Shared and monopolistic gTLDs can co-exist and this will remain. Opening up .GOV or .MIL for example would be unwise. Introducing user specific gTLDs may be appropriate.

Can exclusive and non-exclusive gTLDs coexist?

1. Yes, and their viability will depend on the reputation, quality of service and price of each gTLD.

16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have?

1. The IAHC have set good threshold standards and parameters for registrar responsibilities. The technical competence of each applicant should also be assessed. Enforcement procedures to ensure conformity to the registrar's standards need to be addressed. Will Registrars be removed if they fail the set standards (with potential liability for wrongful dismissal) or will the standards be dropped?

Who will determine these and how?

1. If IAHC is responsible for the IAHC gTLD brands and XYZ is responsible for the ZYX gTLD brand, standards will remain high as reputation, price and quality of service and competition between IAHC and XYZ will distinguish the "brands" and determine their success.

17. Are there technical limitations on the possible number of domain name registrars?

1. No. The more Registrars the more complicated the process for the Registry to monitor progress.

18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

1. Price. The InterNIC Registry charges consumers $100 per domain for 2 years, 30% of which goes to the Internet Structure fund. The real cost of an InterNIC domain is therefore $70 or $35 per year. Is the new gTLD registry system going to see consumers paying Registrars less than $35/year per domain? What happens if the basic domain registration fee payable by consumers is more than $35, is it anti-trust, a cartel or subject to competion law investigation?

19. Should there be a limit on the number of different gTLDs a given registrar can administer?

1. My preference is for only a few gTLDs, but the act of introducing one more gTLD is a slippery slope to their being hundreds. If all gTLDs are to be accessible to all equally, users and Registrars alike, then access must be unlimited. The effect of no limit would be to drive prices down and increase standards as gTLD Registries compete. In the short to medium term the growth rate in gTLDs may be exponential and the number of gTLDs and Registries may number in the hundreds and Registrars in the thousands in a very short period of time. Conversley, there may hundreds of Registries each with their own gTLD yet few Registrars, so the consumers must be able to buy directly from the Registry.

1. If the number of gTLDs is restricted, the central administration for DNS (IANA) may have calls of restricted practices raised and claims of operating a cartel with litigation forcing them to open up the Root Zone Files to incorporate other gTLDs. If the IAHC is the only gTLD Registry, then competition within the registry industry will be constrained and CORE is likely to be sued in jurisdictions where it is not possible to contract out anti-competitive practices.

Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

1. A Registrar submits applications to a Registry and has no explicit control over the registration practices of the Registry. The Registry administers the DNS for the gTLD and "owns" the gTLD under its control. The Registry can be exclusive, like InterNIC, but as mentioned in 6. above, should be distributed shared if a monopoly situation is to be avoided.

1. Currently, Applicants can apply to InterNIC for a gTLD or they can use the services of a Registrar. Registrars provide value added services to Applicants helping them to apply for a Domain Name. Both options provide direct access to the Registry and via the additional services of the Registrar should be available to Applicants in the new gTLD framework. Thus in an exclusive scenario, the applicant should be able to apply to the Registry directly or via a Registrar if additional value added customer services are required. In a non-exclusive scenario the Applicant should have to use the services of a Registrar, who compete with each other in the process of "forwarding" the application to the application to the Registry.

20. Are there any other issues that should be addressed in this area?

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-à-vis domain names?

1. Trademarks and Domain Names are two distinctive identifiers used by consumers to associate a particular product or service with a specific company. The two are not related. Even if one had exhaustive procedures for vetting Domain Name applicants, consumers could be duped by a technically aware DNS wizard to make the domain name an irrelevance when accessing their spoof Web site. (Email is more robust) A suggestion is to tackle the "virtual" problem in the same way as it is established in the physical world..

1. In the physical world, the Trademark Owner, be they registered, common law or regional rights, has a duty to police and protect granted intellectual Property interests using the courts of a specified jurisdiction. In the virtual world the same scenario is possible, if not imperative. Trademark Owners should take active steps to identify themselves to their consumers. One of our clients, http://www.trademark.org, have been proactive in this area. Their aim is to enable consumers to identify Trademark Owners, their Web Address, irrespective of their domain name, with security and confidence using Intellectual Property Practitioners as the third party vetting authority.

1. Traditionally Trademarks are identified on registers or in practical use. The Trademark Owner must apply to be registered or must be active in using a mark, then must police and be responsible for protecting their derived intellectual property. In the virtual world, technology can support the generation of physical evidence in the event of counterfeiting. Such technology should be used to enable judges to make determinations if and when fraud has occurred.

1. Preliminary review of an application requires determination. Such determination may be contentious and has little benefit if the objective is to protect a Trademark Owner. Technology at the control of a fraudsters does not respect a Domain Name application process, however diligent. Recently, someone used a flaw in the DNS system, to technically manipulated the "official" Servers to recognise their alternative gTLD. This flaw has been known by many for some time. The distinction being that in this instance this person sought publicity for their actions. If this attack had been malicious the technique could have been used to map and divert the .COM domain, traffic destined for genuine .COM sites and the consumer would not have known.

1. Promoting domain names as the means of identifying verified site owners as a basis for conducting electronic commerce on the Internet , is without foundation as it will only serve to boost the opportunity value for committing domain name deception as technology will not be constrained by arbitrary rules.

1. A facility operated by one of our clients, makes it possible to identify that the site operator is who they claim to be and is the genuine company and holder of a given Trademark, that has been checked by an Intellectual Property practitioner. The service is currently registering IP practitioners and I understand it will go live once 10,000 IP practitioners have enrolled, giving access to a projected 1 million Trademarks in a secure forum.

23. Aside from a preliminary review process, how should trademark rights be protected on the Internet vis-à-vis domain names?

Preliminary review inherently requires discretionary judgement. This could be contention and lead to litigation. In addition it is futile, please see 22 above.

What entity(ies), if any, should resolve disputes?

71. Mediation, Arbitrators and the Courts.

Are national courts the only appropriate forum for such disputes? No

Specifically, is there a role for national/international governmental/nongovernmental organisations?

72. Yes

24. How can conflicts over trademarks best be prevented?

73. Use technology to marginalise the counterfeiters and fraudsters. We have an operational model and procedures in place today!

What information resources could help reduce potential conflicts?:

74.i) databases of registered domain names - already operated by the Registries. One cannot have two identical Domain Names therefore first come first served model is operational with appropriate databases.

74.ii) registered trademarks - is already available, numerous search firms provide Registered Trademark search facilities, which is far from complete as it excludes common law marks or regionally granted rights.

74.iii) trade names - difficult to assimilate, and of questionable value in the domain registration process due to many identical names.

75. If there should be a database(s), who should create the database(s)? How should such a database(s) be used?

76. There are already numerous databases, provided in the private sector. Their use is directly related to how they are funded. If it is free to be in the database, the charge is to view the data. This inhibits use of the database and has the potential to include information that is not accurate. If it is free to view the database, there is a charge to be included, and to be an authoritative listing the information must be verified. Depending on the scope of verification required before inclusion in the database determines the charge of entry to the database.

77. To encourage open access to information, an authoritative database that is free to search, further marginalise the activities of fraudsters. It is harder to compile but offers more long-term benefits to the community at large.

25. Should domain name applicants be required to demonstrate that they have a basis for requesting a particular domain name?

78. Yes, to obtain additional support from the Registry if the Domain Name is subsequently challenged.

If so, what information should be supplied?

79. Confirmation of organisational / company name, Trademark or family name. Documents can be scanned and attached to files or just the salient details noted where the information is available in the public forum.

Who should evaluate the information?

80. The Domain Name Registry not the Registrar.

On the basis of what criteria?

81. One domain name per person or trading entity or brand.

26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

82. The Registrars are "post-boxes" for the Registry. The number of Registrars is unlikely to have any significant effect on the cost of resolving Trademark Disputes. The more gTLDs the fewer the conflicts if an effective means of identifying the Domain Names of Trademark Owners on the Internet is used. Without an efficient means of identifying sites of Trademark Owners, increasing the number of gTLDs will lead to an exponentially higher the number of passing off, piracy, cyber-squatting, infringement and dilution and inordinately higher costs.

27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

83. Yes. That is what this company has been working on since I attended the WIPO meeting in May. We now have a working model !

END of submission.

###

Number: 341

From: <laina@singnet.com.sg>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 10:49pm
Subject: APIA comments

Dear Patricia Washington of the office of Public Affairs,NTIA

The Asia & Pacific Internet Association is a newly formed (May 1997) industry association of
Internet related service providers from the region. It was born from the efforts of the Asia
Pacific Networking Group, Commercial Working Group over the last two years.

Many in this region have felt that the need for formal channels to pass on their input to policy
making bodies, such as that of your own. Although still a fledging organisation, APIA hopes to
be this channel. Having the key "movers and shakers" in the region as the Board of Directors,
Advisory Board members, Founding Members and Members, we hope that this submission will carry
some weight as industry input from this region.

APIA wishes to applauds NTIA for this process of open consultation, and thanks NTIA for giving
us the opportunity to have our opinions heared.

Looking forward to further collaboration with NTIA in the future.

Regards,

Laina Raveendran Greene
Secretary General
APIA Ltd

PS This submission will also be posted on our webpage at http://www.apia.org

-------------------------------------

Name: Laina Raveendran Greene
Secretary General, APIA
http://www.apia.org
E-mail: laina@getit.org
Date: 8/18/97
Time: 10:49:45 PM

This message was sent by Chameleon
-------------------------------------

CC: NTIADC40.SMTP40("apia-board@apia.org","dfidler@cis...

Before the

U.S. DEPARTMENT OF COMMERCE National Telecommunications and Information Administration Washington, DC 20230

      In the Matter of                    )
                                          )
      REGISTRATION AND ADMINISTRATION OF  ) Docket No. 970613137-7137-01
      INTERNET DOMAIN NAMES               )

Comments of the Asia & Pacific Internet Association (APIA)

Board of Directors

Dr Jin Ho Hur, Inet, Inc, Korea
Toru Takahashi, Tokyo Internet, Japan                        c/o Laina Raveendran Greene
Pindar Wong, VERIFI, Hong Kong                               Secretary General
Barry Greene, Cisco Systems, Singapore                       7th Storey RELC
                                                             Building
Dr Tommi Chen, Asiapac.net, Malaysia                 30, Orange Grove Road
Prof Li Xing, CERNET, China                          Singapore 258352
Roger Hicks, Clear Communications, New Zealand               laina@getit.org

[18 Aug 1997 ]

TABLE OF CONTENTS

Summary

A. Appropriate Principles

Principles a-f

Other principles

B. General/Organizational Framework Issues

Questions 1-9

C. Creation of New gTLDs

Questions 10-14

D. Policies for Registries

Questions 15-20

E. Trademark Issues

Questions 21-28

F. Other Issues

[List other issues addressed]

SUMMARY

POINT 1. Applaud open consultation

APIA is pleased to support the US government attempt to open up consultation on the future of Internet Governance and recognizes that the US government by historical accident, has "ownership" of these issues. With the commercialization of the Internet and the Internationalization of the Internet, the time has come to review the current administrative structures which are "historical accidents". Much as APIA wishes to acknowledge the good work from the US government and US subcontractors such as IANA in keeping the Internet as one global network, it also recognizes that the time has come to review this existing system today.

POINT 2. New legitimate structures needed to reflect the commercialization and internationalization of the Internet.

With a new environment of commercialization and internationalization, the new structure should reflect this carefully. It could for instance, be some sort of a non-governmental international organization, with private sector at the helm and market forces as its driving force. Competition and market decisions should be encouraged as much as possible, within a framework of coordination and collaboration to keep the Internet together. Policies and guidelines of this framework should be made only after open consultation and input especially from the private sector.

Nevertheless, the governments of the world will need to give their "blessing" to these structures and frameworks, to ensure stability and legitimacy. Business interest demands that legitimate and stable frameworks are needed to minimize business and legal risks.

POINT 3. Internet governance should not just focus on domain names

More importantly, the new structure should not just be created to focus on domain names. It should also look at IP addresses, standards processes, financing of the Internet bandwidth costs globally, etc. Basically guidelines and policies that will help advance the vision of a Global Information Infrastructure (GII) and the various National Information Infrastructures (NII). The underlying principle to all these efforts should be interoperability and interconnectivy.

POINT 4. Domain names should not be presumed to be trademarks

Much of the debate on reform to date has focused on trademarks, due to a dozen or so trademark cases. APIA feels that domain names should not automatically be equated to trademarks, and unless there is legitimate infringement, trademarks and other such IP rights should not dominate the solution. In other words, APIA feels that more international study on these issues is needed before attempting to tie legal issues into a solution and create extensive frameworks to solve these handful of problems. There are many other appropriate forums for legitimate cases to be handled.

Scaling and the viability of the Net is more important than protecting the interest of a few. Also, domain names were created to make the Net easier to use, not harder. Where there are technical solutions they should be pursued instead. For example, APIA would like to suggest new naming methodologies that could include numbers and international languages, browsers that will lessen dependence on domain names and directory services. These will help solve scaling issues and recognize the global nature of the Internet.

POINT 5. More international consultation needed to ensure attention to cultural sensitivities and the needs of the developing world..

APIA feels that it is imperative that before new complex structures are created, careful study of the needs of the industry are made first. This includes industry from around the world, and APIA will be pleased to help in this effort from the Asia & Pacific Rim region. In this respect also, APIA also feels that more emphasis should also be given to cultural sensitivities and the needs of the developing world, who are only just catching up to the various issues. For example, domain names in one language can be very offensive in another and local needs also call for non-English computations.

POINT 6. Prudent solutions only after open consultation, and technical solutions not legal ones should be pursued first.

In conclusion, APIA feels that no one solution should be hurriedly applied, and where necessary existing mechanisms should continue (e.g. short term extension of NSI's contract). Prudence is needed to determine that any new mechanisms is devised only after greater consultation and input, and it should then be implemented smoothly to avoid business risks. It is therefore also important that the governments of the world endorse any such private sector solutions, to ensure that decisions are respected by the courts and regulators around the world. This way, legal risks will also be minimized. However, where there are problems, technical and not legal solutions should be pursued first. Lessons from the telecom world, where appropriate should also be considered.

Before the

U.S. DEPARTMENT OF COMMERCE National Telecommunications and Information Administration Washington, DC 20230

In the Matter of ) ) REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01 INTERNET DOMAIN NAMES )

Comments of the Asia & Pacific Internet Association

1. The Asia & Pacific Internet Association (APIA) respectfully submits comments in this proceeding. APIA is a regional trade association, formed in May 1997, whose aim is to promote the business interests of the Internet related service providers in the Asia and Pacific Rim regions.

The goals of APIA include developing and promoting policies and positions in the best interests of their members and represent these interests to the relevant bodies. To date, there has been no formal channel to give a voice to Asia Pacific Internet concerns and APIA aims to fill this gap. Since, issues such as the registration and administration of domain names is very relevant to APIA and is of great concern of our constituents, APIA has decided to participate in this Request for Comments.

A. Appropriate Principles

Whether there should be competition or not, depends whether we are referring to competition in the root server, the Top Level Domains or the Second level Domains. APIA strongly supports competition in TLDs and SLDs, provided the necessary coordination activities and sharing of databases are done to ensure interoperability of the network.

Where competition in the root server is concerned, more research is recommended to see whether this is technically feasible. Again, the principle of interoperability will be key, as APIA would not like to see the fragmentation of the global Internet.

APIA agrees that private sector mechanisms will have its definite advantages, however more than just input, APIA feels that government backing/endorsement will be needed. This applies not just to the backing of the US government, but of international governments, as without such "blessing", the mechanisms will lack stability and legitimacy. From a business perspective, international legitimacy will be a key principle, as the community cannot afford to have differing solutions backed differently by governments and courts, in differing jurisdictions.

Government backing should not, however, mean that government or any inter-governmental organization should take an active role in the management or policy making of the DNS. In fact, this applies to all Internet governance issues, not just DNS. Rather those functions should be left to private sector and determined on market principles, but government should recognize these mechanisms as legitimate.

c. These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.

Indeed, the Internet is no longer a US centric network. It is today a truly global network of networks. The governance mechanisms of the past came out of an informal academic and research community, with key personalities as pivots of such informal structures. Besides this type of Internet culture, much of the informal mechanisms stem from a US culture of private enterprise and of suspicion of government involvement sentiment.

Today, the Internet is global and recognition needs to be given to the diverse international approaches to Internet governance. It may no longer be appropriate to determine the future of Internet governance based on its historical accidents or personalities. More importantly from APIA's point of view, the commercialization of the Internet calls for more formal structures which provide legitimacy and stability to protect business interests. Internet governance today has to face up to these changes.

d. The overall framework for accommodating competition should be open, robust, efficient, and fair.

One can only underscore the need for openness, and fairness. Stakeholders of this network need to have their mechanisms for input. Having decisions that impact business interest, without a mechanism to minimize negative impacts will surely be detrimental to the future development of the Internet. As for robust and efficient, this comes back to the previous comment on stability and legitimacy, however with an aim to avoid bureaucratization of the process. The Internet is a dynamic network and its administration has to be too.

e. The overall policy framework as well as name allocation and management mechanisms should promote prompt, fair, and efficient resolution of conflicts, including conflicts over proprietary rights.

APIA feels that the administrative framework should be prompt, fair and efficient, but does not feel that this necessarily applies to conflicts over proprietary rights. The history of the NSI's involvement in trademark issues is clear evidence that any future administration should stay clear of such issues.

If such a messy litigious situation within NSI is created in a single jurisdiction(US) with a single federal law on trademarks (perhaps with state variations), would this not be ten times worse in a situation where there is no universal Intellectual Property laws in this regard. Global TLDs such as .com, .net, .org, etc are global while trademarks laws are national. APIA feels that it will be premature to have a global administrative body resolve Intellectual Property conflicts without clear international laws and only a dozen or so cases as the basis. APIA is therefore particularly concerned about the International Ad Hoc Committee proposal to adopt the creation of the Administrative Challenge Panel , which in fact protects rights that are not universally recognized, and is even more than that found in the US Federal Trademark Dilution Act of 1995. (The adoption of presumed versus proven infringement, which is far greater than any existing national law to protect the interest of the big trademark holders, is of great concern.) The use of the WIPO Center for meditation and arbitration on the other may have its benefits, but APIA feels more study is required before such mechanism is made compulsory.

APIA would rather support efforts to try and diminish reliance on domain names as trademarks, such as browser, directory services, etc. Meanwhile, while APIA does recognize that there are some legitimate instances where domain names can be equated to trademarks, APIA would rather not support an approach that assumes that all domain names are equated to trademarks. Trademark laws came out of its own history and so did domain names. Applying one to the other should be done with prudence otherwise we would stifle the growth of the network. Therefore, for example, APIA feels that actual infringement should be proven through the normal procedures rather than adopting mechanisms within the domain name administration that assumes infringement.

f. A framework should be adopted as quickly as prudent consideration of these issues permits.

Given that there is no real threat to the fragmentation of the Internet to date, APIA feels that decisions have to be carefully made (not quickly) after consideration of the different viewpoints and interest (prudent). If a technical solution to this problem can be found, then it should be pursued rather than trying to solve a problem that is not really there. A handful of domain name hijacking cases and competing TLDs such as that of ALTERNIC, does not justify rushing into constructing a new global structure, without due consultation. If decisions cannot be made before the end of the NSI contract, consideration should be given to extending it for a short term, to allow for smooth transitioning to a new system.

APIA would also like to highlight that to some extent, the telecommunications world faced a similar situation. There was a time when telephone numbers were alphanumeric but they ran into similar problems as we are facing with domain names. Rather than create complex solutions to those problems, the issue was solved at the root of the problem. APIA is not necessarily implying that we should go back to numbers (as this may not be practical to rename the existing names), however would like to encourage other more practical solutions if they exist.

Other principles

Another important point that APIA would like to note is that domain names are only part of the issue. In fact, to some extent it is the lesser part of the issue. What needs to be addressed is the bigger picture of reform of the Internet Governance issue. The time has come to review whether a hierarchical administrative structure with a US government contractor at the helm (IANA) is still relevant in today's global commercial Internet infrastructure.

The principle of the need to internationalize Internet Governance will have to be considered. IANA will have to be converted to some sort of International organization with governmental blessing and private sector input (perhaps the creation of a new Internet International Organization). It would need accountability to the stakeholders of the Internet, allow for global input and be "blesses" by governments to provide legitimacy and stability. One could also consider IANA being converted to a membership organization much like APNIC and RIPE. The point stands that the time has come to relook the current structure.

Another part of this process of internationalizing is the promotion of different cultures and languages on the Internet. There are many cultures where the Internet has yet to develop to the extent seen in the US and other "European" countries. These cultures must not be disadvantaged in developing their own NII and connecting to the GII for their own cultural and economic activities. The governance structures must be suitably international so as to provide the leadership and support. For example, it should endorse domain names not be based on the Roman alphabet? Why not support Korean, Hindi names etc?

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name registration systems?

One clear advantage is that the current system has kept the Internet together as one network to date. Keeping one root server in a hierarchical naming structure definitely had its benefits to keep the network interconnected and interoperable.

Many disadvantages, however, stem from the fact that policies are set arbitrarily without any open processes to develop guidelines on the allocation of domain names. Much of it has either been an IANA decision or the registrars of the allocated exclusive TLD. As a result, especially with national TLDs, there are many inconsistencies in their allocation. Apart from this fact, in some countries IANA may allow private entities to be registrars while in others allow government entities who insist on control. Some restrict national TLDs to the country geography while others do not. The ones who do not restrict national TLDs to their countries, therefore in fact become global TLD registrars. This has lead to abuse such as marketing gimmicks.

Another disadvantage includes the fact that domain names have not quite scaled as people have begun to use domain names as trademarks. The number of computations are immediately reduced. Furthermore, since only Roman alphabets are used, this has further reduced the scalability of the names globally. Besides scaling issues, the use of Roman alphabets leans in favor of languages using such characters and does not offer the opportunity to support diverse names on the Net. Just like postboxes, while they need to be standardized for "network" recognition purposes (Internet Protocol addresses), names can be personal to the individual culture.

The greatest disadvantage APIA contends is the imputation of legal issues into what was intended to be addressing mechanisms. Domain administrators should not be directly involved or actively involved in legal issues, but rather function more as a listing service. No doubt they cannot run away from all legal issues, but these should be restricted to those involving them as any other organization or legal entity.

2. How might current domain name systems be improved?

The key improvement to the current system would be the improvement of PROCESSES. There needs to be open and fair processes to ensure that input from stakeholders can be gathered and considered, i.e. OPEN CONSULTATION.

Next, legal issues should be left to the appropriate authorities to decide, and not have policies created within the DNS administration around the uncertainties of the law.

Competition in TLDs with coordination and future introduction of non-Roman alphabet names will also help. The bottom line however, is that the market should decide what it needs and wants.

3. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

With the internationalization of the network, there needs to be a new type of international organization (non-governmental, but with government "blessing"), with private sector at the helm of the structure. (There is no reason why government or any intergovernmental agencies such as the ITU and WIPO should be part of the management and policy making structure.) One could consider either making IANA the basis of creating a new type of international organization or make it a membership organization much like RIPE and APNIC but with government endorsement of the structure. IANA or ISOC its current form and within the IAHC proposal, as the veto body and at the helm of the administrative structure is however unacceptable.

The commercialization of the Net, demands more formal structures with government backing (not just US government backing) or endorsement to ensure stability and legitimacy. Business interests cannot afford to be part of a system open to countless litigations and insecurity.

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

APIA endorses the need to learn from other arenas which may have had similar problems. The network numbering plan in fact had the alphanumeric problems of the domain name world, and we could learn from how that was handled. While spectrum allocation is not a good analogy to domain names as it is a scarce resource and domain names a private resource, one can learn from the system of registration and coordination, rather than allocation. To a large extent, even there it should be market forces and not some central coordinating body that decides what should be allocated.

The ITU-T model of consensus building among private sector, with the necessary government backing may also serve as a model to start from. There are however many problems with that system, and the Internet world has to be careful not to import it wholesale. The standards setting process, however, does allow for market decisions to be made and consensus building helps ensure interconnectivity and interoperability of standards adopted. This is what the domain name system can definitely learn from.

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the .com registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

Retiring it completely from circulation will be highly impractical. There will be business costs associated with renaming. Retiring it from future uses may have some validity, however this decision will have to be left to the market to decide. Also, merely retiring it without diminishing the role of trademark law will not solve anything as application of dilution laws or its enlargement (such as applied by the IAHC Administrative Challenge Panel) will mean that companies will block the use of other TLDs as well, creating new problems.

Likewise, country TLDs should also be determined by market needs. Countries such as Cambodia should not be forced to use .kh in today's world. They are more apt to decide what is appropriate to them than some central body who decides without their input. Even ISO country codes were developed for a separate purpose and developed at a separate time.

6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

Not known, but worth pursuing.

7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and coordinate?

Of course this question assumes competition in root servers. APIA will strongly suggest more study to see if this is technical and practically possible, to ensure non-fragmentation of the Internet. Nevertheless, sharing of databases and simultaneous registration in all root servers, are some mandatory principles need to be adopted to ensure interoperability and interconnectivity.

8. How should the transition to any new systems be accomplished?

Prudently and smoothly. If necessary, NSI contract should be extended for 6months to a year to allow for more open consultation and input. This way, a more acceptable, coordinated and universal mechanism can be put into place.

9. Are there any other issues that should be addressed in this area?

Just as with telecommunications networks that laid out common principles for interoperability and interconnectivity, it will be important to have a framework dealing with other such issues, such as international standards of quality of service, international guidelines on how bandwidth for the Internet is financed etc. The latter issue is important to APIA, as the Asia Pacific region ISPs are still paying for the entire bandwidth into the US, even though the 100:0 ration of traffic is no longer true. APIA would like to call for governmental backing to support industry to develop guidelines to ensure a more equitable settlement of costs in building the Global Internet Infrastructure. These issues are equally important to ensure the continued viability of the Net and its growth.

In other words, organizational changes should not just focus on domain names. Domain names are just but a tool that was developed to help the Internet to grow. There are many other tools that need to be coordinated and international standards developed, that such a "new" organization should focus on, beyond just domain names.

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

gTLDs , unlike IP addresses or telephone numbers, are not technologically essential to the operation of the Internet but provide the ease of use of alphanumeric identifiers. There should therefore be no reason why there should be any technical constraints to this. In fact practical and policy considerations will demand its expansion as the use of the Internet grows.

11. Should additional gTLDs be created?

From a cultural perspective, there is a call to have additional TLDs created to accommodate the different languages and characters. Creating gTLDs with numbers should also be considered. However the decision to increase should be left to the marketplace and not to some decision making authority. Even if decided by such body, it should be done only after open consultation and business input.

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

Efficiency and effectiveness of use will be more important than the technical limitations of the issue. There will have to be sufficient mechanisms in place to avoid hoarding. Global coordination and consistency of policies, will also help ensure effective usage. As such, APIA is concerned that the IAHC stance of putting trademarks as the crux of the mechanism, as this will seriously affect scaling as big trademark holders exclude others from TLDs using extemsions of trademark dilution laws. This we believe goes beyond the intention of trademark laws in the first place, and is not an appropriate extension to this new medium of the Internet and domain names, which have different purposes as well.

13. Are gTLD management issues separable from questions about ISO country code domains?

APIA recognizes the problems associated with .us TLD, and sees no reason to distinguish between gTLDs and country TLDs. In the end, the price, the market needs and the service will determine the final choice of the user. Furthermore to date, the ISO country code may impose non-user friendly names, rather than allowing the country to decide an appropriate code (for example, Cambodia is .kh reflecting the days of the Khmer).

To date, each national TLD lays out its own policies, and this leads to inconsistencies which affect the others in the community. For example, some national TLD registrars insist that it be used geographically, while others do not. Those who don't are in reality a gTLD registrar, which is why APIA feels that there needs to be greater coordination in TLD policies just as for gTLDs, and there may be no real reason to distinguish between country TLDs and gTLDs.

14. Are there any other issues that should be addressed in this area?

The charging policies of TLDs are important issues as well. It may be useful to reflect why telephone numbers and names we choose for ourselves are free, while domain names are charged. This is particularly important if there is only one TLD organization. If charges are imposed, then the body imposing these charges have to be a legitimate authority and clear accountability of how the money is used needs to be established.

If however, the position is taken to have competition in TLDs, then of course the market will decide the legitimate body and the price it feels is right. Likewise, the same argument applies for SLD registrars, depending on whether there is competition or not.

D. Policies for Registries

15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

There does not seem to be agreement among experts whether this is possible or not. More study will be required to see if this is technically possible and more importantly, if this is what the market wants and needs. As far as APIA is concerned, if the existing system if improved (incorporating the international nature of the Internet, open consultation, incorporating private sector input and participation, having the requisite inter-governmental backing/endorsement, charging only for costs for domain names, etc.) then there will be no real reason for competition. Nevertheless, APIA believes that competition will certainly help solve many other problems, provided there has to be clear and consistent guidelines globally to avoid registrar shopping (i.e. shopping for reasons other than for price and service).

16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have? Who will determine these and how?

Yes for the protection of the users, but this threshold has to accommodated to the differing legal systems and corporate structures. The one proposed by the IAHC is too US-centric. The legitimate body (to be decided after greater public consultation) would be the one who determines the guidelines and its accommodation to local needs. Otherwise, ideally where there are sufficient consumer laws within each country, it should be the market that decides this threshold.

17. Are there technical limitations on the possible number of domain name registrars?

Not known. More study required.

18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

The need for coordination should be underscored. The key principle should always be interconnectivity and interoperability.

19. Should there be a limit on the number of different gTLDs a given registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

APIA believes that this deserves extra study. Again market requirements should prevail within reasonable limits of what the technology will allow.

20. Are there any other issues that should be addressed in this area?

International standards for the quality to the network and market/consumer laws to ensure quality to the consumer are also key issues to be considered as well.

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-à-vis domain names?

Except for a dozen or so cases, there has not been much indication that business treat domain names as trademarks. APIA of course, recognizes that companies who have spent resources to protect their trademarks globally want to be protected even with respect to domain names, but this we feel do not reflect the majority opinion. It is more important to others to obtain domain names to make a commercial presence on the Net, than to exclude others from the Net. Therefore we feel that before devising extensive administrative structures to protect trademark interests, it is important to study the issues.

Domain names are global while trademark laws are national. More study is required to see the impact of this difference. More importantly, APIA feels that more study needs to be done on the relevance of trademarks to domain names. Both come out of their own historical reasons, and matching them de jure may not be the answer, This does not mean that APIA denies legitimate infringements, but infringement should be proven not assumed.

Again we like to stress that actual infringement and not presumed infringement should be the guiding philosophy. Doing so, would be to accept that all domain names is equivalent of trademarks, which is not universally accepted. In this regards, the domain name administration should act merely as a listing service and not a trademark office. It should be the applicant's onus to protect his rights through the normal procedures.

As for automatic referral of disputes, this should be left to the challenger and applicant to decipher the relevant forum. Also an alternate dispute settlement center should not be imposed upon them. The IAHC proposal making all registrars bound to contract with their applicants to bind themselves to the WIPO mediation center, does not give them this choice. Also, since all registrars are bound by the decisions made here, this amounts to superseding national jurisdictions even if it claims not to do so (even if an applicant can get the national jurisdiction of the registrar involved to make a decision in its favor and the WIPO center or Administrative Challenge panel has a different decision, all registrars (except the one in which the local court decision is made) have to abide by the latter. In other words, the applicant may still have the name restored in the Registrar in question but it won't be registrar on the others, making his name useless. In actual fact therefore, the imposed mechanism does supersede national jurisdiction.)

We sincerely therefore stress further study and consultation before adopting measures to solve the problem of but a few.

Specifically in relation to domain names, APIA is of the opinion that issues raised in questions 23 and 24 should not be dealt without further inquiry and research on an international scale. Consultation with commercial, governmental and non-governmental bodies involved in dealing with trademark rights should also be undertaken. As far as possible, the link between domain names and trademarks should be strongly discouraged.

For the moment therefore, APIA is of the view that trademark rights should be protected through existing national laws and judicial processes. A globally devised dispute resolution system such as the ACP (IAHC proposal) cannot be yet be justified (only a dozen or so cases to date), not until its need is determined. To do as was prescribed in the IAHC today, would amount to prescribing trademark protection rights over and above what currently exists under national and international trademark laws, or in the reverse diminishing some other legitimate trademark rights. It also presumes that domain names are trademarks, a presumption that should be proven not assumed.

The Registrar should be a mere listing service, not an evaluation body. Information about actual usage and qualification for a particular TLD is of greater importance, to ensure efficiency and effective use, than trademark evaluations. Infringement should never be presumed.

26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

The DNS is an addressing system. As web browser technologies evolve, less and less emphasis will be placed on the use of domain names as identifiers of products or services/trademarks. The emergence of comprehensive search engines and greater use of directories may also provide for shared use of domain names. Further study into such technologies is strongly recommended.

28. Are there any other issues that should be addressed in this area?

Albeit protecting legitimate interests, minimize attention to trademarks, please.

APIA members include:

Founding Members:

Inet, Inc, Korea
Cisco Systems
NTT, Japan

Corporate Members:
Telekom Malaysia
Concert Communications Services Ltd
Asiapac.net
AUNet
VERIFI, Hong Kong
Singapore Exhibition Services

Organisational Member:
Internet Association of Japan

Individual Members:
Bernard Randall
Bill Woodcock
Sascha Zumbusch
Hiroki Inoue
Philip Klapakis
Vab Goel

Legal Counsel: Melvin Khoo, Harry Elias & Partners, Singapore

###

Number: 342

From: "David W. Maher" <dwmaher@ibm.net>
To: NTIADC40.NTIAHQ40(dns)
Date: 8/18/97 12:50pm
Subject: Comments on Internet Domain Names

TO: National Telecommunications and Information Administration
From: gTLD-MoU interim Policy Oversight Committee

Attached to this message, in Word format, are the comments of the gTLD-MoU
interim Policy Oversight Committee in the U.S. Department of Commerce
proceeding "Request for Comments on the Registration and Administration of
Internet Domain Names", Docket No. 970613137-7137-01.

Respectfully submitted,
gTLD-MoU interim Policy Oversight Committee
David W. Maher, Chair
<dwmaher@ibm.net>

Document available at: <http://www.gtld-mou.org>

Before the

U.S. DEPARTMENT OF COMMERCE National Telecommunications and Information Administration Washington, DC 20230

In the Matter of ) ) REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01 INTERNET DOMAIN NAMES )

Comments of gTLD-MoU interim Policy Oversight Committee

(created by the Generic Top Level Domain Memorandum of

Understanding (gTLD-MoU) signed in Geneva, Switzerland

on 1 May 1997)

gTLD-MoU Interim Policy Oversight Committee
David W. Maher, Chair c/o Sonnenschein Nath & Rosenthal
8000 Sears Tower
Chicago IL 60606

U.S. DEPARTMENT OF COMMERCE National Telecommunications and Information Administration Washington, DC 20230

In the Matter of ) ) REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01 INTERNET DOMAIN NAMES )

Comments of interim Policy Oversight Committee ("iPOC")

1. The interim Policy Oversight Committee ("iPOC") respectfully submits comments in this proceeding.

2. The interim Policy Oversight Committee ("iPOC") was created under the aegis of the Generic Top Level Domain Memorandum of Understanding ("gTLD-MoU") signed in Geneva, Switzerland on May 1, 1997 by the by the Internet Assigned Numbers Authority ("IANA"), the Internet Society ("ISOC"), and approximately 80 other signatory organizations. As of August 18, 1997, this has increased to approximately 160 signatory organizations.

3. iPOC is the successor to the International Ad Hoc Committee ("IAHC") which was formed in September, 1996 at the initiative of ISOC and at the request of IANA. IAHC issued a Final Report on February 4, 1997, which led to the creation and signing of the gTLD-MoU.

4. The charter of IAHC specified that it would "operate in the style of an Internet standards 'design team', formulating criteria and procedures but seeking review, modification and consensus from the rest of the Internet community" and stated further that "Internet standards are developed according to the principal of 'rough consensus' which means a strongly dominant sense of preference within the community that is seeking to achieve forward progress, in spite of differing opinions."

5. Between September, 1996, and May 1, 1997, IAHC operated open mailing lists, solicited and received hundreds of formal contributions, and participated in a number of existing mailing lists, thereby receiving thousands of comments from interested parties which were reviewed and considered in the preparation of the gTLD-MoU. Members of the IAHC and iPOC have provided continuing participation in those public mailing lists.

6. The gTLD-MoU proposes the creation, initially, of seven new top level domains and provides for an unlimited number of new registrars, all of which will share responsibility for registering domain names in all generic top level domains. The process of submission of applications to become registrars began on July 16, 1997. The registrars will sign a Memorandum of Understanding (the "CORE-MoU") and will form a Council of Registrars ("CORE") which will be a not-for-profit association under the laws of Switzerland.

7. Further information on the work of IAHC and iPOC, as well as the full text of the documents referred to above are available at the gTLD-MoU website <http://www.gtld-mou.org>.

A. Appropriate Principles

iPOC has answered a, b, c, d, e and f.

a. Competition in and expansion of the domain name registration system should be

encouraged. Conflicting domains, systems, and registries should not be permitted to jeopardize the interoperation of the Internet, however. The addressing scheme should not prevent any user from connecting to any other site.

c. These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.

d. The overall framework for accommodating competition should be open, robust, efficient, and fair.

e. The overall policy framework as well as name allocation and management mechanisms should promote prompt, fair, and efficient resolution of conflicts, including conflicts over proprietary rights.

f. A framework should be adopted as quickly as prudent consideration of these issues permits.

iPOC believes that the foregoing principles are appropriate and that they are substantially complete. They have been the subject of public debate for more than three and one half years and are in response to urgent Internet operations requirements. iPOC further believes that the program set forth in the gTLD-MoU executed on May 1, 1997 is the only concrete plan now in existence that embodies these principles. In order to foster these principles, iPOC urges the Government to support the gTLD-MoU program.

B. General/Organizational Framework Issues

iPOC has answered questions 1, 2, 3, 4, 5, 6, 7 and 8.

1. What are the advantages and disadvantages of current domain name registration systems?

Although the current system has the advantage that it has worked well enough to enable the Internet to grow at an exponential rate, the disadvantages have now become painfully apparent to nearly all users. The trademark conflicts, and especially the proliferation of "cyber-squatters", pirates who hold domain names for ransom, are becoming increasingly serious. There are two major disadvantages of the current generic domain name system. First, there are only three generic top level domains available to the general user, and second, the policies for resolving conflicts within the system have not worked well. The current ISO country code domain system has worked well, for the most part, and should be treated separately from the generic TLD system.

2. How might current domain name systems be improved?

iPOC believes that the program set forth in the gTLD-MoU offers the best hope for improvement of the generic TLD system. It increases the range of names that would be available, provides for registrar competition based on price and service (and not based on registrar monopoly control over specific TLDs), user domain name "portability", and takes a significant step towards resolving conflicts among those contending for domain names.

3. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

iPOC believes that the governance structure described in the gTLD-MoU, which is designed for expanded participation by industry and public groups, will be best able to administer the domain name system. iPOC further believes that this structure is fully consistent with the principles set forth in the Framework for Global Electronic Commerce released on July 1, 1997, and in particular, principle 1, that "governments should encourage industry self-regulation wherever appropriate and support the efforts of private sector organizations to develop mechanisms to facilitate the successful operation of the Internet."

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

The Internet standards process of the Internet Engineering Task Force ("IETF") serves as a model for the design of the domain name registration system and the program of the gTLD-MoU. The model will be different because of the operational and commercial nature of the gTLD-MoU structure and because the system must be designed de novo, rather than being based on a system with many years of existing operation. The specific principles which have been taken from the IETF standards process are open operations, layers of review and oversight, and community rough consensus. The design and responsibilities of the Council of Registrars ("CORE") is intended as a coordination activity between Registrars and provides for the gTLD-MoU policy oversight activities a single legal point of contact. The design of the Policy Oversight Committee ("POC") and the Policy Advisory Board ("PAB") reflect the higher level review mechanisms needed to provide necessary public policy oversight functions. The POC is somewhat akin to the IETF's Internet Engineering Steering Group, and the PAB is a formal embodiment of the general Internet community representing the large number of stakeholders in the Internet DNS.

iPOC believes that the limited roles proposed for the International Telecommunications Union ("ITU") and the World Intellectual Property Organization ("WIPO") in the overall structure contemplated by the gTLD-MoU will provide additional benefits of public sector input on the system without the disadvantages of putting the system entirely under the control of international governmental entities.

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the .com registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

iPOC believes that it is neither desirable nor (at this late date) possible to retire the existing gTLDs. Existing gTLDs represent a substantial base of users. They have made significant organizational and marketing investment in these names and it would cause serious operational and monetary difficulties for those users to give them up. There is no compelling technical or operational requirement which justifies taking this extreme action.

The continued existence of the ISO country code domains offers flexibility in the system that has no real disadvantages, assuming that there is coordination in management issues, especially in the area of domain name and trademark conflicts.

6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

iPOC believes that the technological basis of the program set forth in the gTLD-MoU is sound, and that the concept of shared data bases of multiple generic TLDs is practical and can be established in a short time, without disturbing the current root server system. The technical aspects of shared registration are well understood, and therefore the enhancement of the DNS registration model to include shared registration is entirely appropriate. National registries such as those in Great Britain have successfully implemented shared registries with a technology analogous to that which CORE is planning to use.

7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and coordinate?

The proposal set forth in the gTLD-Mou is designed to be scaleable. In particular, the gTLD-MoU sets forth a process which is designed to evolve as needs require and understanding permits.

8. How should the transition to any new systems be accomplished?

There must be an appropriate period of technical testing of any new system, but the administrative structure proposed by the gTLD-MoU is designed with enough flexibility to allow a smooth transition to a broader base of control. The DNS tolerates addition of new top-level domains with no impact on existing TLDs, as evidenced by the fact that approximately seven new two-letter ISO country code TLDs have been added every year for a number of years.

When a stable operational base for the new shared gTLDs is demonstrated, it will be appropriate to begin incorporation of the three existing gTLDs (.com, .net and .org).

C. Creation of New gTLDs

iPOC has answered questions 10, 11, 12 and 13

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

iPOC has taken into consideration, in its deliberations, assisted by public comments, the technical difficulties of scaling the existing system. iPOC believes that, from a technical and practical standpoint, the existing system could support 100-300 additional gTLDs, and that technological progress may enable the addition of significantly larger numbers in the future. There is a difference of opinion, among technical experts, as to the maximum number of additional TLDs that can be tolerated by the DNS root, without degrading system performance or reliability. Operational prudence therefore dictates a careful, incremental approach to the addition of TLDs. This caution is reflected in the approach specified by iPOC.

Also, there have been many comments, largely from entities concerned with the conflict of domain names and trademarks, urging that no new gTLDs or only a few be created. The seven new gTLDs proposed under the aegis of the gTLD-MoU is a compromise between these interests and the strong public demand for additional gTLDs.

11. Should additional gTLDs be created?

iPOC proposes the creation of seven new gTLDs as a first step. This permits development of an experiential base, with a minimum of risk.

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

See answer to question 10. However, it should also be mentioned that gTLDs, as currently used, represent a naming scheme much flatter than real-world identifiers. Therefore, in the longer term, it will probably be necessary to require addition of hundreds of gTLDs to provide a sufficiently rich name space as Internet exponential growth continues.

13. Are gTLD management issues separable from questions about ISO country code domains?

Yes, they are separable. iPOC believes that, in the long run, the stability of the Internet requires a unified management structure that will oversee the address space as well as the gTLD and ISO country code domains. This does not mean, however, that the ISO country code domains should be part of the registrar system established for gTLDs or that such domains should be organized in the same way as the gTLD domains. Each country should be free to develop its own policies in this area. iPOC is sympathetic to the position that intellectual property issues require consistent and uniformly applicable rules for both gTLD and ISO domains because of the global nature of the Internet.

D. Policies for Registries

iPOC has answered questions 15, 16, 17, 18 and 19

15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

iPOC proposes, in the gTLD-MoU, that all registrars share control over all seven new gTLDs and believes that, as a general principle, all registries should be shared. However, iPOC also believes that there may be room for some gTLDs (for example, a gTLD to be used exclusively by a governmental or non-governmental organization) which should be controlled exclusively by a single registrar under rules set by the body controlling the domains.

There are no technical limitations on its proposal for sharing and the computer technology required to achieve sharing is extremely stable and well understood

16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have? Who will determine these and how?

In the interests of fostering competition and encouraging lesser developed nations to have registrars, iPOC has set what it considers to be minimum financial and technical requirements for new registrars that are consistent with maintaining the stability of the Internet. The gTLD-MoU establishes a system of governance - a Policy Oversight Committee, a Policy Advisory Board and a Council of Registrars - that will share responsibility for the functioning of the system.

17. Are there technical limitations on the possible number of domain name registrars?

iPOC does not foresee any technical limitations to the number of registrars having access to shared data bases, and has not, therefore, set any limit on the number of registrars that may apply under the gTLD-MoU. Nevertheless, a specific effort is contemplated in the design of the shared databases to assure that no such technical limitations may appear.

18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

iPOC believes that increased competition among registrars will guarantee better service and lower prices for registrants. In particular, iPOC believes that the consumer is served best by having registrars compete solely on the basis of service and price, rather than on artificial control over a specific portion of the name space. In addition to better service and price, shared registration permits a single point of contact for registration activities, since a single registrar can handle requests for any gTLD. With exclusive control, it is necessary to go to different registrars for different gTLDs. As stated in answer to question 17, iPOC sees no technical issues at this time.

19. Should there be a limit on the number of different gTLDs a given registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

See answer to question 15 above.

E. Trademark Issues

iPOC has answered questions 21, 22, 23, 24, 25, 26, 27 and 28

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-à-vis domain names?

iPOC concluded at an early stage in its deliberations that it should do everything possible to insulate the domain name registration system from trademark conflicts, and that registrars should not have to make decisions on what trademark rights should or should not be protected. iPOC further concluded that it would be impossible to construct a domain name system that is free of trademark conflicts, unless a system were adopted that makes all domain names meaningless strings of letters and numbers. There appears to be no significant support for having meaningless strings aside from a few proponents who do not take account of the practical and historical support for a mnemonic based system. The gTLD-MoU program leaves questions of trademark rights to be resolved either by the courts or by alternative dispute resolution procedures (as described below in answer to question 24). The policy, as set forth in the gTLD-MoU, is as follows:

"a policy shall be implemented that a second-level domain name in any of the CORE-gTLDs which is identical or closely similar to an alphanumeric string that, for the purposes of this policy, is deemed to be internationally known, and for which demonstrable intellectual property rights exist, may be held or used only by, or with the authorization of, the owner of such demonstrable intellectual property rights. Appropriate consideration shall be given to possible use of such a second-level domain name by a third party that, for the purposes of this policy, is deemed to have sufficient rights."

iPOC believes that a requirement of preliminary review of applications for domain name registrations would put an impossible burden on the registrars and would ultimately cause the system to break down. The possible benefit from preliminary review would only be to protect a name applicant from investing in a name that will be challenged, but there is currently no assurance that any such benefit would accrue to the applicant. A mandatory review period would inconvenience all applicants, most of whom would derive no benefit. A voluntary review period affords to those who might benefit from the process the opportunity to balance the inconvenience against the uncertain chance of benefit.

23. Aside from a preliminary review process, how should trademark rights be protected on the Internet vis-à-vis domain names? What entity(ies), if any, should resolve disputes? Are national courts the only appropriate forum for such disputes? Specifically, is there a role for national/international governmental/nongovernmental organizations?

The gTLD-MoU proposes a novel plan to insulate registrars from trademark disputes by providing an expeditious and cost-effective alternative procedure for resolving trademark disputes. iPOC recognizes that it cannot and should not attempt to supplant the authority of national courts. Using the established alternative dispute resolution procedures of WIPO, the gTLD-MoU also proposes the establishment of Administrative Challenge Panels which will be able to resolve many disputes without recourse to litigation. These procedures are fully described in the Substantive Guidelines concerning Administrative Domain Name Challenge Panels published on the gTLD-MoU web site<http://www.gtld-mou.org>.

iPOC strongly supports full public access to the domain name registration databases for the purpose of limiting and reducing potential trademark conflicts. The proposed Memorandum of Understanding for the Council of Registrars (also available at the gTLD-MoU website) specifies that all registration databases must at all times make available all significant information needed for trademark searching.

iPOC has proposed that each applicant for second level domain names be required to furnish certain information about the applicant's purpose and reasons for requesting the name. iPOC believes that this information will be useful in combating trademark piracy and "cyber-squatting". The information will not be evaluated by the registrars but will be available if there is a dispute regarding the assignment of the name.

26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

iPOC does not believe that increasing the number of gLTDs and the number of registrars will, in the long run, have any effect on the number or cost of resolving trademark disputes. However, with a sufficiently large number of gTLDs, there is the possibility that trademark disputes could be reduced through the availability of enough categorization that conflicts over SLDs could be minimized.

27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

Various proposals have been made, including one in the Final Report of the International Ad Hoc Committee ("IAHC"), the predecessor of iPOC, for separate domain name registries for trademarks which would allow various owners of e.g. "ACME" to register in specific classes or with numerical designations to differentiate the owners. None of these proposals has received wide support, but the proposals of iPOC provide sufficient flexibility to accommodate such a solution if the demand arises.

Respectfully submitted,

gTLD-MoU interim Policy Oversight Committee

by David W. Maher, Chair

###

Number: 343

From: Ellen Rony <erony@marin.k12.ca.us>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 12:28am

Subject: NOI - Registration & Administration of Domain Names

Herewith, in the attached Microsoft Word 5.1 document--NOIrony.DOC--is my

submission for the U.S. Department of Commerce Inquiry into the

Registration and Administration of Domain Names.

Before the

U.S. DEPARTMENTOF COMMERCE National Telecommunications and InformationAdministration Washington, DC 20230

In the Matter of )

)

REGISTRATION AND ADMINISTRATION OF ) Docket No.970613137-7137-01

INTERNET DOMAIN NAMES )

ELLEN RONY

21 Juno Road

Tiburon, CA 94920

Director, Alexander Works

erony@marin.k12.ca.us

17 August 1997

Ellen Rony respectfullysubmits comments in this proceeding. I am a16-year veteran of the computer industry. I have been following domain name issues since January 1996 as co-author (withDr. Peter Rony) of a forthcoming book called The DomainName Handbook: High Stakes and Strategies in Cyberspace, to be published by R& D Books, an imprint of Miller-Freeman Publishing.

I am pleased that the Department ofCommerce is proactively soliciting comment from the public. We urgentlyneed a responsible system of governance that is truly representative andaccountable to replace NSI's monopoly franchise that will end on March 30, 1988.

A. Appropriate Principles

a. Competition in and expansion of the domain name registration systemshould be encouraged. Conflicting domains, systems, and registries shouldnot be permitted to jeopardize the interoperation of the Internet, however.The addressing scheme should not prevent any user from connecting to any other site.

tx6842a. First we shape our structures, and afterwards they shape us.

--Winston Churchill

While members of the Internet communitycannot seem to agree upon a framework for the Internet administration andallocation of international namespace, it is clear that few wish to havefragmentation within that realm.The current monopoly administration of five top level domains hasencouraged the development of experimental root server confederations.Errant registries that are not universally resolvable and potentialconflicts in TLD namespacedo not favor the smooth interoperation of the Internet.

SUGGESTED ACTION ITEM: A competitiveregistration system with universally shared root zone files shouldalleviate the pressure to create parallel TLD systems.

b. The private sector, with input from governments, should develop stable,consensus-based self-governing mechanisms for domain name registration andmanagement that adequately defines responsibilities and maintainsaccountability.

b.The Internet will become a critical infrastructure on which most ofsociety's communication and commerce flow. Already it has infused everyaspect of our society. As commercialization on the World Wide Webproliferates, collisions between competing financial, legal and operational interests are mounting. The stakes in this new,emerging technology are very high, especially for electronic commerce.

SUGGESTED ACTION ITEM: It doesn't seemprudent to rely on "consensus" for domain name administration andaccountability. There needs to be a system of checks and balances, andcertainly of accountability, to minimize the opportunity for abuse of control.

SUGGESTED ACTION ITEM: Internet namespaceseems most akin to bandwidth; the government should examine the model usedby the FCC.

c. Theseself-governance mechanisms should recognize the inherently global nature ofthe Internet and be able to evolve as necessary over time.

c. Broad consensus may bedifficult to achieve.

As the dimensions of the tree arenot always regulated by the size of the seed, so the consequences of thingsare not always proportionate to the appaent magnitude of those events thathave produced them.

-- Charles CalebColton (1780-1832)

SUGGESTED ACTION ITEM: The organizationsundertaking the administration of the Internet namespace need to be heldaccountable in some way. I see no reason not to promulgate some guidingprinciples that areresponsible, representative, accountable and equitable.

d. The overallframework for accommodating competition should be open, robust, efficient,and fair.

d.It is very distressing to read NSI's attempts to "brand .COM and assertionof ownership over the.COM database. In its IPO filing with the SEC (July3, 1997), NSI states, "The Company [NSI] has compiled a database ofinformation relating to customers in its registration business. While a portion of this database is available tothe public, the Company believes that it has ownership rights in thisdatabase and is seeking to protect such rights."

SUGGESTED ACTION ITEM: NSI should not beallowed to continue its monopoly franchise over the valuable .COM TLD.(andthe other TLDs)

SUGGESTED ACTION ITEM: Operational andadministrative arrangements need to begin to establish multiple competitorsto register domain names in the same TLD.

e. The overallpolicy framework as well as name allocation and management mechanismsshould promote prompt, fair, and efficient resolution of conflicts,including conflicts over proprietary rights.

e. The question remains, howto resolve disputes between different countries where both opponents have aright under their respective laws to claim use of an identical name incommerce.Only a first-come, first served system will work, but there must be somejustification required in the number of domains an individual, organizationor company can legitimately claim.

f. A frameworkshould be adopted as quickly as prudent consideration of these issuespermits.

f. Based on the growthrate of the past two years, the Inspector General says name registrations"will reacha plateau in mid-1999 of approximately 4 million, generating annualrevenue from registration fees of approximately $200 million." (posted athttp://zeus.bna.com/e-law/docs/nsfnsi.html#FNREF82001.)

SUGGESTED ACTION ITEM: Any framework thatis adopted must be scalable to a continuing and enormous influx of newparticipants. A many-layered approach, where responsibility is obfiscatedand registration fees are duplicated should be avoided.

SUGGESTED ACTION ITEM: It has to assurethat those who were born too late, who are developing new businesses as wedebate these issues are not in some way locked out of participation invirtual commerce.

B.General/Organizational Framework Issues

1. What are theadvantages and disadvantages of current domain name registration systems?

"Locus abauctoritate est infirmissimus." (The argument from authority is theweakest.)

-- ThomasAquinas

1.The number of registrations--exceeding 1 million--attests to the ease withwhich the process may be completed. This is a plus. On the other hand, themonopoly administration over registration of 5 international TLDs hassubjected Internet users to self-serving policies and misguided attempts to graft trademark assumptions onto anew communication medium. Unilateral promulgation and retroactiveapplication of domain dispute policies have had a material adverse affecton the Internet community.

SUGGESTED ACTION ITEM: The fact thatanyone can register many domains specifically for resale at substantiallyinflated fees is, in my belief, contrary to the open spirit of the Internet community. Procedures should be put in place to haltthis destructive practice.

2. How mightcurrent domain name systems be improved?

2. The current approach,funneling 88% of all domain registrants through .COM, is clearlyunsustainable

SUGGESTED ACTION ITEM: If 30 to 50 millionpeople are using the DNS in the manner of a directory, then efforts shouldbe made to facilitate this use. A WHOIS lookup that takes one directly to the relevant home page would bea useful tool. Also useful would be standardization among search engineprotocols, so that an average user can remember how to make a well- definedsearch.

SUGGESTED ACTION ITEM: Name reservationsshould be prohibited.

SUGGESTED ACTION ITEM: I have noobjections to the registration fee and believe that it should be payable assoon as the registry verifies that the desired name is available.

SUGGESTED ACTION ITEM: The practice bywhich domain hoarders will hold names up to 90 days without payment andthen reregister them as an expiration date approaches should not beallowed.

3. By what entity,entities, or types of entities should current domain name systems beadministered? What should the makeup of such an entity be?

If voluntaryguidelines worked, Moses would have come down from Mount Sinai with the Ten Guidelines.

-- Robert Orben,

3. SUGGESTED ACTION ITEM: Administrationof the US domain should be shared among a small group of registrars. Thereshould be ethical principles in common amongall the registries.

4. Are there decision-making processes that can serve as models fordeciding on domain name registration systems (e.g., network numbering plan,standard-setting processes, spectrum allocation)? Are there private/publicsector administered models or regimes that can be used for domain name registration (e.g., networknumbering plan, standard setting processes, or spectrum allocationprocesses)? What is the proper role of national or internationalgovernmental/non-governmental organizations, if any, in national and international domain name registration systems?

4. SUGGESTED ACTIONITEM: As these issues have global implications, I think that it would beappropriate to develop an international roundtable of representatives fromthe existing registries todiscuss transnational issues. Three-letter TLDs are presumed to beinternational, but the international perspective of commercial, legal andtechnical issues at stake have not been fully represented.The international community has been working since the Paris Convention of1883 to develop standards for registration and protection of intellectualproperty, so, unfortunately, I do not expect we will see similarcoordination in the realm of cyberspace any time soon.

SUGGESTED ACTION ITEM: All domain registrations should be country specific.Individual countries will have their own rules and policies regarding disputes, but each registry should issue domain names only to thosewho meet residency requirements or provide a business license to operatewithin the country.

5. Should generictop level domains (gTLDs), (e.g., .com), be retired from circulation?Should geographic or country codes (e.g., .US) be required? If so, what should happen to the .comregistry? Are gTLD management issues separable from questions aboutInternational Standards Organization (ISO) country code domains?

5. Defending theinternational .COM, .ORG, and .NET domains is a losing proposition.Law is based on geopolitical boundaries. Every corporation, business orindividual, has a country of origin, a headquarters, or a domicile. Thecurrentborderless approach to TLDs makes every domain holder potentially liable tothe statutes and regulations of every state, country, principality andkingdom, even when they conflict and contradict.

SUGGESTED ACTION ITEM: The current"international" TLDs should scrapped in favor of national TLDs.

SUGGESTED ACTION ITEM: The mapping of .COM to .COM.US (or .COM.xx forcompanies resident in other countries) would significantly alleviate thepressure and thorny issues we can expect regarding registration ofcompeting international trademarks as domain names.

SUGGESTED ACTION ITEM: Better identifiers are needed to help consumersidentify the source of goods or services, but I advocate a differentapproach. Currently, the .US TLD requires a geographic specificity thatmakes it unwieldy and unpopular. Instead,I recommend appending .US to the current "international"three-letter TLDs, as other countries do. This should at least alleviatethe potential for disputes between two entities in different countries,each with a legitimate claim to the same name. Although the largermultinational companies want to enjoy thebenefits of an international TLD, in the high velocity world of electroniccommerce, it is doubtful that appending .US to a name would hurt a businessevent a whit.

SUGGESTED ACTION ITEM: .COM, the target ofmost disputes, should be subdivided. The biggest source of litigation isbetween those with trademark registrations and domain registrants without.

SUGGESTED ACTION ITEM: Those with registered trademarks should get portedto a TLD that is based on the 42 international classes of goods andservices, simply called TM, preceded by the class number in which thetrademark is registered, xxTM.

Class 15, for example, includes "Musicalinstruments (other than talking machines and wireless apparatus."NAME.15TM.US would tell a web user a lot more about what would be at that address than what we have now, NAME.COM. It would befar easier to familiarize the public with 42 classes of goods and servicesthan have us plow through 200 similar monikers, as is now the case. Thisalso would free up some namespace for companies in a non-competing business identical names used indifferent market channels.

SUGGESTED ACTION ITEM: There also must be a repository for those who haveaccrued common law rights to a name. Perhaps COM.US, would be reservedfor those who had accrued common law rights to a name which are notcurrently acknowledged by NSI's present domain dispute policy.

6. Are there anytechnological solutions to current domain name registration issues? Arethere any issues concerning the relationship of registrars and gTLDs

6. A conclusion is justsimply the place where someone decided to stop thinking.

-- Marsala

SUGGESTED ACTION ITEM: Currently, one ofthe requirements of domain name registration is that the name be activatedwithin 90 days. This provision should be enforced.

SUGGESTED ACTION ITEM: The bright minds of the Internet community shoulddevelop an autoresponder to PING a web site a month after registration. Ifthere is no response, it should automatically PING again a month later. Atthe end of the third month, perhaps a PING should be made about four times in a 24-hour period on the lastday. If there is no response, the name should revert to the availablepool.

SUGGESTED ACTION ITEM: An online directoryis desperately needed to help Internet users differentiateamong the many confusingly similar names. The model provided by DISC.COMis a step in the right direction. An intermediate home page lists two verydifferent businesses using the identical DISC moniker. A user merelyclicks on the desired one. Search engines return too much extraneous information to help a person quicklydiscern from among 20 similarly named domains which one is the source ofthe particular goods or services being sought. A lookup databaseintegrated with the WHOIS records, that takes a user directly to a particular web site, might provide anefficient solution.

7. How can weensure the scalability of the domain name system name and address spaces aswell as ensure that root servers continue to interoperate and coordinate?

7. Certaintygenerally is illusion and repose is not the destiny of man.

--U. S. Supreme CourtJustice Oliver Wendell Holmes, Jr. (1841-1935)

8. How should thetransition to any new systems be accomplished?

Change is not madewithout inconvenience, even from worse to better.

--Samuel Johnson (1709 -1784)

8.SUGGESTED ACTION ITEM: The brilliant technical minds not occupied with theYear 2000 problem perhaps could create a mapping program that would switch.COM to .COM.US or even to .15TM.US transparently. I t would be theelectronic equivalent of call forwarding -- domain forwarding -- over the course of a six month period. It wouldstart on the assumption that all .COM domains originate in the .US and secondarily transition out the smaller group of domains that are headquarteredelsewhere.

I do notbelieve that a transition from an international TLD to one that is country-based will extract a huge "social cost". Next month, the San FranciscoPeninsula will change over from the 415 area code to 650 to accommodategrowth in this state. The change affects 2.5 million numbers. A newprefix means new stationary, new business cards, reprogrammed cell phones, fax machinesand modems for business. There doesn't appear to be any hue and cry aboutthe social cost of this changeover, and new area codes are popping up allover the country as the existing reserve of numbers runs out. It should be easier to update URLs on line thanto change a phone number with all its ripple effects. Some websites arehuge, but now there are HTML generators that can perform global search andreplace.

C. Creation of NewgTLDs

10. Are theretechnical, practical, and/or policy considerations that constrain the totalnumber of different gTLDs that can be created?

10. Neither lawyers norgovernments can make ten pounds of names fit into a one-pound bag.

-- John Gilmore, ElectronicFrontier Foundation

The name shortage is fueled by a speculative fever and a misguided policythat allows anyone with access to the InterNIC registration forms andperhaps an acronym generator to grab as many domain names as they want. IfgTLDs are opened to public registration, we will just ratchet the speculative, fever up to a higher level.

SUGGESTED ACTION ITEM: Halt the gTLDmovement immediately. It is unnecessary to create new international TLDs,and such activities will create a Pandora's Box of cross-borderproblems.

11. Shouldadditional gTLDs be created?

Businesses, organizations, andindividuals should absolutely not have the right to name and "own" anyTLDs. A TLD should be seen as a public resource and not as individualproperty.

11. SUGGESTED ACTION ITEM: I believe weonly need one new gTLD, which I shall call.NDX. This would be the beginnings of a global lookup, akin to the WHOISdirectory, that would take people directly to the domain name of choice.When a domain name is registered, a copy of the the record couldautomatically be included in the NDX directory.

SUGGESTED ACTION ITEM: New TLDs that are not country -based will only exacerbate the "root" of the problems alreadymanifested in .COM. All TLDs should be based on the ISO 3166 country codesexcept for .INT and .ARPA, which should be grandfathered in., along withthe new one, NDX.

12. Are theretechnical, business, and/or policy issues about guaranteeing thescalability of the name space associated with increasing the number ofgTLDs?

SUGGESTED ACTION ITEM: Thename protocols should drop down one level across the board, with anexpansion of the SLD regime.

SUGGESTED ACTION ITEM: New SLDs shouldinclude the major areas of concern to human society, such as (in noparticular order):finance, politics, sex, religion, commerce/no trademark, commerce/withtrademark, non-profit organizations, individuals (personal), media, arts,education, government, health, and military: FIN, POL, SEX, REL, COM,xxTM, NPO, IND, MED, ARS, EDU, GOV, HEA, and MIL.

The advantage to these SLDs is tht theywill help Internet users better identify the particular domain holder. Myemail address is long by average standards, but it lets people know: 1) who I am;2) where I am located; and 3) generically, what I do. I think itaccomplishes a lot in just 21 characters: erony@marin.k12.ca.us.

SUGGESTED ACTION ITEM: Expansion of theSLDs followed by the country code would be the appropriate way to handlethe massive numbers of domain applications arriving daily.

SUGGESTED ACTION ITEM: Conflicts would bereduced if all the non-commercial entites that are now funneled under COM could be more appropriately categorizied. Individual web sites that areinformational, not commercial, certainly don't belong there. Neither do theNPOs who gravitate to .COM because their desired .ORG name is taken. Thiswould also help the consumer, who is hit with a staggering assortment of very similar names.

SUGGESTED ACTION ITEM: Let the registrantchoose which category best applies to the specific business/organization..A company wanting to caste a net over more than one TLD would have toprovide a very good reason for doing so. That does not preclude a company with anon-profit foundation from registering in NPO, as well.

The focus should be more on how the DNS isactually used, not on ways to avoid pigeon-holing businesses. Because ofthe latter perspective, we now have a glut of registered but inoperative domains, ascompanies busily scoop up every trademark, brand, division, committee,slogan, variation, and plural they can think of associated with theircommercial operation.

Recently, I tried to figure out which of 25 domain variations of the WarnerBrothers name would take me to its web site. Such confusion is anoutgrowth of companies fears that they will lose control over their name incyberspace. But the end result isa bewildering array of choices that more likely to encourage a Internetuser to look elsewhere.

13. Are gTLDmanagement issues separable from questions about ISO country code domains?

13. If the US does nothingand new gTLDs appear, what is to stop other countries from issuingadditional TLD? Is this direction truly going to serve the globalcommunity in the succeeding decades?

SUGGESTED ACTION ITEM: This question ismoot if the only gTLDs involving the US are .US, .INT, .ARPA and .NDX.

14. Are there anyother issues that should be addressed in this area?

D. Policies forRegistries

15. Should a gTLDregistrar have exclusive control over aparticular gTLD? Are there any technical limitations on using sharedregistries for some or all gTLDs? Can exclusive and non-exclusive gTLDscoexist?

SUGGESTED ACTION ITEM: Since this is a global medium, efforts should bedirected toward harmonization and integration of registration policies. Theestablishment of exclusive registries would move Internet administrationfurther away from this important goal. Most assuredly, NSI should not be allowed tocontinue its monopoly control over .COM and the other "international"TLDs.

SUGGESTED ACTION ITEM: An exclusiveregistry for .INT, .ARPA and .NDX would be acceptable.

16. Should there bethreshold requirements for domain name registrars, and whatresponsibilities should such registrars have? Who will determine these andhow?

16. SUGGESTED ACTION ITEM: NSIshould be required to immediately open an escrow account for all pro ratamonies on registrations that extend beyond September 1988. Domain registrants should be allowed to chose among a variety of registrarsall servicing the same .US TLD. Those who select another registrar shouldincur NO financial penalties and should be allowed continuous use of the domain name as it transfers under a differentadministration.

SUGGESTED ACTION ITEM: There absolutelyshould be threshold requirements. Those who register a name for Internetuse have much at risk when they begin to develop content on a web site andemail contacts under the name. it is in the public interest to assure continuity and consistency ofservice.

SUGGESTED ACTION ITEM: Again, I wouldsuggest that the model of allocating bandwidth and call letters to radiostations be explored to see ways it can be applied to the virtualrealm.

17. Are theretechnical limitations on the possible number of domain name registrars?

17. I have read that there arenot technical limitations, but there certainly are practical ones.Citizens of modern society are already overwhelmed by a plethora ofchoices. Although I believe that the registry operations should be shared, too manyregistrars would slow down or complicate the process.

SUGGESTED ACTION ITEM: Multiple registrars will need equal access to adynamically updated, shared database to assure that identical domain namesare not granted in separate locations simultaneously (the airline ticketdatabase comes to mind).

SUGGESTED ACTION ITEM: It is essential notto permit monopoly control over any gTLD.

SUGGESTED ACTION ITEM: I believe it would serve the public interest tohave, perhaps, a dozen geographically distributed registrars. The numberwould assure competition without make the choice burdensome upon theaverage netizen.

18. Are theretechnical, business and/or policy issues about the name space raised byincreasing the number of domain name registrars?

18. While those at thetechnical end want a piece of this lucrative pie, perhaps 95% of the domainname holders have minimal interaction with this part of the process. Theysimply want to establish a name and get on with their business without interference. Currently there areabout 180 registries worldwide, but they operate with different levels ofexpertise and different policy requirements. Since the Internet is a global medium with every website potentiallyaccessible to anyone with a ISP connection, I don't think a business shouldfind it necessary to register domain names in multiple countries and inmultiple TLDs.

SUGGESTED ACTION ITEM: TI think that nouseful purpose is served by establishing a plethora of registrars. Thegreater the number, the greater the likelihood of discord and confusion. .

19. Should there bea limit on the number of different gTLDs a given registrar can administer?Does this depend on whether the registrar has exclusive or non-exclusiverights to the gTLD?

19. This NOI earlieraddressed the issue of whether a registrar should have exclusive rights toa gTLD.

SUGGESTED ACTION ITEM: Since I feel that the domain name system should be country-based, I can only addressthe .US TLD, which should be shared by various registries. (Perhaps INT,ARPA and .NDX fall into a different class and would be administered by asingle registry.)

SUGGESTED ACTION ITEM: If the gTLDs areall country-based, then this questions is perhaps moot. The US cannotdetermine for other countries how they should handle their registries. Ifa minimal number ofadditional gTLDs are established, there should not be a limit on the numberthat a given registrar can administer.

SUGGESTED ACTION ITEM: There should be a limit on the number of multiple SLDs that are availableunder which individuals and organizations can register domain names. Forexample NAME.NPO.US or NAME.12TM.US are two different SLDs.

E. TrademarkIssues

21. What trademarkrights (e.g., registered trademarks, common law trademarks, geographicindications, etc.), if any, should be protected on the Internet vis-a-visdomain names?

21. If the was a magicway for us to take care of trademarks, we would. But the trademark issue is just a mess.

-- Jon Postel

By removing the costs of distribution andthe barriers of time and distance, the Internet is changing the rules ofcommerce and the structure of our social order.There are ways people use domain names other than expressing trademarks.

SUGGESTED ACTION ITEM: The attempt tograft trademark ownership onto domain registration is unsustainable. Suchefforts should cease.

SUGGESTED ACTION ITEM: NSI hasestablished itself as an arbiter of legal rights in a misguided attempt toobtain a balanced position between the competing demands of trademarkowners and domain name registrants.While NSI's dispute policy may quell some legitimate concerns of trademarkowners (e.g., infringing use, extortion and dilution of a mark), it placesthe registration authority in a judicial role.Forcing the suspension of a domain name because a registered trademarkholder asserts a baseless infringement claim is in contradiction toestablished trademark law. Such arbiter efforts should ceaseimmediately.

SUGGESTED ACTION ITEM: Trademarks used onthe Internet should be protected by the body of trademark law as regardsinfringement and dilution. The mere registration of a domain name, WITHOUTMORE, should not be held by either NSI or the courts as justification forpreliminary relief under a claim of jtrademark infringement.

SUGGESTED ACTION ITEM: Personal domains, for example, or names that are sodescriptive that they aren't eligible for a trademark, or names that havenot yet achieved secondary meaning. All of these may be perfectlylegitimate domain names even though they are identical to a company trademark.

22. Should some process of preliminary review of an application for registration of a domain namebe required, before allocation, to determine if it conflicts with atrademark, a trade name, a geographic indication, etc.? If so, whatstandards should be used? Who should conduct the preliminary review? If a conflict is found, what should be done, e.g.,domain name applicant and/or trademark owner notified of the conflict?Automatic referral to dispute settlement?

22. The problem now is thatso many businesses are moving operations onto the web and just about every name willfind some kind ofconflict with another. Theonly appropriate standards according to trademark law are in the realm ofactual confusion as to the source of goods and services. That means therehas to be content on the web site which would be the source of confusionor, in the case of a famous mark,of dilution. A domain name alone is not enough.

SUGGESTED ACTION ITEM: A preliminaryreview is unappropriate, except by the applicant. Any names that areunavailable because of federal law should be posted, along with a link to the applicable statute (e.g. "Olympic","NASA").

SUGGESTED ACTION ITEM: An online trademarkdatabase would certainly help registrants be aware of trademark pitfallswaiting in the wings for a particular name.

23. Aside from a preliminary review process, how should trademark rights beprotected on the Internet vis-a-vis domain names? What entity(ies), if any,should resolve disputes? Are national courts the only appropriate forum forsuch disputes? Specifically, is there a role for national/international governmental/nongovernmentalorganizations?

23. Our system ofjurisprudence is designed to accommodate conflicting claims, defineproperty rights, establish rules to guide conduct and, of course, to resolve disputes.Cross-border disputes are tricky because different countries applydifferent restrictions, rules and limitations to the registrations.

SUGGESTED ACTION ITEM: Thetrademark-cum-domain issues would find huge relief if, to the USC ISIgeographical hierarchy, we would add the following:

.COM.US, .ORG.US , .NET.US, .EDU.US,.GOV.US, .MIL.US and so forth.

SUGGESTED ACTION ITEM: U.S. organizationswould be required to register for the .US domain. Corporations in theinternational marketplace would register according to the country whereintheir headquarters are located.

SUGGESTED ACTION ITEM:An independenttribunal should be established by the WIPO to adjudicate internationaldisputes.

The focus of most of the domain friction is upon whether a domain name,this new identifier-hybrid, is property of just an address. A domain namehas been likened to an alias, address, brand, directory, handle, trademark,telephone number, and even a cow.

SUGGESTED ACTION ITEM: It seems time forthe USPTO to recognize a new category of intellectual property: domainnames.As distinct from patents, trademarks and copyright as those other forms ofintellectual property are from each other.

SUGGESTED ACTION ITEM: An individual should not be required to trademarkhis or her surname to protect a similarly named domain from a trademarkinfringement claim (see TY.COM v. Ty, Inc.).

SUGGESTED ACTION ITEM: A company with earlier common law rights in a nameshould not have to yield the coveted domain name to a registered trademarkowner.

SUGGESTED ACTION ITEM: Wouldn't "unbundling" COM into somesociety-recognized categories as SLDs relieve the congestion and, mostimportant, help out the average Internet user?

24. How canconflicts over trademarks best be prevented? What information resources(e.g. databases of registered domain names, registered trademarks, trade names)could help reduce potential conflicts? If there should be a database(s),who should create the database(s)? How should such a database(s) be used?

24.This question treads into murky territory. Only 51 nations have ratifiedthe Nice Agreement as of January 1, 1997. Other countries apply differentclassification standards to their trademarks. And country-to-country, therequirements for granting a trademark differ, as well.

SUGGESTED ACTION ITEM: Internet use is evolving at a blistering pace. Atrademark alone should give no automatic right to the name as an Internet address or to the appropriation of the name from another party to whom it isalready registered. Nor should a domain name confer trademark rights in the name. Rightnow, anyone who is a domain holder but does not possess a registeredtrademark for the name is vulnerable to a challenge FROM ANYWHERE ON THEPLANET.

SUGGESTED ACTION ITEM: There absolutely needs a directory of registeredtrademarks accessible to the public without the need to hire a trademarkattorney and spend $$$$$$ simply to register a domain name.

SUGGESTED ACTION ITEM: Domain registrants deserve the opportunity to feelsecure that their time and investment in developing a web site is not atcontinual risk merely because they chose not to trademark a name.Expensive and protracted litigation to lodge a response to a name challenge imposes a significant burden upon alegitimate domain registrant and can devastate an on-line business.

SUGGESTED ACTION ITEM: Assuming thecontent at a site is not of an infringing or diluting nature, a registrant should have publiclyavailable the resources (e.g. trademark directory, WHOIS of registrants)that can provide peace of mind in this regard.

SUGGESTED ACTION ITEM: A [erson who uses his or her given name as a domainshould be exempt from competition with a registered trademark, at leastbased on who is the senior user of the contested name.

25. Should domain name applicants be required to demonstrate that they havea basis for requesting a particular domain name? If so, what informationshould be supplied? Who should evaluate the information? On the basis ofwhat criteria?

25. SUGGESTED ACTION ITEM: Aregistrant seeking more than one domain name should provide some evidenceof a legitimate need for each name that is desired.

Acceptable evidence could be defined quitebroadly. A separate company division, brand, product line, or geographicalregion would all be acceptable, if undesirable, reasons for assertingrights to another domain name.

SUGGESTED ACTION ITEM: Not acceptable would be those who wish to hold aname for resale. Thus, a business name such as "For Sale by Owmer.,""Domains for Sale" or my personal favorite, "If You Want This Domain,Please Contact Me," would not be allowed.Some may argue that an after market is a legitimate industry. I think itis just an unproductive offshoot that exacerbates the frustrations of manymembers of the Internet community.

26. How would thenumber of different gTLDs and the number of registrars affect the numberand cost of resolving trademark disputes?

26.If new gTLDs are established, many businesses will feel obligated toregister their trademarked name in every one that they can, by any stretchof the imagination, qualify for. This activity, already evidenced in .COM,seems to be driven by some sort offear that if the trademark owner doesn't throw a safety net around thename, another user will not only claim it, but will impose infringing usesupon it. A directory that facilitates user lookup would minimize the opportunityfor confusion based solely on the domain name.

27. Where thereare valid, but conflicting trademark rights for a single domain name, arethere any technological solutions? How would the number of different gTLDs and the number of registrars affectthe number and cost of resolving trademark disputes?

I don't know how trademarkscan be "conflicting" when they are issued for different classes of goodsand services. To have "valid, but conflicting trademark rights," wouldmean that two registered trademark owners of an identical mark would feel they were each entitled to a particular domain name.That doesn't seem the proper interpretation of trademark rights unless itis determined that domain names = trademarks. They don't. Trademarkrights include the right to protection against unauthorized use whichcauses confusion by consumers as to the source of goods or services, or, in the case of famous marks, protectionagainst a "whittling away" of the value of the mark. The same trademarkapplied to different classes of goods will not have a "conflicting" use.Many individual domain holders with a legitimate nexus to a name havefought battles to keep Network Solution Inc.'s hands off the domain while atrademark challenge was pending.

SUGGESTED ACTION ITEM: One field of thedomain application could be for a date of first use of the domain name name(without the .COM.US suffix) in commerce. For many registrants, that firstuse might likely be the day the web site is activated. This would, at least, put others on noticeof priority rights in a name but would not assure priority rights in adomain name.

28. Are there anyother issues that should be addressed in this area?

28. Yes.

SUGGESTED ACTION ITEM: Take decisive action immediately, before the March 31, 1998cutoff of the Cooperative Agreement National ScenceFoundation and Network Solutions, Inc. The U.S. government (NSF) hasallowed this mess to build, now it should resolve it without further delay.The online community wants to know that the transition from a monopolyfranchise to a shared database with competitive registries will occur ina timely and orderly manner. Theworst result would be that confusion would reign and by default, thecurrent registrar would be able to continue its flawed control over thefive TLDs.

[END]

###

Number: 344

From: Karl Denninger <karl@mcs.net>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 1:05pm

Subject: NOI RESPONSE - Docket number 970613137-7137-01

Please find enclosed our NOI response in Microsoft Word format.

Karl Denninger (karl@MCS.Net)| MCSNet - Serving Chicagoland and Wisconsin

http://www.mcs.net/~karl | T1's from $600 monthly to FULL DS-3 Service

| NEW! K56Flex modem support is now available

Voice: [+1 312 803-MCS1 x219]| 56kbps DIGITAL ISDN DOV on analog lines!

Fax: [+1 312 803-4929] | 2 FULL DS-3 Internet links; 400Mbps B/W Internal

CC: NTIADC40.SMTP40("karl@mcs.net")

To: Department of Commerce, NTIA REQUEST

Docket number 970613137-7137-01

Regarding: DNS Governance Issues

From: Karl Denninger, President

Macro Computer Solutions, Inc.

Two Prudential Plaza, Suite 2625

Chicago, IL 60601

312-803-6271 x219 (Voice)

312-803-4929 (Fax)

karl@mcs.net (email)

Gentlemen,

Please enter the following commentary as an official response from our organization to the above-referenced document and comment request. We look forward to a constructive dialogue, with the Federal Government's appropriate intervention and oversight, in resolving the significant policy questions which have arising in the management of the top-level domain namespace.

Sincerely,

Karl S. Denninger

President

Background

Karl Denninger, MCSNet's President, has been actively involved in the electronic and online world for nearly 15 years. Author of the first electronic networked conferencing system, AKCS, in 1981, Mr. Denninger has been at the forefront of electronic commerce and online issues during his entire professional career. Mr. Denninger currently owns a majority interest in Macro Computer Solutions, Inc., a privately-held Internet Service Provider based in Chicago with a service area encompassing northeast Illinois and southeast Wisconsin.

Macro Computer Solutions, Inc. (dba MCSNet) is one of the Chicago area's largest Internet Service Providers. MCSNet spearheaded the formation of eDNS, and alternative top-level domain name registry system which was originally constituted in March of 1997 and reorganized in May of 1997.

The eDNS system is predicated on the basic principle of "first come, first served", recognizes the right to develop, market, and own brand names in cyberspace just as in the "old fashioned" paper world. The eDNS system also recognizes that both government and industry have a legitimate interest in protecting competition and preventing the development of monopoly forces in the registration of domain names for the purpose of protecting consumers and promoting diversity.

As a consequence of this focus and background, our comments will necessarily be directed towards the purpose of protecting these principles and ensuring that open, competitive markets are able to form and flourish.

Current Situational Facts

Today we have what amounts to a monopoly in domain registrations within the "gTLD" arena. This name, given by the IAHC and IANA, is actually a misnomer. All TLDs are necessarily at the same level in the "authority" tree, even though the IANA and IAHC would like to pretend otherwise. Different TLDs have different policies and procedures for both registration and operational criteria, including requirements on location of the organization, fees, dispute resolution, and possibly line-of-business or even affiliation requirements. For example, .MIL is limited to organs of the US Department of Defense. .GOV is limited to organs of the United States Federal Government. .ARPA is, in fact, the most special-case it is limited to the definition of reverse-mapping records for the resolution of host names from ip numbers.

It is misleading and in fact fraudulent to suggest that we can "categorize" TLDs into some kind of artificial set of groups. In fact, the assignment of ISO country codes (whether ISO2 or ISO3) as TLDs is in fact a distinction which is based on nothing more than historical precedent just as is .GOV, .MIL, and .ARPA.

The National Science Foundation (NSF) originally designated certain registries to operate different TLD zones, based on either national sovereignty or administrative convenience. NSI, along with AT&T and General Atomics, were given a grant to operate .MIL, .GOV, .COM, .NET, .ORG and .EDU several years ago. While there were several complaints about the speed of processing and general performance over the years, no severe controversy arose until September 1995 when NSI unilaterally changed the terms of not only new registrations, but renewals of existing domains through the imposition of user fees.

There were and remain several serious complaints related to these changes. Some of the most pressing were:

1) This change was undertaken without any comment period for the community.

1) The change was made retroactively, and amounted to a unilaterial modification of an agreement (some would even say "contract") by one party.

1) The change included a modification of the dispute resolution policy (which was previously "take it to a judge"). In many cases this was far more significant than the demand for monies.

The community responded with a number of requests and suggestions, all of which were ignored. The most obvious fix for a firm taking such action would be competition but the IANA refused to issue additional TLDs and allow firms to attempt to develop those brands under its auspices.

The result was the birth of a new industry, the alternative DNS registry.

Since that time, the IANA has effectively boycotted these alternative registries. Further, a non-profit organization, ISC.ORG, delivers the DNS software, called "BIND", with a pre-loaded file which contains only the IANA/NSI root nameservers. No mention is even made of the alternative possibilities for this file, nor where to find them, when a user downloads and installs this software.

In addition, computer manufacturers and other software vendors package this software with their operating systems, including the cache file. The consequence of this is a restraint of trade, as in many if not all cases network operators are completely unaware that choices even exist. As a result, the monopoly control over the TLD namespace is perpetuated.

MCSNet and eDNS feel that this situation must be addressed.

Response to NOI Questions

A. Appropriate principles

MCSNet and eDNS feel that the principles espoused in this section, if implemented fairly and appropriately, will encourage competition and permit the entry of free competitive forces into the DNS registry marketplace. So long as principle (A.1) is not permitted to result in de-facto restraint of trade due to false technical constraints which have been claimed (but in fact are not real) the guiding principles set forth are appropriate.

A. General/Organization Framework Issues

1. Advantages and disadvantages of current system:

a. The current system is an effective monopoly. By exerting control over the number of TLDs which are issued, and who controls them, the IANA and NSF have effectively precluded effective competition.

a. The advantage to any monopoly is that the costs, pricing models, and operation are pre-determined for all registrants under any given monopoly system.

a. The primary disadvantage is that the free market is not able to address inequities in pricing or dispute resolution by offering alternatives to the consumer.

a. A secondary, but possibly more-important disadvantage, is that the development of natural language forms of site naming has been completely prevented. For example, to find American Airlines today on the web, you must type http://www.americanair.com. Instead, should you not be able to type http://american.airlines? Today, the second form will not function. It appears obvious that allowing the virtually-unlimited advancement of top-level domains will enable the use of natural language within the Internet a huge boost to usability by consumers.

1. How might current systems be improved:

By opening the current system to free and open competition, subject to the normal "first use" restrictions which exist for any other line of business in the United States or elsewhere, and permitting brand development in TLDs to occur, the problems described in point (1) will be effectively addressed.

1. Who should administer the domain system?

The TLD domain system (ie: the "root") needs to be administered by an entity which is free of coercion and also accountable to both free market economic realities and the legal protections which exist for firms both in the US and abroad (ie: Anti-trust laws). The current IAHC proposal does not meet this criteria, as the moving of this function outside of the US to Switzerland effectively negates US anti-trust legal protections.

It is our contention that it is essential that all legal remedies available to registrants, and in particular firms that wish to enter this marketplace, remain fully available. It is particularly important that any malfeasance, deliberate obstruction of business process, or other actions which would constitute either a civil or criminal violation of a firm's or person's rights not be shielded from the jurisdiction of a court by either contract or selection of jurisdiction. The current IAHC proposal is particularly onerous in that regard, as their application procedure contains a clause that the applicant agrees that the IAHC may not be pursued legally even if the IAHC acts in a capricious, malicious, or arbitrary manner. This sort of unconscionable language would not be allowed to stand in a US-based agreement but it may be perfectly acceptable under Swiss law.

1. Are existing decision-making processes available?

We do not believe that there is a current model that applies to this problem. TLD assignment is very different than spectrum or IP numbering plans, in that there is an effectively unlimited namespace available. Even if we only consider five-character and less TLDs (in fact there is no technical reason to limit TLDs to less than 32 characters) we still have a namespace of more than 60 Million potential names or one for every 4 people in the United States. Technical limitations do place an upper boundary on the number of TLDs which can be reasonably supported, but that number is in the tens of thousands and the technical requirements of providing effective, commercial service will prevent this space from becoming exhausted in the forseeable future.

MCSNet believes that national and international governmental bodies should act only to prevent the monopolization of the namespace and its control. The free market has been proven over time to be the only "fair" means to allocate resources when there is no natural scarcity of a given resource. Since the potential namespace is so enormous (many quadrillion potential TLDs), and reasonable technical requirements act in the requirement of provisioning reliable service, there is no operative reason for governments to begin "rationing" TLD resources.

1. Should generic TLDs be retired?

It is MCSNet's contention that not only should they not be retired, but they should be opened to free competition. The disruption to the marketplace should generic TLDs be retired would be enormous in fact, beyond enormous to the potential disruption of all Internet-related commerce. It is simply not feasible today to force over 1 million registrants to change their addresses, stationary, and other business communications. Forcing this change now would effectively disconnect most businesses on the Internet for an indefinite amount of time.

Further, the disruption to existing ISPs cannot be underestimated. MCSNet, for example, currently houses over 1,500 customer domains. Each and every one of these zones would have to be individually changed should such a plan go into force. This would be repeated 1,000-fold at ISPs and other providers around the globe. The manpower requirements to do this would be astronomical and run into the tens of millions of dollars or more.

The gTLD system must recognize and provide for the ISO country code namespaces in both the 2 and three-character formats. Both of these, for any given country, should be delegated to an arm of the respective governmental body, and administration left to that government. For example, management of .US and .USA should immediately be given to an arm of the US Federal Government to manage as they see fit. In the United States, this might be the National Science Foundation, or perhaps the Department of Commerce.

1. Technological Solutions to current issues?

The issues facing us today are not technological they are political in nature. There is no bar to even 100,000 TLDs being placed in the root servers other than Mr. Postel's refusal to do so. There is no bar to the roots synchronizing their zones again, other than Mr. Postel's and the IANA's refusal to permit it.

The problems facing this area of industry are political and must be handled as such.

1. How can we ensure the scalability of the DNS system?

Appropriate technical standards for TLD "owners" or "delegates" make this requirement moot. The DNS system is a remarkable distributed database in fact, it is one of the most robust distributed systems ever invented. If the requirement to operate a TLD were, say, to have at least two DS-1 (1.5 Mbps) connections to disparate networks and two distinct nameservers for that TLD, you would create a "pool" of potential registries of approximately 1,500 US organizations, and perhaps another 2,000 world-wide. This would bring the total pool to approximately 3,500 organizations. If we then limit, to prevent anti-competitive activity, the number of TLDs that an organization can control to 10, we have a maximum of 35,000 TLDs in service.

Assuming that 50% of the organizations which could meet the requirements would choose to enter this market, we have under 20,000 TLDs in total. This is less than 1/5th of the existing capacity of the TLD DNS root servers by the most conservative estimate. This is also less than 1/50th of their current operational load (the roots currently have COM on them, which consists of 1 million active domains).

Clearly, there is no problem with scalability of the DNS system even in a fully free-market model.

IP numbering is an entirely different problem and should not be tied to nor attempted to be addressed by these same processes.

1. How should transition to these new systems be accomplished?

Clearly, someone needs to "take the bull by the horns". It is clear that at the present time the IANA has no intention of actually permitting free-market forces to come to bear. Since the NSF originated this system, it appears appropriate for it to let, by contract, an agreement to an agency to manage the TLD delegation system, recognizing all prior use claims consistent with the protection of free commerce and prevention of monopoly activity. This could be done immediately and with no disruption to the existing infrastructure.

1. Are there other issues that should be addressed?

MCSNet believes that a formal inquiry by the Department of Justice into the actions of the IANA is warranted at the present time to determine whether violations of US law, particularly the Sherman and Clayton acts, have occurred. We support and request that such an inquiry be conducted with respect to Mr. Postel's refusal to open the DNS system to free and full competition upon the origination of user fees and policy changes by NSI. We further request that all appropriate actions consistent with the laws of the United States be taken.

A. Creation of new gTLDs

1. Are there practical an/or policy considerations that limit the number of TLDs?

Yes, but they are very large. Under point (7) above, we have addressed some of these concerns and shown that within the constraints of reasonable technological requirements, the natural outcome in a free market system will fall well within current capabilities.

Since the computer capacity of servers and systems is expected to continue to expand, we do not believe that any problems with scalability will arise in the future.

1. Should additional TLDs be created?

Yes. Only through the free market can the inequities of the current system be addressed. The TLD namespace should be immediately opened to free and full competition, consistent with protection against monopoly interests grabbing large portions of the namespace.

1. Are there technical, business and/or policy issues related to scalability of the namespace?

Not within the guidelines set forth in this response. The current COM zone is running on the roots and has more than 1 million entries in it. This is constructive proof that contrary to some assertions, the system will not melt down even with similar numbers of TLDs.

1. Are gTLD management issues separable from questions about ISO country code domains?

Within the 2 and 3 character ISO namespace, it is clear that the ISO country code delegations must take precedence. EDNS, for example, has reserved all two-character combinations for this reason. However, in the 4+ character namespace, there is no reason for interference to occur, and in fact even in the 3 character space, the ISO requirement is "sparse" compared to the total number of combinations available.

1. Are there any other issues that should be addressed?

We believe this has been relatively exhaustive of the salient issues related to this section and applaud the thorough treatment you have given this issue.

A. Policies for Registries

1. Should a gTLD registrar have exclusive control over a particular gTLD? Are there technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs exist?

Yes, no, and yes. Registries should be free to organize under any business model which they deem proper. The free market is quite capable, in our opinion, of determining the merits of any particular business model. Some examples of this would be strong brand-identification (and exclusive control) of a given TLD, a shared and completely open model, or a shared-but-restrictive model with an association of registrars tightly controlling business policies and practices. We believe that only the free market is qualified to determine the best business model for administering a particular TLD, and that the successful models will flourish while others fail of their own accord.

16. Should there be threshold requirements for registrars? Who determines them?

We believe that there should be threshold technical requirements for TLD servers, but not for registrars other than the promotion and protection of free market interests (ie: protection against a firm gaining a significant percentage of the total TLD namespace in use). Technical requirements exist to protect the network from instability and lack of ability to resolve a given name. There is no technical reason to set thresholds on the registrars themselves, and the two organizations, while related, are not directly tied to each other in all business models (ie: a registrar could "outsource" the DNS service for a given TLD)

17. Are there technical limitations on the number of domain name registrars?

No, other than those imposed by a particular software design (properly the decision of a gTLDs ownership, whether individual or by consortium).

18. Are there technical, business and/or policy issues about the namespace raised by increasing the number of registrars?

Yes, but only to the extent of communication and clear statement of policies and practice. Since the United States, as well as most other countries, already have statutes related to the misleading of consumers, this appears to be adequately covered by existing laws.

19. Should there be a limit on the number of gTLDs that a registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to a gTLD?

We believe that there does need to be a strong limitation on the number of gTLDs that a given registrar can administer to prevent the concentration of market power and the potential deleterious effects on the free marketplace. EDNS and MCSNet have proposed that the limit be 10 gTLDs under control or registration authority by any single registrar, and that this limit be independent of the organization method of the gTLD(s) in question.

20. Are there other issues that need to be addressed in this area?

Once again, we believe you have been very thorough and covered the matter under consideration well.

E. Trademark Issues

21. What trademark rights if any should be protected on the Internet vis-à-vis domain names?

The Internet is a communications media. It is not fundamentally different than print, radio, television or any other form of media, including individual spoken words. For this reason, the rights and enforcement mechanisms related to trademarks should track the existing legal protections which are accorded under the law. Specifically, there should be no further protection afforded a trademark owner on the Internet than in any other media. Today, this is not the case. In any other media, a trademark owner who alleges infringement must prove this in court in order to enjoin a potential infringement. However, under NSI's current policy, a simple allegation is often enough to cause a domain to be rescinded or placed out of service.

22. Should preliminary review be done?

No. The policing of trademarks is properly the domain and responsibility of their holders. The Internet must not become a place where trademark owners are able to run roughshod over people who would, in any other form of commerce or means of communication, be found not to be infringing a given mark.

23. Aside from a preliminary review process, how should trademark rights be protected vis-à-vis domain names?

In any other form of commerce, we have existing means of resolving those disputes, usually revolving around the court system of the country where the infringement is alleged to have occurred. The Internet is no different those who allege a violation of trademark rights should take their issues to a court of competent jurisdiction for adjudication. It is impossible for the Internet to recognize the disparate legal systems which exist around the world. The Internet is not a new world order it exists within the existing legal codes of the sovereign states which it traverses. Appropriate protections already exist for trademark owners under these legal systems, and disputes should be addressed to those tribunals.

24. How can conflicts over trademarks best be prevented?

A requirement to publish the registration information for any given domain under a TLD would significantly ease the protection of trademarks. Similar to the publication requirement that exists today for corporate and dBA registrations under United States law, this would be simple to implement and very inexpensive.

Steps beyond this are, once again, attempts to impose a different set of legal requirements on the Internet than exist in any other form of commerce and are inappropriate.

25. Should domain name applicants be required to demonstrate that they have a basis for requesting a particular domain name?

No. Trademark owners have the responsibility under the law for policing their own marks and enforcing their rights with relationship to violations they believe have occurred. There exists no reason to expand a trademark owner's rights on the Internet, and we should, as a matter of public policy, not expand trademark rights based on the media in use at a given time.

26. How would the number of different gTLDs and registrars affect the number and cost of resolving disputes?

Greatly expanding the number of gTLDs would actually reduce the number of disputes. For example, http://apple.records and http://apple.computers would not conflict. Today, there can only be one http://apple.com, which is bound to lead to a dispute. A publication requirement (say, through a widely-distributed Usenet newsgroup) would further reduce the cost of policing and searching for potential infringements from the situation we have today where no effective notification of use exists.

27. Where there are valid, but conflicting trademark rights for a single domain name, are there technological solutions?

Yes, but they require cooperation between the parties (ie: a "meta page" which references both companies on the initial load). The best solution is an expansion of the namespace so that http://american.airlines doesn't conflict at all there is only one American Airlines, but there are thousands of firms with the potential right to use the domain name http://american.com.

28. Are there other issues that should be addressed in this area?

We believe you have, once again, been very thorough in your discussion of the issues related to this section of the NOI.

We thank you in advance for consideration of our comments.

###

Number: 345

From: Jim Dixon <jdd@vbc.net>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 2:29pm

Subject: comment on Internet domain names

Attached is a statement on behalf of EuroISPA, the European Internet

Services Providers Associations, in reply to the Department of

Commerce Request for Comments on the Registration and Administration

of Internet Domain Names, docket number 970613137-7137-01.

Jim Dixon VBCnet GB Ltd http://www.vbc.net

tel +44 117 929 1316 fax +44 117 927 2015

CC: NTIADC40.SMTP40("chief@ispa.org.uk","secretariat@i...

EuroISPA

STATEMENT ON THE DOMAIN NAME SYSTEM

This statement is in reply to the US Department of Commerce Request for Comments on the Registration and Administration of Internet Domain Names [Docket No. 970613137-7137-01]. All sections of the Request for Comments are addressed. Our comments are italicized.

EuroISPA Background

EuroISPA is a federation of the Internet trade associations of Europe: AFPI in France, ECO for Germany, ISPA Belgium, ISPA UK in Great Britain, NLip for the Netherlands, AUI of Spain, and, shortly, AIIP of Italy. Representing nearly a thousand members, EuroISPA is the largest Internet trade association in the world.

EuroISPA is at 30 Rue Montoyer Bte 8, B-1000 Brussels, Belgium. Telephone +32 2 230 7560, fax +32 2 230 6491

Author

Jim Dixon, VBCnet GB Ltd, 78 Queens Road, Bristol BS8 1QX, United Kingdom. Telephone +44 117 929 1316, fax +44 117 927 2015.

A. Appropriate Principles

We agree. However, it must be recognized that much of the success of the Internet is due to co-operation, consensus, and above all respect for what works. The management of the domain name registration system must be based on models which are known to work. We believe that there is serious risk that the system that now seems most likely to be adopted is based on a flawed model, one which will generate endless disputes.

We concur with this as well. However, the influence of government is likely to be restrictive and damaging until there is better understanding by governments of the Internet. The practical reality is that the Internet is changing at a blinding pace and there are few in government who have any understanding of it.

Also, the consensus that is needed is consensus among those who manage and use the Internet. It is important that industries that feel threatened by the Internet, in particular the telecommunications community, should not be given undue influence over it.

In short, what is needed is self-governance: governance of the Internet by the Internet community, by those who understand it and wish it well.

c. These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.

Absolutely.

d. The overall framework for accommodating competition should be open, robust, efficient, and fair.

Of course. However, attention should be paid to non-competitive models where these are clearly successful.

e. The overall policy framework as well as name allocation and management mechanisms should promote prompt, fair, and efficient resolution of conflicts, including conflicts over proprietary rights.

Yes. But the Internet is primarily a system for transferring data. It is not a judicial system. Whatever domain registration system is adopted should be designed to avoid or at least minimize any legal entanglements. Where legal disputes arise, they should be adjudicated by courts, not by domain name registrars.

f. A framework should be adopted as quickly as prudent consideration of these issues permits.

Indeed. But there is at least one very serious problem with this general principle: the .com/.net/.org domains, which are the most widely used in the world, are currently managed by one company, NSI, as a monopoly. It is important that arguments about prudence not be used to perpetuate that monopoly.

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name registration systems?

Large parts of the current domain name registration system work very well. In the UK, for example, Nominet handles registrations in the .uk domain. Nominet is a non-profit company owned by those involved in registering domain names; there are several hundred members. Members of the general public can register either directly or through Nominet members (Internet service providers, Web site operators, etc.). The company is run very well and there are very few complaints about it.

Nominet operates on a first-come, first-served basis. Anyone registering a .uk domain name agrees to an arbitration process in case of disputes. Nominet makes every attempt to stay out of legal disputes over domain names. Nominet's success is due in large part to the fact that it is managed by people who have been involved in the Internet for a long time and understand and respect its traditions, particularly the need for consensus.

The InterNIC, which handles registrations in .com/.net/.org, is generally managed very well. However, NSI, which currently runs the InterNIC on a contract basis, has a generally poor reputation in the industry and in particular has been involved repeatedly in lawsuits, often because of actions perceived by the Internet community as arbitrary.

NSI is managed by people who have no particular interest in the Internet and no understanding of its traditions. This is not a minor consideration.

2. How might current domain name systems be improved?

There are three fundamental problems in the current system. First, while the way that the root name servers and "." (pronounced 'dot'), the basis of the name space, are handled was appropriate for an Internet meant for academics and amateurs, it is not appropriate for today's Internet, which has become crucial to the operation of many businesses. Secondly, .com/.net/.org are controlled by NSI, operating as a monopoly. Thirdly, .com is too popular: there are simply too many names in .com, and handling it is a serious technical problem.

The root name servers and "." are currently the responsibility of IANA and specifically Jon Postel. For whatever reason - presumably lack of resources at IANA - management of "." has in effect been delegated to NSI. This gives them the power to decide which Top Level Domains (TLDs) can be created and who has control over them. Given NSI's abrasive management style, this is unfortunate. Furthermore, NSI has no particular interest in managing "." well, and so it doesn't.

The root name servers are the machines which actually hold the top level directories. One of these is run by NSI. The others are run on a volunteer basis; most are in the US.

The management of "." would be much improved if authority over it were formally transferred to a new organization with endorsement from the Internet community at large; this would be IANA's successor. We believe that it would be beneficial if the individuals currently responsible for IANA, in particular Jon Postel, had seats on this new committee. Others should be appointed by Internet bodies such as the existing registries (RIPE, APNIC, ARIN), trade associations (the CIX in North America, EuroISPA in Europe, APIA in Asia), and academic institutions and networks such as JANET in the UK. There would be no benefit at all in the appointment of members from outside the Internet community, and possibly a great deal of harm.

Currently the root name servers contain a mixture of top level directory (".") information and .com/.net/.org data. These should be split up and put on different machines. The "." name servers should be controlled by IANA or its successor. That organization must have enough funding to be able to do this job, although it might very well choose to continue to leave the root name servers where they currently are.

It would be appropriate for IANA or its successor to also have enough money to fund continued development of the DNS software; this is currently done on a volunteer basis by Paul Vixie.

The NSI monopoly over .com/.net/.org should be ended. It is not necessary that NSI stop handling these domains. But it is urgent that other registries for these domains be created and that NSI be forced to share data with them. Furthermore, obstructionism by NSI should be expected and severely penalized if it does occur.

The third problem, .com being too popular, is a serious technical problem, but it is nothing more than that. This problem should be addressed by the Internet community, which has historically been very good at solving serious technical problems.

3. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

We believe that what would be best is a diverse system allowing different (but co-operating) models to be used simultaneously. There should not be one entity. There should be many.

For technical reasons, there must be one group, the successor to IANA, responsible for ".", that is, for the allocation and management of TLDs and for the root name servers. This should be a non-profit organization with members appointed from various organizations in the Internet community. Appointments should be long term. The group should be funded by the Internet community; possibly by the TLD registries.

At the next level down are the ISO country code registries and the gTLD ("generic" TLD, like .com) registries.

The ISO country code registries already operate well. They follow different models; most, like Nominet in the UK, are quite successful.

The gTLD registries could follow any number of models. For .com in particular, because of its size, it is preferable that a number of registries be created. These will have to share data. The CORE system proposed in the gTLD MOU is a good model for doing this.

However, it should also be possible to create gTLDs that are independent of the CORE system.

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

It is important not to overlook the fact that some of the existing domain registration systems are quite successful. Nominet, the registry for .uk, is one of these.

In so far as possible, the Internet should manage itself. Given the rapid pace of change in the Internet, governmental involvement or any involvement by outside organizations such as the ITU or WIPO is likely to be counter-productive.

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the .com registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

The gTLDs are popular for good reasons and many companies have substantial investments in them. It would be exceedingly unwise to force their abandonment. On the other hand, it would be desirable to apply some pressure to move US users into the .us name space where this is appropriate, simply to relieve pressure on the .com name space.

Companies with an international presence are going to want a name in .com. This is rational. But there is no reason for every corner store in the United States to have an international name.

The use of ISO country also simplifies some trademark disputes and should be encouraged.

6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

These questions have been answered above.

7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and co-ordinate?

It is widely accepted that the Internet address space will be exhausted in the foreseeable future. This problem is being dealt with and is largely irrelevant to DNS.

There do not appear to be scaling problems with the DNS in the near term, except possibly with .com. Once again, this is a technical problem of the type that has been solved very successfully by the Internet in the past and its discussion here does not seem useful.

Interoperation of the root name servers is not currently a serious problem but may become so in the fairly near future, because of NSI's role. As discussed above, nine of the root name servers are currently mixed "." and .com/.net/.org name servers. It is important that the "." domain be split out and put under the clear authority of IANA's successor as soon as possible, preferably before the 1 April 1998 end of the NSI contract.

8. How should the transition to any new systems be accomplished?

See the preceding paragraph.

9. Are there any other issues that should be addressed in this area?

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

Yes of course. However, the DNS system can easily tolerate many more gTLDs than currently exist, at least hundreds more.

11. Should additional gTLDs be created?

Yes. It would be quite beneficial to create gTLDs that correspond to specific industries or business areas. This would relieve pressure on .com.

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

There seem to be no technical problems, unless thousands of gTLDs are created. However, it is important that new gTLDs be created reasonably slowly and carefully, to minimize disruption and disputes. Also, whatever mechanism is established for creating new gTLDs must not be such as to lock in only one approach to the problem.

Specifically, there is no technical reason why all registries for new gTLDs should belong to one super-registry, like CORE in the gTLD-MOU proposal. It should be possible to create new registries or groups of registries that do not belong to CORE or establish alternatives to CORE.

13. Are gTLD management issues separable from questions about ISO country code domains?

They have little to do with ISO country code domains.

14. Are there any other issues that should be addressed in this area?

None that have not been already addressed.

D. Policies for Registries

15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

We believe that variety is desirable. Probably where a gTLD registry has exclusive control over a gTLD, it should be a non-profit organization. There are serious technical problems in developing shared registries, but these can be dealt with so long as they are addressed soon enough (specifically, in time to permit an orderly transition away from NSI's monopoly control of .com/.net/.org). Exclusive and non-exclusive gTLDs can certainly exist.

16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have? Who will determine these and how?

It is quite important that registrars be financially stable and technically competent. They are responsible for the orderly operation of their registries and interoperation with other registrars where registries are shared. Orderly operation covers a wide range of factors, from good billing practices, backing up data, and avoidance of legal disputes. The group succeeding IANA should ultimately have responsibility for setting down requirements and operational standards for registrars. Where a group of registrars share a common database, they may also create a common supervisory body (CORE in the gTLD-MOU model).

17. Are there technical limitations on the possible number of domain name registrars?

For all practical purposes, no. Nominet, who manage the .uk name space, have over 600 members who are independent registrars.

18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

Where the domain names are not shared, no. Where registrars share a common data base, yes, but these are obviously solvable. The operational requirements of the domain name system are many times simpler than those of airline reservation systems or stock exchanges.

19. Should there be a limit on the number of different gTLDs a given registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

In general, diversity is to be preferred, so fewer gTLDs per registrar is better. If a registrar has exclusive control over a gTLD, financial collapse of the registrar would be catastrophic for users, so multiple registrars per domain is better. However, there may be special circumstances arguing for different policies, so this sort of consideration is best left to whoever has authority over ".", IANA's successor.

20. Are there any other issues that should be addressed in this area?

None that can be addressed in the time available.

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-à-vis domain names?

Wherever possible this question should be addressed by local law and local courts.

However it has to be recognized that in general trademarks should have nothing to do with domain names. In the UK, for example, trademarks are specific to an industry, and there can be many companies with an equal claim on a name. Where an "international" gTLD such as .com is involved, there is no rational way to invoke trademark law. What we would suggest is that in nearly all cases registrants asking for domain names with international scope should be required to forfeit the protection of national trademark laws.

Requiring preliminary review is quite undesirable. It is likely to drive costs up and introduce delays that will be unacceptable to users. The only exception is where the review can be automated and based on generally acceptable simple principles.

The preferred model is always first-come, first-served. Registries should require anyone obtaining a domain name to accept an arbitration process. Wherever possible the registrar should not be involved in the actual arbitration.

The Internet does not make law and cannot judge legal disputes. Any domain name registration system should be designed to minimize the number of disputes, require that any disputes go to arbitration, and then insofar as possible direct disputes that cannot be arbitrated into existing legal channels.

Where a registry confines itself to an ISO country code and the ISO country code corresponds to a legal jurisdiction, this question can be easily answered, but the answer depends upon the local legal system.

Where international domain names such as .com are involved, the best principle would be that registrants have no rights under trademark laws. They would now forfeit their right to argue that the user of a domain name was "passing off", that is, attempting to impersonate another party. The claimant could seek redress in the courts. Where the registry is given conflicting orders by two different courts of competent jurisdiction, the registry must have the right to withdraw use of the name entirely.

Once again, for ISO country code registries this question is easily answered, but the answer varies from country to country.

However, the general principle is that this would impose an undue cost and an undue responsibility on the registrar. "First-come, first-served" should be the default rule.

26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

If the principles set out above are followed, trademark disputes would be minimal. However, if all of the complexity of international trademark law is allowed in, the costs involved would explode as the number of gTLDs and registrars increased and the financial stability of the registrars would be seriously at risk. It is vital that the connection between trademarks and domain names be severed.

Furthermore, as a general principle, legal responsibility should always remain with the registrant.

27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

No. DNS is universal. Such cases should always be settled on a first-come, first-served basis.

28. Are there any other issues that should be addressed in this area?

We believe that it is of fundamental importance that trademark conflicts be kept out of the Internet. Old law should not be used to cripple new technology. And therefore we believe that as few trademark issues should be addressed as possible.

###

Number: 346

From: Richard Hovey <HoveyR@mail.dec.com>

To: "'dns@ntia.doc.gov'" <dns@ntia.doc.gov>

Date: 8/18/97 2:57pm

Subject: DIGITAL comments on DNS noi

I am responding on behalf of Digital Equipment Corporation (DIGITAL) to

the Department of Commerce "Request for Comments on the Registration and

Administration of Internet Domain Names."

On May 1 DIGITAL signed the generic Top Level Domain Mememorandum of

Understanding, and attached the following statement of concerns

[ also posted at http://www.itu.int/net-itu/gtld-mou/decadd.htm ].

We believe this action and statement are responsive to questions on which

the Department has requested comment.

Richard Hovey

Corporate Strategy and Technology

Digital Equipment Corporation

--------

DIGITAL EQUIPMENT CORPORATION ATTACHMENT TO ESTABLISHMENT OF A

MEMORANDUM OF UNDERSTANDING ON THE GENERIC TOP LEVEL DOMAIN NAME SPACE

OF THE INTERNET DOMAIN NAME SYSTEM (gTLD-MoU)

As the first computer vendor to be connected to Internet and the first

to register in the ".com" domain, DIGITAL is particularly pleased to be

the first to support the efforts of the IAHC and other founding

organizations to sustain the Internet's growth and stability in an

international setting. We also thank both the IAHC for bringing the

proposal forward and those organizations running the naming mechanisms

for the current Top Level Domains.

DIGITAL believes that the principle enunciated in the MoU that the

Internet Top Level Domain name space is a public resource and subject

to the public trust is fundamental to the stability and future evolution

of the Internet.

We view the step of signing the Memorandum of Understanding as a

beginning and not the end of a process, and are attaching to our MoU

this set of concerns we believe must be addressed as the process

progresses. These concerns are among the issues Digital intends to focus

on as we participate in the next phase of work.

o Consensus building:

Many affected parties, including governments, consumers, and commercial

concerns, have not yet fully considered the issues embodied in the MoU

or represented their concerns to their satisfaction. We expect the

process going forward to be fully open to views from all interested

parties in order to extend the public consensus which is the foundation

for the Internet.

o Stability of existing generic Top Level Domains:

The MoU indicates that the existing gTLDs; such as ".com", are not

immediately subject to the provisions of the MoU, but foresees that

eventually they may be included. Plans need to be developed which

ensure the existing gTLDs will not be disrupted in the future or

subjected to IP address re-assignments.

o Stability of name registration:

Competitive, geographically dispersed registration may increase the risk

of name conflicts. We will work to see that the structure and

implementation of any system for any new gTLDs will not compromise the

integrity or accuracy of the Internet domain name system and will

continue to support the rapid rates across the Internet to avoid

barriers to growth.

o Preserving Intellectual Property Rights:

Administration functions for evaluating and resolving trademark issues

need further clarification and we intend to work with other parties to

develop a consensus-driven regime for the allocation of domain names that

respects the intellectual property rights of all and the established

legal mechanism for protecting these rights. The process for resolving

conflicts in domain names must extend and not replace established legal

procedure.

DIGITAL believes the MoU is a very good beginning. However, DIGITAL

reserves the right to reconsider its support until these issues, among

others, are successfully resolved. We are supporting the MoU now because

we do think it is a good beginning, and we are optimistic that the issues

DIGITAL and others will raise can be successfully resolved.

04/30/97

###

Number: 347

From: Barbara Dooley <bdooley@cix.org>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 3:37pm

Subject: CIX Response with Member List

Request for Comments on the )

Registration and Administration ) Docket No.

970613137-7137-01

of Internet Domain Names )

Please accept this as the filing with the Member List attachment.

Respectfully,

Barbara Dooley

Executive Director

Commercial Internet eXchange Association

1041 Sterling Rd. Suite 104A

Herndon, VA 20170

v. 703-709-8200

f. 703-709-5249

www.cix.org

Before the

DEPARTMENT OF COMMERCE

NATIONAL TELECOMMUNICATIONS

AND INFORMATION ADMINISTRATION

Washington, DC

Request for Comments on the )

Registration and Administration ) Docket No. 970613137-7137-01

of Internet Domain Names )

Comments of the Commercial Internet eXchange Association (CIX)

Introduction. Within the last four years, the Global Internet has been transformed from a largely research and academic-oriented network to one that is predominantly commercial. The commercialization of the Internet has been led by the private sector, principally but not exclusively US information technology and new media firms, with the active encouragement of the US Government.

This process of commercialization will reach an important milestone in early 1998 when the National Science Foundation's (NSF) cooperative agreement with Network Solutions, Inc. (NSI) to register and administer domain names for certain generic top level domains (gTLDs) ends. Administrative responsibilities for the DNS pursuant to this agreement at the US Government's behest will soon be exercised through new institutional arrangements, possibly by new individuals and entities. A smooth transition over the next few years to a permanent self-governing Internet administrative structure is essential if the Internet's commercial growth is to continue.

The specific issue of gTLD expansion within the DNS is significant as a result of the Internet's growing commercial role. Domain name space is one of the critical elements of the Internet along with IP addresses, IP protocols, the physical transmission and routing plant, and Web sites. Achieving the best solutions on issues of this magnitude warrants time since major modifications to the Internet's crucial components such as DNS may occur slowly once changes are adopted and implemented.

In l991 the Commercial Internet eXchange Association (CIX) established a private commercial exchange point in California as an alternative to the network that was operated for the US Goverment. Today CIX is the largest and oldest trade association of Internet Service Providers (ISPs) with almost 170 members from the United States, Asia/Pacific, and Europe (a recent list of members is attached) and continues to operate the California exchange point. CIX members represent a wide variety of ISPs, ranging from large providers to backbone carriers to smaller regional and local networks and state trade associations.

ISPs are an essential component of the Internet community. They provide the Internet user's physical connection from the local telephone loop to the Internet from their local point of presence. Of particular relevance to this docket is the fact that ISPs also help commercial customers to apply for a domain name, renew it if desired, and obtain their Internet addresses. Some ISPs may also provide Web design or hosting services or help process transactions for their commercial customers although these functions are not considered part of their Internet access service. In some countries where the registry is operated commercially, ISPs may have a formal role in managing the country registry. Thus, ISPs play vital roles both in helping to construct parts of the Global Information Infrastructure and facilitating the presence of commerce on the Internet.

CIX welcomes the US Government's request for comments on changes to DNS registration and administration of generic TLDs since a stable DNS is vital to the smooth operation of the commercial Internet. Destabilizing changes to the present system could profoundly affect our members' operations and services and the consumer's Internet service quality. While changes to the current DNS will be necessary in the future to accommodate the Internet's explosive growth, maintaining network stability, security, and connectivity during this transition period must be the US Government's and industry's primary goal. Stability and connectivity must also result as well from proposals to change the DNS as the US Government's role diminishes in importance.

The US Government has a unique historical responsibility for the foundation of the modern Internet as a result of its funding and developing of IP network technologies and the role of the Department of Defense in its early history. This historical role and authority have crucial legal consequences and shape the debate on Internet administration during this transition.

CIX will (1) address the five issue sets identified in the notice: appropriate principles; general/organizational framework issues; creation of new gTLDs; policies for registries; and trademark issues; and (2) propose changes to improve the management of existing commercial namespace such that it will help meet the burgeoning demand for commercial names for the immediate future while creating relatively minor disruption in the Internet and user communities.

CIX believes that its proposed modifications (1) will be relatively inexpensive, efficient and simple to administer; (2) result in lower costs for DNS users; (3) stimulate competition among registrars; (4) will be competitively neutral; (5) give policymakers additional time in which to devise long-term solutions; (6) better meet the needs of the commercial sector better than some more drastic proposals; and (7) respond to marketplace forces.

The proposals are intended to apply to commercial situations and to address issues that affect business. They incorporate and extend current practices existing in the domain names system and are intended to be evolutionary in character and appearance and can work with other proposals to alter the DNS.

A. Appropriate Principles

CIX enthusiastically endorses the principles enumerated in Part A as there can be no disagreement about their value. CIX proposes three additional principles for inclusion in this section.

a. CIX agrees strongly with the proposed statements affirming the leading role of the private sector, consensus-based management, competition, and self-governing mechanisms. Any proposal to expand the number of gTLDs or significantly alter current DNS procedures must satisfy basic due process requirements both formally and substantively.

Critical questions have been raised about the authority of decision makers and will subside only if those asserting leadership consult with representatives of significant Internet stakeholder groups. Therefore, CIX recommends the addition of the following principle on openness and inclusiveness:

"_. Policymakers with authority to make decisions on domain name issues should consult widely with the representatives of affected stakeholder groups and ensure that its processes are inclusive and that creditable views receive appropriate consideration."

b. Consistent with the growing commercial importance of the Internet, CIX proposes the following new principle that stresses a reliance on market forces to determine which domain name changes are needed by the commercial sector. This proposed new principle also provides CIX's answers to several questions posed in the instant notice, such as whether ".com" should be dropped. If business users prefer a ".com" domain, that gTLD should continue. When additional business-related gTLDs are needed, representatives of commercial users, including small business, should be actively consulted for their views on the number and names of the new gTLDs through formal, open, and public procedures.

"_. Modifications to the registration and administration of gTLDs and country TLDs which affect commercial enterprises, including the addition of new gTLDs, should be responsive to market forces and take into account the views of business users and their Internet vendors."

CIX also endorses the principles expressed in (b) relating to the responsibility and accountability of policymakers. Only enforcement of due process can ultimately sustain responsibility and accountability, and CIX recommends the addition of the term "enforcement" in Paragraph A, so that policymakers will clearly understand that public accountability is expected and required and will be sustained by appropriate and effective enforcement.

c. Finally, CIX recommends adding a third new principle to Part A on the legal status of the namespace. It has been treated as a private resource for many years, which does not necessarily mean that there should be trafficking in desirable second level domain names as private property. On the other hand, there have been recent assertions, notably in the gTLD-MoU, that the namespace is a public resource held in public trust, a position that implies resource scarcity and, indeed, that can be used to justify governmental regulation. Unlike IP addresses or telephone numbers, the number of domain names is far less severely constricted and can be devised easily. They are not technologically essential to the operation of the Internet but provide the ease of use of alphanumeric identifiers. Therefore, CIX recommends adding the following paragraph at the end of the principles section:

"_. The TLD namespace is a private resource subject to reasonable limits developed and agreed to by the Internet community on a consensual basis."

The translation of principles into a workable transition plan would require considerable time even if there were a consensus about the problems and remedies. To give the interagency committee adequate time in which to conclude its task, at the conclusion of this comment period, the committee should first determine the US Government's principles. Second, it should then solicit proposals for a transition to a permanent DNS namespace and registry system based on these principles, which might include ideas which have already been formally proposed such as those suggested by the IAHC, eDNS, and other organizations. Alternatively, the Executive could impanel a US Government committee to craft the transition plan, similarly as the NSF led the thoughtful and stable transition from NSFnet to commercial backbones. As part of this process, the panel should continue to solicit the views of non-American individuals, companies, non-profit organizations, academics, and governmental entities.

Since the highest priority is to ensure network stability and security during the transition, the US Government should take concrete steps to achieve these goals. Furthermore, if the requirements for a stable, proven DNS gTLD structure are not in place by April 1, 1998, the United States Government should extend its agreement with NSI for a time certain to prevent registry or administrative problems. Sensitivity to the concerns of individuals and entities located outside the United States should not automatically preclude a transition plan that is anchored in the United States with a potential US Government oversight role during the transition, since currently the majority of Internet users and resources are located in the US, as long as the Internet continues to evolve into a fully global infrastructure with international administrative infrastructures and governance.

B. General/Organizational Framework Issues

Proposals to change DNS registration and administration have been stimulated by the imminent end of NSI's current cooperative agreement with the NSF and, to a lesser extent, complaints about the quality of registry services, trademark disputes or limitations in the current DNS design. There is an appealing, superficial attraction to the argument for fundamental institutional changes since the original DNS was created to serve a far smaller user community. Thus far, however, there is disagreement as to the need for and the timing of additional gTLDs, as the large majority of gTLD registrations appears to be concentrated in the commercial space and from North American customers.

CIX proposes the following propositions for this section on framework issues:

Incremental evolutionary recommendations that result in minimal necessary changes are usually preferable to radical changes on which there may be no community consensus.

Changes which will significantly affect commercial users and their Internet providers should be undertaken after formal and substantive consultation with their representatives in developing these proposals. It is understood that only the views of those stakeholders and stakeholder groups who choose to participate in open processes can be used to develop policy and new administrative forms.

Membership in the permanent policy and administrative bodies should be balanced and inclusive of major Internet stakeholder groups.

Minimal governmental involvement and private sector leadership continue to be valid principles. Any significant increase in governmental participation or influence on Internet policies or in Internet governance can be justified only after comprehensive public debate supported by the public record.

National governments and international governmental organizations' appropriate roles are to help stabilize the Internet (if needed); facilitate necessary changes if the private mechanisms fail; and to offer counsel.

Market forces will most efficiently determine the need for new gTLDs and the disposition of old names.

During the transition to a new DNS regime and thereafter, it is appropriate for the US Government to play a stabilizing role based on its historic legal authority as long as this role is limited and the system moves towards a genuinely global character.

On the basis of these principles, CIX will comment on the existing system, selectively discuss proposed alternatives and propose a potential compromise consisting of ideas drawn from a variety of sources.

Question 1. What are the advantages and disadvantages of current domain name registration systems?

The disadvantages of the current gTLD system were largely inherent in its initial design, and administrative decisions such as national monopolies on registration authority, and lack of sufficient commercial resources and attention. National anomalies in administrative practices, and legal uncertainties present problems to public and private sector policymakers. There was no procedure prescribed through which new gTLDs could be created when needed.

There were advantages as well, including DNS' private sector origins, the design's simplicity, ease of use and understanding and versatility. On the whole, there can be little question that the DNS has worked quite successfully and has been an important factor in the expansion of the Global Internet.

Question 2. How might current domain name systems be improved?

CIX believes that only simple steps are needed to mitigate the alleged lack of desirable commercial names. In addition to the possible creation of new gTLDs, CIX proposes that commercial users in the US should be offered the use of the ".us" country TLD, with a "com us." second level domain (SLD) available to them.

Second, an experts panel should study and report why the ".us" is perceived as too complex by commercial users, comparable practices of other national registrars and the "us" registrar and mandate reforms to the ".us" TLD space to attract more business users, thereby reducing demand for ".com". Similar reforms to simplify and improve the use of country TLDs are taking place elsewhere, most notably in Canada.

A second way to expand namespace to satisfy commercial demand would entail the "cloning" of the existing ".com" name space. Such cloning could enable, for example, multiple "smithcompany.com" with virtually the identical domain name except for a slight modification to the ".com" with the addition of an extension in that space (for example, ".com1," ".com2," etc.) While cloning does not resolve trademark conflicts, it could mitigate substantially the demands for the same namespace by identically or similarly named companies.

Question 3. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

The limitations of the current DNS have sometimes been attributed to NSI even though the naming system conventions, with their inherent problems, and the number of gTLDs were fixed prior to its cooperative agreement with the NSF. Moreover, there is widespread disagreement whether there is a systemic DNS "crisis" that justifies new institutions and bureaucracies.

CIX proposes that the private sector organization assume responsibility for registering, administering, and managing the commercial domains in the US TLD through a non-profit entity with delegated authority from either the Federal Networking Council or NSF. This group would be the US Registrar for the ".com.us" second level domain (SLD) and set SLD commercial use and policies and be responsible for marketing, registration, and administration. It would be governed by a board of representatives of the commercial sector.

The new commercial board members, after its initial organization, would be elected to two-year terms and would represent stakeholder interests rather than an individual company or trade association. It is intended that elections ensure wide participation by representatives of qualified organizations, promote responsiveness, encourage diverse views and inclusive policies, and prevent the emergence of entrenched, self-perpetuating interests.

With the exception of an Internet standards organizations like the IETF, no other individual organization would enjoy guaranteed representation. In carrying out its functions, the new commercial board could issue an RFP for an organization(s) to handle registration and administer the domain, develop the operational procedures and rules under which the registries would operate, as appropriate.

The board would be required to operate under open and transparent rules and regulations and be subject to audit and due process requirements.

CIX suggests that the interagency committee raise with the NSF the possibility of using the funds being collected for infrastructure purposes by NSI at the direction of the NSF (perhaps exceeding $25 million) to support financially the start-up costs of the new ".com.us" registry and its non-profit board and perhaps to support similar efforts in Latin America where Internet service is new and expanding. Registry activities would be particularly appropriate for being seeded since the money was originally raised though levies on domain names.

An additional financial issue should also be resolved at this time. NSI's current fees include a substantial (30% of the total) amount for infrastructure purposes at the direction of the NSF, which has not yet determined how the fund should be used. This "infrastructure surcharge" is in effect a tax which has been applied arbitrarily to only one segment of the domain name user community and has fallen disproportionately on US users. If an "infrastructure surcharge" needs to be collected, it should be applied and distributed in an equitable manner. This current 30 percent supplement should be discontinued and the fees reduced accordingly.

Question 4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

While no situation is precisely analogous to the Internet governance model, the example of the North American Numbering Plan, which was formerly a monopoly function within the private sector, might be applicable to the current situation. However, it is also important to remember the differences as well, which include the essentially commercial significance of name space and the fact that names are not limited as are numbers.

This question also raises the critical issue about the proper roles of national or inter-governmental/non-governmental entities. The Internet has flourished in large part precisely because until recently it lay outside the traditional regulatory attention of governments, which have concentrated instead on the centralized telephone monopoly. Instead, both inside and outside the US, the Internet was largely constructed by private entities operating through non-governmental structures.

CIX believes that the proper role of governments and intergovernmental organizations is limited to that of counsel and support and is deeply concerned about the growing influence of international governmental organizations, best illustrated by ITU's and WIPO's inclusion as permanent members of the gTLD-MoU Policy Oversight Committee. CIX reaffirms its views that government regulation tends to slow economic development and global growth of the Internet if extended beyond legitimate but limited consumer protection functions.

Question 5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the ".com" registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

The disposition of old gTLDs and creation of new ones are properly left to be determined by market forces. If companies want to continue using the ".com" name, it is because they have determined it has value to them. Commercial users' decisions should be respected and not be dictated by political or bureaucratic fiat which ignores their preferences by terminating the ".com" gTLD or creates unneeded new ones. Likewise, the private sector will indicate when additional names are needed and what they should be. A comment period could be opened to solicit opinions on the need for additional gTLDs and their proposed names. Only after the business sector has had an opportunity to comment for the public record should new gTLDs be created.

Question 6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

Question 7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and coordinate?

There are no publicly available technological solutions to solving the many domain name registration issues at this time. There are questions as to the timeline for the development of these shared databases with heavy real-time demands. Therefore, extensive technical trials and experiments must still be conducted constantly to ensure that answers devised on the basis of computer models workin practice.

The domain names controversy could be readily resolved through use of directory databases, which can easily scale to growth and would diminish the demand by some firms for distinctive domain names if they can be reached easily. Moreover, directories would resolve some trademark issues as well since a multinational corporation would have a single address with only a limited prospect for trademark disputes. The creation of global Internet directories would be a complex technological task requiring additional research and development, but efforts are underway.

The possibility of using existing Internet private sector organizations such as ARIN, RIPE and APNIC for the coordination and implementation of root server architecture should be explored.

Question 8. How should the transition to any new systems be accomplished?

The transition to a permanent DNS structure will be critical to ensure network stability. With legal and technical stability foremost in mind, NSF has alternatives to ensure that it is clear who has the ultimate legal and operational responsibility for its current databases and root server. It will also be necessary to begin working with the major technical and professional Internet organizations such as the IETF and IAB to develop protocols, schedules, flow charts, and work plans, objectives, and identify areas for additional research.

Question 9. Are there any other issues that should be addressed in this area?

This question asks specifically for issues that were not listed above. CIX believes that there are legitimate concerns about the great disparity in the treatment of the gTLDs and the country TLDs by the former IAHC for reasons that were never explained. As a result, it appears that the United States and, to a lesser degree, Canada, would be disproportionately affected by its proposals for restructuring the existing system since the use of ".com" is widespread in North America. Moreover, the US will be affected by any attempts to shift the infrastructure ownership to another locale, e.g., Switzerland.

CIX also notes that there are critical questions about the authority for various parties to take certain actions affecting Internet administration in the absence of either explicit and verifiable delegation. Clearly, the authority to administer certain critical tasks derives from the work performed in behalf of the US Government. As a result of its early volunteer work, the IANA, located at the University of Southern California, assumed authority to designate the national registrar for country top level domains in each country and to allocate IP addresses. In addition, the IANA also assumed responsibility for administering the ".us" TLD. However, the legal authority to do perform these tasks is ambiguous as it was never codified nor is it substantiated in any legally verifiable form. The interagency committee should clarify the questions of authority by drafting a statement that reaffirms the position of the FNC and NSF as the sole legitimate and legal sources of authority, which will be delegated permanently to a successor at the end of transition.

An important issue that is not addressed in the questions is future management of the root server infrastructure, which is currently handled by NSI and which is crucial to the operation of the commercial Internet.

Finally, CIX recommends that the means be found to open up country TLDs as well as the gTLDs and that international initiatives be taken to harmonize varying DNS policies which are found to affect gTLDs.

C. Creation of New gTLDs

Decisions on creating new gTLDs should be left to the market, which will determine whether new ones are needed, when, and, if so, which ones, far better than any politician or engineer. Market data will show where there is a demand and users' preferences. Both sound commercial and engineering principles indicate the need to make no changes which might destabilize the infrastructure or create consumer confusion.

Question 10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

As a general proposition, there probably is no technical problem to adding gTLDs or country TLDs. Country TLDs have been added at a fairly rapid rate recently with no problems reported. From a business perspective, the problem may not be technical but rather potential market confusion. At this time, further research, planning and development of ways to accommodate changes to the existing system beyond a limited number of gTLDs or country TLDs is needed to achieve support for our goals of a more competitive, flexible DNS.

Question 11. Should additional gTLDs be created?

Concerns that name space in the ".com" registry is being depleted and that complementary business gTLDs are required were once a rationale for restructuring DNS registration and administration. Documentary evidence of systemic problems with the ".com" gTLD have yet to be substantiated, but anecdotal evidence suggests the need for increased options for commercial users. Irrespective of the heightened demand for commercial names, CIX believes that its proposed changes to increase use of the ".com" as second level domain within ".us"; operate that SLD as an attractive namespace for many commercial enterprises, and to consider an expansion of the generic space under some schema, for example by the cloning of ".com," could satisfy foreseeable business demand and give policymakers and the Internet community ample time in which to develop sound, fair long-term solutions which scale to future needs.

Question 12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

According to some experts, the issue is efficiency -- not technical possibility. The business manager may also question the additional complexity if the system becomes more confusing -- and perhaps more difficult to use for the ordinary consumer.

Question 13. Are gTLD management issues separable from questions about ISO country code domains?

CIX sees no reason to distinguish between gTLDs and country TLDs in some cases as when names could be substituted, for example, when an American firm has the option for registering for ".com" or ".com.us". This is also true for non-US applicants who can choose between ".com" or a country TLD and a space in the "com" SLD. In the end, price and service may determine the final choice.

Different national registrars have diverse policies, some of which may reflect national conditions without any external impact. The desirability and feasibility of harmonizing DNS policies deserves serious study since national policies can have international effects, and a review of the registration and administration of ".us" is particularly urgent because of this country's heavy demands for domain names, the brand value and symbolic importance of the US's country code, and the leadership which the US exercises on Internet matters. Decisions to modify and adapt ".us" for commercial purposes and create a new registration and management structure should be undertaken prior or in parallel to implementation of the final DNS system.

D. Policies for Registries

Question 15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

While there has been strong disagreement even among experts, there appear to be significant technical and systems obstacles at this time to the use of shared registries for gTLDs and hence to having multiple registrars, even those domains with a relatively low level of activity - not to mention busy ones like ".com" and ".org". Shared registries for gTLDs presently do not exist, and many questions must be answered first before the software is available to enable the competitive model to be achieved. Policymakers and managers must define their business requirements with granularity before the necessary software programs can be developed, such as order of processing applications, effect of delays in receipt of e-mail applications, etc. It is unclear now how long it will require to overcome technical limitations. The interagency committee's transition plan should be realistic in assessing the technical problems, provide flexibility for unanticipated delays, and base the transition plan on successfully solving problems rather than artificial schedules.

Once a gTLD (or country TLD registration) becomes competitive, registrars must follow similar policies to ensure uniformity in information collected and processed and in policies to be followed in that domain.

There is no reason why exclusive and non-exclusive registries should and could not coexist. Experience and research may show that it will be easy to migrate some of the existing gTLDs to a competitive environment, while others may move more slowly. Further, new gTLDs which are designed to be competitive from their inception obviously would not face the problems of existing domains and might succeed as competitive registration and administrative business models.

Question 16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have? Who will determine these and how?

CIX's proposal for a private sector board to administer the commercial space in the US TLD envisions this task to be farmed out to technically experienced organization[s] to develop registrar qualifications and procedures. There are alternatives to delegating this task of setting registrar requirements, as proposed by CIX. Standard qualifications might include technical knowledge and skills, ample back office operational skills, billing and financial experience, financial resources, and information technology experience. CIX recommends that registrars be required to escrow their billing records with a designated authority in the event they leave the business or go bankrupt, so customers are adequately protected. It is not clear that the handling of applications for domain names alone is a reasonable business without providing additional services such as Internet access, design, hosting, etc.

Question 19. Should there be a limit on the number of different gTLDs a given registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

Limiting the number of different gTLDs a registrar can administer is the type of detailed policy questions that should be deferred until we have more experience with the permanent DNS structure and the market for registration services.

E. Trademark Issues

Trademark issues have tended to dominate discussion of the business problems generated by gTLDs registration and administration although such issues are a small, albeit highly significant, subset of the overall business issues which should influence policies in this area.

Trademarks are critical to multinational corporations with global brands and which deploy ample legal resources to and protect their brands. This is not true for the bulk of companies globally whose primary consideration is simply to establish a viable commercial presence on the Internet rather than a domain name will mislead or confuse. Therefore, increasing the availability of commercial namespace is at least as important as the trademark issues which are dealt with extensively in the gTLD-MoU.

Question 21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-à-vis domain names?

The Internet vastly increases the scale and immediacy of trademark issues. However, it is unclear exactly how current national protection policies conflict with the Internet and how, if at all, national policies are deficient. Before new international arrangements and institutions are created, there must exist a better understanding of national trademark regimes in the digital environment.

An essential early step in improving understanding of DNS trademark conflicts and solutions should include comprehensive hearings by the US Patent and Trademark Office, its foreign counterparts, and, perhaps, the appropriate Congressional committees. Such hearings will improve our understanding on the impact of the DNS on trademark law and which trademark policies and institutions are inadequate in the new digital environment. The Department of Commerce's efforts in this docket begin this important educational process, but in and of itself, may not elicit sufficient information from trademark experts and ISPs for a comprehensive overview of the problems or solutions to DNS trademark disputes.

Question 24. How can conflicts over trademarks best be prevented? What information resources (e.g. databases of registered domain names, registered trademarks, trade names) could help reduce potential conflicts? If there should be a database(s), who should create the database(s)? How should such a database(s) be used?

CIX strongly supports intellectual property rights as a cornerstone of the modern information-based economy and acknowledges the problems some trademark owners have confronted with DNS domain names. ISPs would like to be part of a solution that is balanced, effective, and consistent with Internet technology.

Timely resolution of trademark disputes is essential, and the gTLD-MoU attempts to deal with this issue through WIPO constituted challenge panels. That function is valuable, but it is not at all clear why it justifies granting an international governmental organization a seat on its policy council or whether DNS trademark is so important as to justify allocating two committee seats to organizations concerned solely about trademarks even as business and commercial organizations lack representation. If, in the end, it is agreed that WIPO's dispute resolution services are needed, they can be provided under contract to the appropriate organization.

Question 25. Should domain name applicants be required to demonstrate that they have a basis for requesting a particular domain name? If so, what information should be supplied? Who should evaluate the information? On the basis of what criteria?

Trademark issues should not be allowed to dictate registration procedures to the detriment of all commercial users and other business considerations. For example, preliminary delays in processing registration applications would inconvenience applicants and could even cause a substantial business problem for those whose needs are time-sensitive. Additional databases devoted solely to checking registered trademarks and tradenames would likely increase registration costs to all applicants and ISPs. Applications which require the applicant to demonstrate a basis for requesting a domain name would surely cause processing delays, raise preparation costs (perhaps even involving legal help), impose a burden on small businesses, and involve more administrative burdens for uncertain benefits. Before policy and organizational changes are adopted, there should be a base level of understanding of the extent and gravity of problems, benefits and costs of various proposals, alternatives, and the time frame in which the solutions are needed.

The proposal in this question stems an earnest attempt to solve a problem, but it deserves a separate NOI since its implications are so far reaching. As in several other questions requesting responses to detailed proposals, it is possible that a satisfactory solution will gradually only with experience and time. CIX's concerns, noted above, are motivated by the potential impact on innocent commercial customers and the fact that this proposal shifts part the burden from the prospective plaintiff/challenger to the applicant and the service provider who handles the registration application.

Question 27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

Additional namespace as proposed by CIX can be created by exploiting existing technologies and may help solve some trademark disputes because two or more identically denominated entities could share a similar domain name. Otherwise, a domain name address would be exclusive to the original applicant within the gTLD or country space. Disputes over registered names of strong global or regional brands would not be solved by CIX's proposal.

CIX appreciates the opportunity to provide its views on these important issues involving the development of the Internet as a commercial network, its critical components, and the establishment of permanent administrative mechanisms in this process. The roles of Internet and Online Service Providers and the business community across the globe are central to the successful growth of the Internet. CIX looks forward to continued participation in the interagency committee's process and in other domestic and international fora which deal with critical Internet issues and the

interests of network service providers.

Respectfully submitted,

COMMERCIAL INTERNET EXCHANGE ASSOCIATION

Barbara A. Dooley

Executive Director

Eric Lee

Public Policy Director

Robert D. Collet

Chairman of the Board

Commercial Internet eXchange Association 1041 Sterling Rd. Suite 104A Herndon, VA 20170

(703) 709-8200

Commercial Internet eXchange Association Members

August 1997

@ Home

a2i Communications

ACSI

Apex Global Information Services

Aliant Communications

American Network Inc.

ANS CO+RE Systems

Ascend Communications

Ashton Communications (AICnet)

Asociados Espada

AT&T

AT&T Jens Corporation

ATMnet

Atson, Inc.

BBN Planet

Bekkoame Internet, Inc.

British Telecom

Bull HN Information Systems Inc.

Cable & Wireless Internet Exchange

Centnet

CERFnet

Comnexo

CR Internet

CRL Network Services

Crocker Communications

CTD Technologies, Inc.

CTS Network Services

Cybergate, Inc.

Dart Net Ltd.

Data Research Associates, Inc.

Data Xchange

Datanet Communications Ltd.

Demon Internet Limited

Digital Equipment Corporation

Digital Express Group

Dimension Enterprises

DirectNet Corporation

E-Z Net

easynet DV GmbH

Easynet Group Plc

Electronic Systems of Richmond, Inc.

Emirates Telecommunications

EPIX

Epoch Networks Inc

Eskimo North

EUNet BV

EuroNet Internet BV

Exodus Communications

Fiber Network Solutions, Inc

Fibrcom, Inc.

Fujitsu Limited

Genuity, Inc.

GetNet International

Global One

Global Village Communication

GoodNet

GridNet International

GST Internet, Inc.

Hitachi

Hong Kong Supernet Limited

Hookup Communications Corp.

Hewlett Packard

Hurricane Electric

I-2000

IBM Global Network

ICon CMT

Inet, Inc.

InfoCom Research Inc.

Integrated Network Services

Intermedia Communications Inc.

Internet Bermuda Limited

Internet Corporativo, SE de CV

Internet Exchange Europe

Internet Initiative Japan (IIJ)

Internet Prolink SA

Internet Public Access Corp.

Interpath

Interserve Communication (H.K.) Ltd.

IPF.Net International

ITnet SpA

IUnet s.p.a.

JC Information Systems

JTNET Research Institute

Kokusai Denshin Denwa, (KDD)

Korea Telecom

Lafitte, Morgan & Associates

LDS I-America

Logic Telecom S.A.

Logical NET Corp. (Micros)

MCI Telecommunications

Mikrotec

MIND (Mitsubishi Electric Network Information Co.)

Minnesota Online

Nacamar Data Communications GmbH

NEC Corporation

Netcom

NetDirect Internet

netINS, Inc.

NETRAIL

NetNet

NetVision

Netway Communications

New York Net

Novia Internetworking

Octacon Ltd.

On-Net

Osaka Media Port Corporation

OSI de Guatemala, S.A.

OTSUKA SHOKAI Co.,Ltd

Pacific Bell Internet

Pearl Vision

Pilot Net Services

Planet Online Ltd.

PSINet

Qwest Communications

RACSAnet

Renater

Rapid Systems, Inc.

Red Creek Communications

Singapore Telecom

SOVAM Teleport

Sprint

Sun Microsystems

Synergy Communications

Tchui Data, Ltd.

Telecom Finland

Teleglobe, Inc

Telewest Communications, Ltd.

The Internet Mainstreet (TIMS)

TheOnRamp Group, Inc.

Thoughtport

Threeweb Corporation

TogetherNet

Tokai Internetwork Council

Tokyo Internet Corporation

Total Connectivity Providers

Toyama Regional Internet Organization

U-NET Ltd.

USIT United States Internet, Inc.

UUNET PIPEX

UUNET Technologies

USAGate

VBCnet (GB) Ltd

VoiceNet

Voyager Networks, Inc.

Web Professionals

WebSecure

Verio

###

Number: 348

From: president- <parker@mibor.net>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 3:02pm

Subject: domain system

The domain system works and should not be changed. If the new systems is

implemented it would cause confusion to the end user along with a flood of

domain dispute policies. While the new system will solve some problems, it

will likely create far greater ones that cannot be reversed. Not only will

the new system create more trademark disputes but will also create market

confusion to the end user. We need to remember its the end user who really

controls the FUTURE, and yes the Internet leaders need to hold there hand,

but confusion and frustration detracts the end user from enjoying the

Internets full potential. I feel the Internet needs to be more stable

before the new system is and IF implemented. I think there needs to be more

than one register to the existing domains. However, we need to maximize

and stabilize the all ready existing domains. Please do not add more

top-level domain names. These views are only my opinion. thanks

Peace and Productivity

Chad Folkening-President

DigitalPCS.com

317-844-0074

Parker@mibor.net

###

Number: 349

From: "Robert L. Shearing" <robert@priori.net>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 3:04pm

Subject: Priori DNS Comments

Attached are the comments of Priori Networks, Inc. in Docket No.

970613137-7137-01. Your cooperation is filing them is much appreciated.

Filing Instructions

Mail comments in paper form to Patrice Washington, Office of Public Affairs, National Telecommunications and Information Administration (NTIA), Room 4898, 14th St. and Constitution Ave., NW, Washington, DC 20230,

Mail comments in electronic form is dns@ntia.doc.gov. Comments submitted in electronic form should be in WordPerfect, Microsoft Word, or ASCII format. Detailed information about electronic filing is available on the NTIA website, Error! Bookmark not defined..

Paper submissions should include three paper copies and a version on diskette in the formats specified above. To assist reviewers, comments should be numbered and organized in response to questions in accordance with the five sections of this notice (Appropriate Principles, General/Organizational Framework Issues, Creation of New gTLDs, Policies for Registries, and Trademark Issues). Commenters should address each section on a separate page and should indicate at the beginning of their submission to which questions they are responding.

Horzontal lines below represent page breaks.

NB. The filing deadline is 1700 hrs EDT (UTC-4), 18 Aug 1997.

Document available at: http://www.priori.net/NOI

Before the

U.S. DEPARTMENT OF COMMERCE National Telecommunications and Information Administration Washington, DC 20230

In the Matter of )

)

REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01

INTERNET DOMAIN NAMES )

Comments of Priori Networks, Inc.

Robert L. Shearing 2893 El Camino Real Redwood City, CA 94061

President and CEO

August 18, 1997

TABLE OF CONTENTS

Summary

A. Appropriate Principles

Principles a-f

Other principles

B. General/Organizational Framework Issues

Questions 1-9

C. Creation of New gTLDs

Questions 10-14

D. Policies for Registries

Questions 15-20

E. Trademark Issues

Questions 21-28

F. Other Issues

SUMMARY

1. Priori Networks, Inc. (hereinafter "Priori") believes that the existing Internet DNS is politically unstable, and that there is a serious risk of Internet partition at the end of the existing contract between Network Solutions, Inc. ("NSI") and the National Science Foundation at the end of March, 1998.

1. Priori submits that the largest source of that instability is the lack of a democratic self-governing mechanism to administer the "root" of the DNS. The creation of such a democratic self-governing root administration is an urgent priority to the Internet Community as a whole.

1. Priori submits that the second largest source of instability is the existence of a series of artificial barriers to the registration of new TLDs directly with the root administration. These barriers are an accident of history, and are not justified on technical, legal or political grounds. In the future a corporation such as IBM corporation should be able to register the domain "ibm.", and not be forced to choose "ibm.com." or "ibm.xyz."

1. Priori submits that a Constitutional Convention to formally create a new root administration should be held as soon as possible, with participation by all interested individuals and organizations, including representatives of the U.S. Government and other national Governments.

Before the

U.S. DEPARTMENT OF COMMERCE National Telecommunications and Information Administration Washington, DC 20230

In the Matter of )

)

REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01

INTERNET DOMAIN NAMES )

Comments of Priori Networks, Inc.

1. Priori Networks, Inc. respectfully submits comments in this proceeding. Priori Networks, Inc. is a national Internet backbone specializing in selling Internet connectivity to Internet Service Providers throughout the U.S. Like other every other business that depends on the continued growth and vitality of the global Internet, Priori would be hurt severely if the Internet's Domain Name System were to fragment into political chaos.

A. Appropriate Principles

1. An inherent characteristic of the Internet's Domain Name System is that it is hierarchical in nature. This means that, for the system to function effectively, all of the networks that constitute the Internet must at least tacitly accept that there is a single, authoritative root or "." Once below the root, however, there can be any number of domains at each level in the hierarchy.

1. This has two significant consequences for competition. First, competition at the root level is simply impossible. The day that "competition" appears is the day that the interoperability of networks that makes the Internet so useful begins to break down. DNS policy must never forget this fact. It is a huge political challenge to get millions of networks to agree to anything, let alone that there must be one unchallenged authority at the root level of the DNS, but it must be achieved if the Internet is to avoid partition.

1. The second consequence is there is no technical barrier to unlimited competition at every other level of the DNS hierarchy. Although people can argue about what they think is the "best" ratio of the number of domains at one level of the hierarchy to the number at another level, there is nothing inherent in the structure of the DNS or the BIND software that operates it that limits the number of domains at any level other than the root.

1. This point is not well understood by the Internet community, but it has major implications nonetheless. Specifically, there is no necessary reason to limit the number of TLDs, or SLDs, or another LDs. In other words, the assumption that URLs need end in .com or .firm or .us is false. There is no technical or administrative reason that IBM should not have as its domain simply "ibm" so that its web site would be found at www.ibm. Priori Networks could just as easily be www.priori. The requirement to type a ".com" or ".org" or ".us" at the end is an accident of history. Because this particular accident is largely responsible for creating the current legal and political chaos, this would be an appropriate time to change the system.

1. Some have argued that the BIND software would not be able to use caching properly if there were hundreds of thousands or millions of TLDs, but this has never been proven, nor is it an inherent part of the software design. For those who feel strongly about this argument, however, we point out that it would not be difficult to create a "hidden" group of TLDs, one for each letter of the alphabet. Thus, www.ibm would actually be www.ibm.i, where the "i" is the first letter of the SLD and is added automatically by the software (it would not be typed by or visible to the user - he or she would see only www.ibm).

1. Elimination of the artificial limit on TLDs would interfere with a number of business plans that are premised on becoming a "registry" for one of a limited number of TLDs, but it should be done nonetheless. This does not mean, however, that companies should not be free to create new TLDs and attempt to convince others to register in them for a fee. However, logic would dictate that most organizations (and individuals for that matter) would prefer not to add a .com or .xyz at the end of their domain name if there is no requirement that they do so.

1. It follows from this that the concept of "competing registries" is dually flawed. On the one hand, there can be no competition with the administrator of the root that allocates TLDs (because it is the root that must "point" to each and every one of these TLDs). On the other hand, companies that attempt to become a "registry" are not really necessary if companies and individuals are free to apply directly to the root administrator to obtain a new TLD (although it is within the realm of possibility that many or most people would prefer domain names with a .xyz suffix at the end).

1. This does leave room for companies that seek to assist applicants for TLDs in registering with the root administration. Indeed, this is the exact equivalent of what many companies are doing right now in assisting applicants with their applications for .com, .net and .org SLDs from Network Solutions, Inc. (NSI). There is no reason that companies should not be allowed to purchase, for a fee, assistance in dealing with a root registry that is effectively a quasi-governmental agency. Attorneys, lobbyists patent and trademark agents and many other industries have been created by the free market to assist organizations and individuals in interacting with governmental and quasi-governmental authorities.

1. These mechanisms should be developed by the "private sector" with private sector defined as "anyone who seeks to interconnect their network with the networks that constitute the Internet."

1. Input from governments should be as welcome as inputs from any other interested individual or organization.

1. Stability of the DNS is an absolutely critical goal.

1. Misunderstanding of this term "consensus" is responsible for much of the political crisis surrounding domain names. Consensus is absolutely necessary in the sense that every network that wishes to interconnect with the other networks that constitute the Internet must accept the decisions of a single root administration. Consensus is impossible in the sense that it defies human nature to expect all of the diverse users of the Internet to agree on what the appropriate policies of the root administrator should be. Differences of opinion will always exist, and, in order to win the consent of all parties, these differences must be resolved in favor of the will of the majority, determined by an actual vote taken by an authorized and fair method.

1. The mechanism for root administration must be a self-governing one for the simple reason that there is no official government entity on the planet that has jurisdiction over the Internet as a whole. Governments are, of course, users of the Internet, however, and therefore an important part of the necessary self-governing.

1. It is crucial for the root administration to adequate define responsibilities and maintain accountability if the Internet is to avoid fragmenting into various political factions.

c. These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.

1. Because of the inherently global nature of the Internet, governments, no matter how well-intentioned, are not capable of governing the root administration on their own.

1. The key to successful evolution over time is to create a stable framework for the root administration via a written Constitution with provision for Amendment over time. Below is a Draft Constitution for the Agency for Internet Names and Numbers (AINN), one possible name for a self-governing organization that would administer the root.

CONSTITUTION OF THE AGENCY FOR INTERNET NAMES AND NUMBERS

We the People of the Internet Community, in order to promote more complete interoperability of the individual Networks that constitute the Internet, insure harmonious relations between the various Networks that constitute the Internet, and to secure the Blessings of Liberty to all the Networks that constitute the Internet, do ordain and establish this Constitution for the Agency for Internet Names and Numbers (AINN).

ARTICLE I.

SECTION 1. All policymaking Powers granted herein shall be vested in the Membership of the Agency for Internet Names and Numbers.

SECTION 2. The Membership of the AINN shall be composed of those individuals and organizations that pay the required Membership fee and submit a duly-completed application.

The Membership of the AINN will elect a Board of Trustees on an annual basis and will determine the responsibilities of the Board on an annual basis.

SECTION 3. The Membership will vote on major policy issues at regularly scheduled elections held at least twice/year; provided, however, that a special election may be called at any time by a petition signed by at least five (5%) of the AINN Membership.

SECTION 4. The Membership will keep a Journal of its Proceedings, including the results of its Elections, and publish this journal electronically on the Internet.

SECTION 5. The Membership will have Power to determine the prices charged for globally unique identifiers, whether administered by the AINN itself or by contracted outside entities.

To provide for organizing, operating and maintaining whatever registries of Internet globally unique identifiers it determines are necessary to

To make all policies, rules and regulations necessary and proper for carrying into Execution the foregoing Powers, and all other Powers vested by this Constitution in the AINN, or in any Department or Officer thereof.

SECTION 6. No bill of attainder or ex post facto policy, rule or regulation shall be passed.

No Money shall be drawn from the Treasury of the AINN, unless appropriated by the Membership, and a regular Statement and Account of the Receipts and Expenditures of all AINN Money shall be published from time to time.

ARTICLE II

SECTION 1. Management of day-to-day operations of the AINN shall be vested in a General Administrator of the AINN. The General Administrator will be selected by the Membership.

SECTION 2. The General Administrator shall from time to time provide the Membership with Information on the state of the AINN Administration, and recommend measures for their Consideration.

ARTICLE III

SECTION 1. The judicial Power of the AINN shall be vested in one Judicial Review Board, and in such inferior tribunals as the Membership may from time to time ordain and establish. The members of the Judicial Review Board will be elected by the Membership.

SECTION 2. The judicial Power shall extend to all Cases arising under this Constitution and the policies, rules and regulations of the AINN.

ARTICLE IV.

Full Faith and Credit shall be given to the public Acts, Records, and judicial Proceedings of every court of competent jurisdiction in every sovereign nation. The Membership may by general policies, rules and regulations prescribe the Manner in which such Acts, Records and Proceedings shall be proved, and the Effect thereof.

ARTICLE V.

Amendments to this Constitution shall be proposed by a Majority Vote of the Membership or a Majority Vote of the Board of Trustees, and shall be valid as Part of this Constitution when ratified by a two thirds vote of the Membership.

ARTICLE VI.

This Constitution, and the policies, rules and regulations which shall be made in Pursuance thereof, shall be binding on all Networks that choose to be connected to the Internet.

ARTICLE VII.

The Ratification by two thirds of the Delegates to the AINN Constitutional Convention shall be sufficient for the Establishment of this Constitution.

d. The overall framework for accommodating competition should be open, robust, efficient, and fair.

1. Yes. See paragraphs 1-16 above.

1. Yes.

f. A framework should be adopted as quickly as prudent consideration of these issues permits.

1. Yes. The Internet DNS is at great political risk. There is presently only one entity with the administrative capability to administer the entire root. That entity, NSI, is a corporation whose contract to administer the .com, .net. and org domains expires at the end of March, 1998. It is extremely uncertain whether or not that organization, or any other organization, will be accepted by the Internet community as having binding authority over the root administration after that time. If no single organization is accepted as binding authority over the root administration, the Internet will partition, with potentially catastrophic consequences.

1. Support of the U.S. Government for a Constitutional Convention to create a new democratically governed root administration would significantly increase the likelihood that such a Convention would take place and that it would succeed in preventing the DNS from disintegrating into political chaos after March, 1998.

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name registration systems?

1. The principal advantage of the current domain name registration system is that it works. Not perfectly, but well enough to make the Internet tremendously useful to individuals and organizations around the world.

1. The principal disadvantage of the current domain name registration system is that it has fostered tremendous political instability. The corporation that largely administers the root, NSI, has a fiduciary duty to maximize the investment returns of its shareholders, not to maximize the effort expended on behalf of the Internet community or to minimize the costs to the Internet community. This is the source of much hostility toward NSI in the Internet community, even though it is nearly impossible to define what specifically would have been done differently if NSI had been a self-governing organization controlled directly by the Internet Community for the last four years.

2. How might current domain name systems be improved?

1. A new Membership-based self-governing root administration should be created by the Internet Community. (See the proposed Constitution of AINN above.) This entity would exercise binding authority over the root administration. This entity might choose to contract with private entities, including NSI, for the performance of certain functions, but the organization would be responsible for directly setting policy based on the majority will of its members.

3. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

24. See paragraphs 1-23 above.

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

1. Given the sheer scope and scale of the global Internet, by far the most relevant models are the actual governments of the Western democracies. The concepts of a written Constitution, separation of legislative, executive and judicial powers, democratic voting mechanisms, a procedure for Constitutional Amendments - all of these are as crucial to the survival of Internet self-governance as they are to the survival of Western democracies. (Yes, democracy is possible without a written Constitution, but there are number of advantages to having one.)

1. Governing mechanisms appropriate for a small private club or an industry trade group can be used successfully for something as narrowly focused as a network numbering plan or spectrum allocation. They can no longer be used successfully in Internet governance. The "governed" are now so numerous and diverse that obtaining direct consent of the governed is absolutely necessary if the Internet is to avoid disintegrating into political chaos.

1. The AINN would be an international non-governmental organization, and would have the critical role described in paragraphs 1-23 above.

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the .com registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

1. gTLDs should not be forcibly retired from circulation, but it is likely that the majority of Internet users would prefer to go directly to the root for an individual TLD (www.ibm rather than www.ibm.com, see above.)

1. Requiring geographic or country codes would be a mistake in a Multi-National world. Multi-National corporations do not want to be known by 200 seperate TLDs (ibm.co.uk, ibm.co.us, ibm.co.to, etc).

1. The .com registry should be allowed to shrink naturally according to market demand.

31. There is no necessary link between gTLDs and ISO country code domains.

6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

32. See paragraphs 1-23 above.

7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and coordinate?

1. The only way to ensure that root servers continue to interoperate and coordinate is to create a democratic self-governing organization to administer the root (root servers plus their relationships to TLD servers). The most likely way that such an organization could be created is if for the Internet Community to call a Constitutional Convention for the purpose of creating such an organization. See paragraphs 19-32 above.

1. Another possibility would be for NSI to be converted directly into the necessary non-profit self-governing organization. This latter option would have a number of advantages, but would be much more difficult to achieve politically. NSI's shareholders would have to be "bought out" somehow, and determining a price at which such a transaction would take price would be fraught with difficulties. The only realistic possible sources of funds to buy out the NSI shareholders would be future revenues from domain name registration or U.S. government funds. The former involves difficult valuation issues; the latter relies on U.S. taxpayer funds that are not readily available for this purpose.

8. How should the transition to any new systems be accomplished?

1. The new self-governing root administration should be created immediately. The organization would then work closely with NSI in effecting a transition.

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

1. No. The number of TLDs, generic or otherwise, can and should be virtually unlimited, so that most of what are today called SLDs are tomorrow called TLDs.

11. Should additional gTLDs be created?

37. Yes. See paragraphs 1-23 above.

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

1. There are no technical problems with increasing the number of TLDs that are incapable of ready solution. The problems with the DNS are by and large political rather than technical.

13. Are gTLD management issues separable from questions about ISO country code domains?

39. There is no necessary link between the two.

1. Are there any other issues that should be addressed in this area?

N/A

D. Policies for Registries

15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

1. In a DNS without restrictions on TLDs, the notion of "gTLD registrar" would tend to become obsolete. The owner of the domain ibm. (presumably IBM corporation) would itself be a "TLD registrar." Still, some individuals and organizations would prefer to register in other TLDs, including gTLDs administered by others, so the question remains relevant. We suggest that gTLDs be handled in the same manner that all TLDs are handled, on a first-come, first-served basis.

1. Any sharing between "registries" should be on a purely voluntary basis. In a system where there is no restriction on the creation of new TLDs (anyone can apply for and receive any one that is not already taken), there are no competitive issues interfering with the use of exclusive TLDs.

1. There is no reason exclusive and non-exclusive gTLDs could not coexist.

16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have? Who will determine these and how?

43. There should be no barriers to becoming the owner/registrar of a new TLD.

17. Are there technical limitations on the possible number of domain name registrars?

44. No. See paragraphs 4-5 above.

18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

45. None that are not readily solvable. See paragraphs 1-23 above.

19. Should there be a limit on the number of different gTLDs a given registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

1. No and No.

20. Are there any other issues that should be addressed in this area?

1. N/A

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-a-vis domain names?

1. Trademarks should be protected as they have been for hundreds of years, by the parties involved through either private dispute resolution or via the courts. The Internet community does not have jurisdiction over trademark law.

1. The self-governing entity responsible for root administration should behave in the same manner with respect to trademarks as any large corporation would. Would Time Magazine sell a full page ad to a General Motors Cars, Inc., a maker of toy cars? Probably it would decline to do so in absence of proof that no trademark conflict exists with the "real" General Motors car company. Likewise, in "obvious" cases like this one it would be reasonable for the self-governing body to deny registration of the domain generalmotors. to anyone other than the car company.

1. In all non-obvious cases the self-governing body should decline to take responsibility for researching the potential conflict, for notifying the parties to a potential conflict, or for referring one or more parties to "dispute settlement." Instead, it should simply register applicants for TLDs on a first-come, first-served basis, and let any conflicts that develop be resolved between the parties, either privately or through the courts.

1. Aside from a preliminary review process, trademark rights on the Internet should be protected via national courts. The Internet community, and therefore any self-governing organization that it creates, does not have jurisdiction to resolve trademark disputes. Therefore, the self-governing organization should only take a position on a trademark dispute where it believes it is clear that the governing law of the appropriate jurisdiction would clearly require that a domain name registration be denied.

1. There is no appropriate role for national/international governmental/nongovernmental organizations, other than as stated above.

1. By allowing anyone to register any domain in any new TLD. See paragraphs 1-23 above. In this system, speculators intending to register domains (i.e. gucci. or kodak.) solely for the purpose of extracting settlement payments from corporations eager to avoid the expense of litigation would stand out like a sore thumb. These registrations could easily be denied as obvious cases by the root administration, and, if they inadvertently slipped through the self-governing organization, could be reversed easily in national courts via summary proceedings rather than prolonged litigation.

1. No basis should be required or even requested unless the officials of the self-governing organization believe that the registration would represent an obvious trademark infringement. If there is any ambiguity at all in the context of the application, the registration should be granted and the burden of enforcement placed on the trademark holder, where it traditionally lies. Any other policy is doomed to swamp the root administration in the resolution of trademark disputes.

26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

1. Allowing anyone to register any TLD absent an obvious trademark violation would drive the cost of resolving trademark disputes to the cost of resolving trademark disputes in almost any other context. The existing artificial shortage of TLDs is creating large numbers of unnecessary trademark conflicts.

27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

1. No. Trademark issues are legal and political rather than technical.

28. Are there any other issues that should be addressed in this area?

N/A

F. Other Issues

N/A

Annex 1 Service List

*********************************************************

Robert L. Shearing voice: +1-650-482-2840

President/CEO fax: +1-650-482-2844

PRIORI NETWORKS, INC. http://www.priori.net

"The People You Know. The People You Trust."

*********************************************************

###

Number: 350

From: Tony Rutkowski <amr@netmagic.com>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 3:18pm

Subject: DNS NOI Comments

The attached comments are respectfully submitted by

Anthony M. Rutkowski.

(signed)

Anthony M. Rutkowski

13102 Weather Vane Way

Herndon VA 20171

tel: +1 703.437.9236

Document available at: <http://www.wia.org/pub/amr-dnsnoi-comments.html>

Before the

U.S. DEPARTMENT OF COMMERCE

National Telecommunications and Information Administration

Washington, DC 20230

In the Matter of )

)

REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01

INTERNET DOMAIN NAMES )

Comments of Anthony M. Rutkowski

[signed by]

Anthony M. Rutkowski

13102 Weather Vane Way

Herndon VA 20171

USA

18 Aug 1997

mailto: amr@netmagic.com

tel: +1 703.437.9236

fax: +1 703.471.0596

TABLE OF CONTENTS

Summary

A. Appropriate Principles

Principles a-f

Other principles

B. General/Organizational Framework Issues

Questions 1-9

C. Creation of New gTLDs

Questions 10-14

D. Policies for Registries

Questions 15-20

E. Trademark Issues

Questions 21-28

SUMMARY

The subject matter of this proceeding is a key component among several fundamental transitions in the evolution and scaling of the Internet and its applications. The creation, development, operation, and administration of the Internet over the past 20 years has - until only the past few years - occurred under the auspices of the United States Federal Government. The responsibility and overall management of Internet identifiers including generic domain names has remained under Federal auspices. In a fashion similar to the well-considered transition of ownership and control of the Internet backbone to a stable private-sector arrangement three years ago, it is now appropriate through this proceeding to accomplish the same for the administration of Internet identifiers, including domain names.

The existing arrangements have been and remain effective and satisfactory for the vast preponderance of Internet providers and users. Generic domain names are used and of interest almost entirely in the United States especially for branding and marketing. There is no need to act urgently. There is, however, a need: to understand much better what users want; to understand what is technically and operationally feasible; to devise appropriate policy objectives; to allow some experimentation to occur in providing root, root-server, and registration services, to avoid attempts by various secondary groups especially intergovernmental organizations - to claim "governance" authority; and to generally put in place the lawful and effective processes, arrangements, and models by which the private-sector can undertake the policy making and administrative responsibilities long undertaken by the federal government and its contractors.

This situation is similar to the same kind of transition effected more than a decade ago in conjunction with the implementation of the A&T Consent Decree and the creation of arrangements now known as the North American Numbering Plan Administration (NANPA). Similar open, industry-driven institutional and process arrangements with antitrust liability protection and appellate review provided by an appropriate federal government agency. Because some identifier services like those supporting DNS are more akin to directory services rather than physical number assignments, there is also likely greater flexibility in allowing a relatively unfettered market to occur.

Continuing flexibility and responsiveness are particularly significant for the Internet arena where technologies and applications are changing so dramatically. For example, the existing domain name system may need to evolve significantly to support the use of mobile software agents or new electronic commerce implementations.

The least desirable outcome is one where one particular set of parties unilaterally assumes dominant control by sheer assertion, where intergovernmental organization responsibilities and processes are put in place by formal agreements, or where top-down institutional arrangements become effected that are closed, sheltered, and unresponsive to real stakeholders and users.

The diverse Federal agencies instrumental in conducting this proceeding are to be commended for doing so, thereby providing a critically important open public forum for gathering views on the important questions and issues identified. Going forward, those same agencies not only have a unique opportunity, but an important obligation to take the necessary subsequent steps to put into place a NANPA like mechanism that brings industry and users together on a continuing basis to openly develop the information, criteria, policies, and administrative practices for domain names and other Internet identifiers.

Before the

U.S. DEPARTMENT OF COMMERCE

National Telecommunications and Information Administration

Washington, DC 20230

In the Matter of )

)

REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01

INTERNET DOMAIN NAMES )

Comments of Anthony M. Rutkowski

1. These comments are respectfully submitted by Anthony M. Rutkowski. The qualifications of Mr. Rutkowski are well-known and a matter of public record. He possesses degrees in both Electrical Engineering and Law. Over the past thirty years, he has held many positions which dealt with many of the issues raised in the inquiry, including Chief of International Telecommunication Regulations and Relations Between Members at the International Telecommunication Union, FCC Engineer, Adjunct Professor at New York Law School (teaching the graduate program in international telecommunications law), Research Associate at MIT, and industry positions with General Magic, Sprint, Horizon House, Pan American Engineering, and General Electric. He has enjoyed leadership positions in many diverse Internet, WWW, legal, and telecommunication organizations, has appeared on several occasions as an expert witness before committees of the U.S. Congress, and authored scores of related published articles and several books. He is presently Vice President of Internet Business Development at General Magic, Inc, of Sunnyvale, California, but filing in a personal capacity, and not necessarily representing its views. See <http://www.chaos.com/rutkowski.html>

A. Appropriate Principles

2. This principal in its present form is too abstract and general to be useful. The principal subject matter of the inquiry happens to be one instantiation of an array of existing and possible name systems, products, and services that exist, are specified, or may evolve for use with applications running across the distributed, heterogeneous private network of networks commonly referred to as the Internet. Within each of these instantiations including the present predominant one - there are a multitude of possible architectures and implementations that may or may not in whole or in part be capable of or benefit from expansion and competition in providing the services and products. From a public policy standpoint, it is not clear why such Internet related name system services and products should be treated any differently than another enhanced, private networks and services.

3. Along similar lines, it is intrinsically difficult to specify what the "Internet," "interoperation," or "connecting to" are, much less specify them as principles. Furthermore, name systems support applications running across the Internet, not the Internet, qua Internet. In practice, it is users, industry, and the marketplace that define these attributes.

4. As Internet pioneer Einar Stefferud has admonished in public statements, the sole technical operational requirement is "coherency" as part of a general desire to further widespread connectivity. There is no other technical operational requirement that can or should serve as a public policy principle. Thus, it is apparent that there are significant antitrust implications underlying this proceeding which needs to be recognized and openly discussed . Private organizations on their own cannot establish and attempt to implement other criteria for who can or cannot enter that marketplace, including attendant enforcement mechanisms. Doing so creates an antitrust conspiracy.

5. This is a desirable principle, although it is unclear what purpose "input from governments serves." With perhaps a single exception, the desirable roles of government in this matter are largely procedural, and associated with: 1) assuring lawful and appropriate transition from existing legacy implementations, 2) stable, open, public, industry-driven processes and institutional mechanisms for achieving consensus on policies and administration, 3) keeping other government and intergovernmental organizations from encroaching into this arena, and 4) providing antitrust guidance, liability protection and associated appellate relief. In light of para. 4., above, the antitrust-related role assumes substantial, inescapable significance. This is not a matter that Federal government can simply abandon to the private-sector. If there are public policy reasons for limiting a freely open market in domain name services and products, it is the government that must determine and implement those policies.

c. These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.

6. This is a desirable principle. However, "inherently global" should not necessarily imply only a conventional international connotation. What we are dealing with here are unique private identifiers for use among private information systems. These identifiers sometimes in the form of language expressions constituting names. There are many different ontologies and communities of interest potentially involved and it is this breadth of scope to which "global" should refer.

d. The overall framework for accommodating competition should be open, robust, efficient, and fair.

7. This seems like a desirable principle, although the quality of "nondiscriminatory" should be included. This non-discriminatory treatment should apply worldwide for the processes employed not only to users of DNS services and products, but also to providers as they seek to enter markets. In addition, the framework itself should be consistently applied for all similar instantiations.

8. This seems like a desirable principle, although the notion of "should promote" seems too narrow and should be amended to allow for multiple external means of conflict resolution such as existing administrative and judicial mechanisms. The number of conflicts in this arena in practice have been comparatively miniscule, and existing judicial processes have proven comparatively very effective. Conversely, the notion that entirely new conflict management mechanisms need to be established especially mandatory, far-reaching, new intergovernmental ones when most of the conflicts are likely to arise in the U.S. seems particularly inappropriate if not ludicrous.

f. A framework should be adopted as quickly as prudent consideration of these issues permits.

9. This principle is inappropriate and unneeded. Although unnecessary delay should not occur, there is no urgency of any kind associated with his matter other than staying actions by some parties especially intergovernmental organizations - to acquire unlawfully and inappropriately the responsibilities and authority in this matter presently effected by the Federal government and its contractors. This is a wholly artificial exigency, and it is far more important to effect a transition to a stable, accountable, responsive, and effective industry framework that will establish significant precedents, serve as a governance model, that everyone must live with for some significant number of years.

Other principles and related issues related to domain names.

10. There are other principles that deserve recognition. These include the private resource character of domain names, related First Amendment expression, and the stakeholders affected.

11. Some parties in the ongoing discourse on domain names especially some intergovernmental organizations have sought to characterize domain names as a radio spectrum like "limited global public resource" for purposes of assuming jurisdiction and effecting traditional international managed cartel governance models. The issue is whether domain names are treated a "public resource" to be administered by an ITU-type bureaucracy or conversely as globally unique, registered private name expressions managed by private entities. Denominating domain names as a global public resource seems at best a perversion that is orthoginal to reality, and a poor public policy making choice that has significant permanent administrative implications. Thus, an additional principle should be adopted that asserts "domain names are globally unique private expressions that describe an open network based information object."

12. Domain names in the final analysis are language expressions. Increasingly, they are being significantly used for marketing products, services, and ideas. Yet no free expression related principles were included in the Notice of this proceeding. At the international level, this is a growing issue that some institutions and governments deign to control with an apparent intent to prevent the use of domain names that may be undesirable or offensive to some peoples or groups. A principle relating to domain name expressions both personal and commercial should be adopted - invoking instruments such as the First Amendment in the US, and Human Rights Declaration and World Trade Organization agreements internationally.

13. Significant economic investments are at stake here - by existing registrants, by existing registrars, by other affected parties. In the real business world, countless millions of dollars have been invested in establishing and using names within markets. In the United States in particular, generic domain names have become an exceedingly important vehicle for establishing and maintaining recognition for corporate and organizational brands, products, and services. It is inescapable and pervasive via every medium from television, radio, and printed publications to billboards and cereal boxes. Some argue that branding is indeed the most significant issue going forward, and that given a market based commerce approach, the subject deserves significant attention in the form of a principle.

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name registration systems?

14. Existing domain name services and products which include registration services, root-server services, and DNS software function well, have found enormous public acceptance, and have produced comparatively miniscule disputes and problems worldwide. This is supported not only by the facts, but also by statements by business leaders and public officials in the U.S. and other regions and countries. Indeed, the primary expressed concerns deal with the instabilities and changes being unilaterally pursued by several parties and intergovernmental organizations purporting to have "governance" authority.

15. However, the existing scheme just grew by default with very little thought or consideration - probably the inevitable result of closed processes attendant to long-standing government management and contractor practices established for the Internet as a component of the Defense Data Network (DDN). Not only have there been no public processes for determining administrative policies and making decisions, there were no public procedures or decision making records. Although this may have been perfectly acceptable and effective in the past, it seems generally agreed that this is not a regime appropriate for large-scale Internet implementation in a commercial marketplace.

16. In addition, grant of monopoly registration franchises to entities in other countries for 2-letter domains frequently government agencies or entities wholly under the control of government agencies with little or no private-sector role seems wholly inappropriate for a private Internet infrastructure. Even in the United States, the DOD DNS root management contractor in conjunction with a National Science Foundation project, assigned the responsibility to a closed arrangement operated by the University of Southern California. Because this proceeding involves Federal government in transitioning its existing responsibilities to future arrangements worldwide, it should playing a proactive role in establishing or shaping open, industry-oriented arrangements for top level domain policies and administration.

17. Even the technical operational arena there are major potential disadvantages to the existing domain name system. For example, there are a small group of TLDs that have organizational connotations, and a much larger group that have geopolitical connotations. By contrast, various agent naming standards groups have given the matter rigorous consideration and developed a system of flexible naming ontologies. See, e.g., Agent Management, FIPA'97 Draft Specification at http://drogo.cselt.stet.it/fipa/spec/17611.htm.

2. How might current domain name systems be improved?

18. As noted above, the improvements involve many elements: transitions in responsibilities away from the Federal government but dealing with attendant antitrust issues; coupled with the establishment of go-forward, open, user-oriented process and institutional models for the system worldwide. In the short-term if the aggressive related activities of intergovernmental organizations could be abated - the existing arrangements could readily be continued, together with a combination of technical study and entrepreneurial experimentation.

19. The agent environment is potentially much more complex and varied than the relatively simple one dimensional existing DNS environment. Since we're now at a fairly significant point of establishing future directions, policymakers, industry, and users should use the occasion to take a somewhat broader, harder look at what purpose the DNS will be serving in the future, and what the ontology (i.e., structure of naming) should be for that future? (The term "ontology" is used extensively within computer science, knowledge exchange (esp. KQML), object, and software agent communities with a connotation and purpose somewhat similar to "taxonomy" in the biology community. It's a useful concept and methodology because it recognizes that distributed knowledge structures across Internet like networks are flexible and autonomous, and that some threshold step of understanding and agreeing on an ontology is necessary for meaningful communication. The existing Internet DNS is somewhat application- specific and uni-dimensional - dealing only with places, and presently has only three ill-defined ontologies: 1) territories (e.g., cnri.reston.va.us), 2) type of sponsoring institution (e.g., fcc.gov), or a hybrid of the two (e.g., dillons.co.uk). As such, it's not suitable for more advanced applications that are beginning to emerge where there are highly distributed arrays of mobile and stationary objects that need to have assigned names. Naming systems for these kind of applications and systems must necessarily allow for diverse and flexible ontologies. It's worth reflecting on why anyone would seek permanent intergovernmental agreements and global administrative structures constructed around a naming system serendipitously created 12 years ago for an Internet that consisted of a DDN operation of 2000 hosts with only a handful of simple applications.)

20. As part of this activity, continued treatment of 2-letter geopolitical identifiers as ISO codes is an open invitation for never ending trouble and formal intrusion of national authorities. Because the Internet is basically a distributed information system, the identifiers aren't really tied to any physical notion of geographical exercise of sovereignty. Along these lines, it seems desirable to have multiple geopolitical identifiers.

3. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

21. Domain name services and products of all types should be provided by private-sector entities except where the domain itself may be the government. The overall responsibilities for establishing and implementing the ground rules should occur through an open, industry-driven body similar to the NANPA.

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

22. As noted above, a NANPA-like model seems appropriate. The nominal sponsoring entity for antitrust and public policy purposes could similarly be a Federal agency. It is not necessary and should not be an intergovernmental organization. The Internet and its applications are private as opposed to public systems, and thus outside the scope of traditional intergovernmental organization management responsibilities.

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the .com registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

23. There is widespread, large-scale acceptance of the use of generic top level domains, and the need for such a change is not evident. Furthermore the economic consequences would be substantial. Lastly, it is not clear on what basis such a change could be mandated. The evolution of the Internet and its applications will require more flexible DNS ontologies, not a more restrictive, less flexible one.

24. As noted above, it may well be desirable to establish in the U.S. and facilitate abroad, a number of alternative country-oriented domains. The history of past .US management and control was instrumental in the paucity of .US registrations. That problem can be corrected in several ways, and the focus should be in that direction rather than pursuing restrictive alternatives.

6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

25. Whatever issues have arisen are non-technical in nature. Within reason, the technology can be applied to support a wide variety of architectures, policies and implementations. Root and TLD administration is bound to root server administration, and effective arrangements between the various providers will be necessary. The same situation exists with respect to the vendors of DNS software who presently provide the default tables that are presently linked to the root servers. This area of implementation and de facto enforcement of policies also raises significant antitrust concerns and potential liabilities. These two categories of providers root-server services and DNS software vendors effectively control the marketplace for services under existing arrangements. This has not been an issue in the past because the overall responsibilities have been born by the Federal government, but will be so in a substantially private-sector oriented DNS provisioning environment.

7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and coordinate?

26. It is not apparent that scalability of the system is a significant issue. It is not clear what is meant by "address spaces." As noted in para 25., above, interoperation and coordination is necessary. However, this can be entrusted to a NANPA like organization.

8. How should the transition to any new systems be accomplished?

27. These are responsibilities being transferred from Federal government and contractor auspices to new arrangements. The transition is very similar to that paced ten years ago with the implementation of the AT&T Consent Decree. Virtually identical processes can be used involving a combination of proceedings, corporation formation, and establishment of the basic model and processes.

9. Are there any other issues that should be addressed in this area?

28. By the force of events, an issue has arisen regarding the role of United Nations. intergovernmental treaty organizations, including so-called depository functions which are in fact highly proactive, open-ended, and jurisdiction setting . Fundamentally, it is plain that any involvement whatsoever of such U.N. organizations in the provision of private network information service operations such as the Internet is not only unneeded and inappropriate at this time, but also constitute a very undesirable precedent. The involvement that has occurred over the past year arose from unapproved, aggressive initiatives of permanent secretariats which itself is a very undesirable precedent.

29. Tolerance of this conduct, much less acceptance of such roles, poses substantial risks for U.N. system related policies, for Electronic Commerce policies and initiatives, and for the evolution of the Internet and its applications. Tolerance and acceptance will also encourage further such behavior. The effects will be permanent and long-term raising the cost of U.N. agency contributions, and entwining Internet users and providers in a maze of new U.N. organization driven legal, policy making and administrative activities over which it will have no control and likely adverse outcomes. This dynamic substantially endangers the ability for an industry-driven model to be put into place not only for DNS services, but potentially an array of other Internet related services as well. These concerns have already been recognized and initial abating actions taken by the Secretary of State. As part of this proceeding, appropriate Federal government agencies should take necessary multilateral, bilateral, and unilateral steps to halt these U.N. specialized agency activities.

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

11. Should additional gTLDs be created?

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

13. Are gTLD management issues separable from questions about ISO country code domains?

14. Are there any other issues that should be addressed in this area?

30. There are no apparent technical considerations that are incapable of resolution other than the single requirement for coherency. There are, however, an undefined array of practical, market, and policy considerations that need exploration and resolution in due course by stakeholders and other affected parties in an open NANPA-like mechanism, and difficult at best to be determined in this proceeding even if it is useful to raise them for the industry.

31. In information system architectural terms, this is a process of deciding ontologies (i.e., structure) and their branches. It involves determining weight given technical, practical, policy, business considerations, including the tradeoffs and tensions among them. Part of this involves determining what exactly are the resulting goals being sought. These are rarely articulated. Is it stability, security, low costs, flexibility, administrative speed? Or is the goal simply an open architecture for distributed naming services. It might be worth considering in the context of this proceeding, exactly what are the public policy or business bases for making these choices. The Internet as a critical infrastructure may modify these choices.

32.. The name "generic" has been pinned onto the current 3-letter TLDs in recent years. From an ontology perspective, however, these are really entrepreneurial-type TLDs preponderantly favored by US entities for marketing purposes. See Generic TLD Trends, Appendix 2. The entire matter might be usefully dealt with as multiple ontologies that could very well have different purposes, models, regimes, and institutions. In this context, an "entrepreneurial-type" ontology (.COM, .ORG, etc) are the least appropriate for binding to complex international regimes that would be endlessly becoming nvolved in matters that primarily relate to information services and branding for US business and marketing.

D. Policies for Registries

15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have? Who will determine these and how?

17. Are there technical limitations on the possible number of domain name registrars?

18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

19. Should there be a limit on the number of different gTLDs a given registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

20. Are there any other issues that should be addressed in this area?

33. These questions largely parallel and amplify the issues raised and discussed in Section C, above. They are part of an array of practical, market, and policy considerations that need exploration and resolution in due course by stakeholders and other affected parties in an open NANPA-like mechanism, and difficult at best to be determined in this proceeding even if it is useful to raise them for the industry.

34. It is important to examine these questions for all gTLDs. There are vast array of possible DNS ontologies, and these policies may well be different for different DNS ontology classes. For example, some country-oriented classes may have one set of policies, while quite different policies might apply to other classes particularly highly desirable marketing and branding classes. It is not apparent why "one size fits all" policies need to be maintained especially considering the very heterogeneous, distributed nature of the Internet and its applications.

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-a-vis domain names?

23. Aside from a preliminary review process, how should trademark rights be protected on the Internet vis-a-vis domain names? What entity(ies), if any, should resolve disputes? Are national courts the only appropriate forum for such disputes? Specifically, is there a role for national / international governmental / nongovernmental organizations?

26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

28. Are there any other issues that should be addressed in this area?

35. Given the existing facts that include 1) more than one million domain names being registered, 2) a few hundred known disputes arising at most, 3) less than two dozen litigated disputes, and 4) virtually all generic TLD use and disputes occurring in the United States, it is appears that the existing arrangements work well and remain efficient and satisfactory. Furthermore, any intractable disputes arising will almost certainly be resolved under national law by national judicial systems. Under these circumstances, any new international regimes are patently unneeded, unwarranted, comparatively highly inefficient, and likely to exacerbate an already complex conflict of laws environment.

36. It appears, however, that a consensus on trademark-related procedures attendant to existing registration activities among trademark law practitioners would be useful and could be attained through a NANPA-like mechanism.

Appendix 1

Internet DNS Historical Timeline

OFFICIAL ACTIONS AND STATISTICS

1985

Q1 DCA/DARPA implement RFCs 920,921, tasks USC-ISI for DNS root management, SRI International for DNS NIC registrations

Mar SYMBOLICS.COM is first registered domain - 15 Mar 1985

Apr CMU.EDU, PURDUE.EDU, RICE.EDU, UCLA.EDU are first registered .EDU domains

Jun CSS.GOV is first registered .GOV domain

Jul MITRE.ORG is first registered .ORG domain

Jul UK is first registered 2-letter domain

Oct 5 .COM, 1 .ORG, 0 .NET, 17 .EDU, 1 .GOV, 2 .MIL, 1 .UK, 1782 .ARPA domains registered

Oct Internet at 2k hosts (Lottor)

1986

Nov Internet at 5k hosts

1987

Nov Mockapetris writes RFC1034 (Domain Name Concepts and Facilities) and RFC 1035 (Implementation and Specification)

Dec Internet at 28k hosts (Lottor)

1988

Oct Internet at 56k hosts (Lottor)

1989

Oct Internet at 159k hosts (Lottor)

1990

Oct Internet at 313k hosts (Lottor)

- DISA rebids DDN NIC contract

1991

- DISA awards DDN NIC contract to Government Systems Inc.

May DDN NIC moves from SRI International to GSI

Oct Internet at 617k hosts (Lottor)

Oct 165 .COM, 22 .ORG, 20 .NET, 285 .EDU, 29 .GOV, 20 .MIL, 1 .AT, 29 .AU, 20 .CA, 2 .CH, 15 .DE, 3 .DK, 1 .IL, 2 .IT, 1 .JP, 1 .KR, 1 .MX, 11 .NL, 2 .NO, 2 .NZ, 2 .PR, 9 .SE, 5 .UK, 3 .US domains registered

1992

Jan 31 two-letter domains in use (Lottor)

Q1 DISA allocates .COM, .ORG, .NET, .EDU. and .GOV to NSF

Mar NSF issues RFP for .COM, .ORG, .NET, .EDU. and .GOV services

Oct NSF awards contract to Network Solutions Inc

Oct Internet at 1136k hosts (Lottor)

1993

Jan NSF 5 yr contract to Network Solutions Inc begins for .COM, .ORG, .NET, .EDU. and .GOV services

Jan 51 two-letter domains in use (Lottor)

Mar 0.6k COM, 1.1k EDU, 0.5k ORG, 0.2k NET, 103 two-letter domains

Oct Internet at 2056k hosts (Lottor)

1994

Jan 11k COM, 1.4k EDU, 1.1k ORG, 0.4k NET, 115 two-letter domains

Jan 61 two-letter domains in use (Lottor)

Oct Internet at 3864k hosts (Lottor)

1995

Jan 30k COM, 2k EDU, 3k ORG, 2k NET 140 two-letter domains

Jan 87 two-letter domains in use (Lottor)

Apr Internet backbone network transitioned to private-sector

Jul Internet at 6642k hosts (Lottor)

Sep InterNIC announces domain name services charges

1996

Jan 133 two-letter domains in use (Lottor)

Feb 232k COM, 2k EDU, 17k ORG, 11k NET, 168 two-letter domains

Jul Internet at 12881k hosts (Lottor)

1997

Jan 796k COM, 3k EDU, 53k ORG, 44k NET, 197 two-letter domains

Jan 177 two-letter domains in use (Lottor)

May 1222k COM, 4k EDU, 84k ORG, 86k NET, 215 two-letter domains

Jul US Federal DNS Inquiry Proceeding published in Federal Register

Jul Internet at 22000k hosts (estimated)

Aug US Federal DNS Inquiry Proceeding deadline for comments

Other meetings and actions

1996

Jan DNS Workshop, Washington DC

Jan ISOC issues initial plan for assuming DNS responsibilities

Jun OECD workshop on domain names

Jul USC staff for DOD-IANA contract issue plan for new domains

Jul Newdom listserve created

Aug USC staff indicate acceptance of new registry applications

Aug Image Online application allegedly accepted

Sep Harvard IIP Conference

Sep Operational alternative registries appear

Oct ACICP meetings on domain name policy

Oct FNCAC advises transfer of registry functions to private sector

Dec ISOC creates IAHC with INTA, ITU, and WIPO

1997

Feb ISOC IAHC committee issues report asserting DNS powers

Feb Image Online files suit in California court

Mar PGMedia files suit in NY Federal court

Mar ISOC IAHC committee issues MoU

Mar First eDNS meeting, Atlanta

Mar European Commission issues demarché on MoU

Apr DNS Brief filed with federal agencies by Rutkowski

Apr US Dept of State sends demarché expressing concerns to ITU Secretary-General

May ISOC-ITU MoU signing ceremony, Geneva

May Image Online hearing on TRO, suit withdrawn

Jun ITU Council considers General Secretariat MoU actions, begins 60-day comment period

Q2 Multiple DNS confederations emerge

Jul Australian root-server confederation commences

Jul Open Internet Congress meeting on DNS

Jul ITAA DNS workshop

Aug Meeting to select first Australia TLD for AURSC and activation of the TLD

Appendix 2

Generic TLD Trends

In absolute numbers or even total percentages, the number of gTLDs in use outside the USA is very small and comparatively shrinking in size. In Jan 1996, the statistics were:

GTLDs Used Outside the U.S.

Percent Count

..COM domains 15% 10,200

..COM hosts 6% 150,000

..ORG domains 13% 900

..ORG hosts 7% 19,000

..NET domains 43% 2,200

..NET hosts 17% 126,000

.NET top level domains tend to be used by Internet Service Providers outside the U.S.

Of these, 5-7 percent are in Canada, 1-2 percent in the UK, and the rest are only a few hundredths of a percent at the most in other countries. It is likely that the percentages for non-USA use of these domains will continue to drop because they preponderantly used for marketing and branding purposes in the United States. The rest of the world appears to prefer country-oriented domain names.

###

Number: 351

From: greg detardo <detardo@nutechsecurity.com>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 3:19pm

Subject: detardo@.travel

To Whom It May Concern:

As operators of ".travel" & "travelnic.net", we would like to be included on

any correspondence or committees that involve the reorganization or future

ownership questions as they may inhibit or cause hardship on our ongoing

business as the registry for ".travel". We have spent thousands of dollars

creating our registry and publicizing, etc. We feel any restrictions on

current TLD's would create an unfair marketplace since those running top

level domains have been the forefront of the internet and are innovators

providing additional alternatives to ultimate end users. We would like to

be included in any decisions that your commitee produces in reference to

this topic and would be available for any panel discussions or questions on

how changes would affect our business.

I can be reached at 407-628-1600.

Sincerely,

Greg S. DeTardo

###

Number: 352

From: Don Heath <heath@isoc.org>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 3:21pm

Subject: Internet Society Response to RFC - Docket No. 970613137-7137-01

Please use the subject line above, that reflects "RFC," instead of

"NOI," which was used in transmitting the response from ISOC.

The document previously transmitted is correct, just the subject line

of the transmittal email was incorrect.

Don

In the matter of REGISTRATION AND ADMINISTRATION OF INTERNET DOMAIN

NAMES, Docket No. 970613137-7137-01, the Internet Society is pleased

to submit its comments as presented in the attached Microsoft Word

document.

We thank you for the opportunity to offer these comments.

Sincerely,

Donald M. Heath

President/CEO

Internet Society

CC: NTIADC40.SMTP40("isoc-trustees@isoc.org")

###

Number: 353

From: Shari Steele <ssteele@eff.org>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 3:27pm

Subject: EFF's DNS comments

This was sent via Fed-X, as well.

Shari Steele, Staff Attorney

Electronic Frontier Foundation

1550 Bryant Street, Suite 725

San Francisco, CA 94103

August 18, 1997

Patrice Washington

Office of Public Affairs

National Telecommunications and Information Administration (NTIA)

Room 4898

14th Street and Constitution Avenue, NW

Washington, DC 20230

Dear Ms. Washington,

The Electronic Frontier Foundation (EFF) is a nonprofit, public-interest

organization working to protect rights and promote responsibility in the

electronic world. We thank you for giving us the opportunity to submit

our comments on the current and future systems for the registration of

Internet domain names.

Appropriate Principles

EFF believes the Internet domain name system is a crucial public

resource caught between the triple perils of monopolization,

inappropriate government involvement, and unclear authority. Each of

these problems leads down the path of increased risk of disruption of

service to the entire Internet and persistent conflicts among its users.

The Electronic Frontier Foundation advances the following long-term

public-interest policy principles for the Internet domain name system.

Nothing in these principles should be taken as an endorsement of

government control over resource allocation in the Internet domain name

system. "Public" here means public in the sense of the air we breathe,

not in the sense of a "public" street.

1. The Internet top level domain name space, an essential component of

the operational infrastructure of the Internet, is a global public

resource and must be developed and maintained for the public good.

2. Any administration, use and/or evolution of the Internet top level

domain name space is a public policy issue and must be carried out in

the interests and service of the public, and with the broadest

international public input and participation.

3. Such public policy should balance the interests of current and future

stakeholders in the domain name space to the extent possible.

4. The public, including domain name space stakeholders, can benefit

most from a self-regulatory and market-oriented approach to domain name

registration, administration and growth.

5. Domain name registration and the generic top level domains themselves

must not be monopolized by a single registry, treated as any entity's

intellectual property, or controlled by or from within any single

governmental jurisdiction.

6. Registration services for the generic top level domain name space

should provide for global geographical distribution of registrars.

7. The administration of the domain name system and generic top level

domain registries must provide for domain name portability rather than

making it difficult or impossible for a customer to switch registries.

8. Use of and access to the domain name system must open and flexible,

with participation in the system available to any capable registrar and

domain names available to any user, on a non-discriminatory basis.

9. Internet technical standards and administrative processes should not

be impeded by lack of consensus on separate issues such as international

trademarks.

This EFF position recognizes the Internet Ad-Hoc Committee/Internet

Society/Internet Assigned Numbers Authority "Memorandum of Understanding

on the Generic Top Level Domain Name Space" (gTLD-MoU). This EFF

position goes beyond the principles expressed in the gTLD-MoU and aims

to establish baseline public-interest principles for long-term stability

and equity of the Internet Domain Name System. The EFF position also

rejects the MoU's "principle" that the interests of one group of

intellectual property holders should outweigh any other group's, or

outweigh any other rights held by individuals and the public. EFF

generally supports this MoU, with these reservations, but is not a

signatory to it.

General/Organizational Framework Issues

EFF is very concerned about the current domain name registration system.

It is a monopoly which has been given to Network Solutions, Inc., a

for-profit company attempting to make as much money as possible and not

acting in the public interest. EFF believes that the National Science

Foundation made a mistake in choosing a for-profit company to manage

such a vital public resource. Now we are afraid this monopoly is going

to take away important public property when its contract expires. In

order to protect the public, the United States government should enforce

the Cooperative Agreement between Network Solutions and the National

Science Foundation and should:

1. Make sure that the public retains ownership in the contents of the

full database of the .com domain.

2. Make sure that the public retains all of the software that Network

Solutions wrote under that government-funded and eventually user-funded

contract. The Cooperative Agreeement specifically requires Network

Solutions to provide a copy of all software code it created at the

termination of its relationship with the National Science Foundation.

Please make sure that this happens.

3. Make sure that people will continue to be able to register in the

.com domain without any involvement of Network Solutions, Inc.

4. Make sure that the $30 million collected for the preservation and

enhancement of the Intellectual Infrastructure of the Internet gets

disbursed to the public. Do not permit Network Solutions to take any of

this money for any reason.

After these four conditions are met, both Network Solutions and all

branches of the United States government should defer involvement in the

domain naming process to the people actually using the Internet, whether

in the form of the Internet Ad-Hoc Committee/Internet Society/Internet

Assigned Numbers Authority or whichever other non-governmental body that

gains the most democratic support from Internet users.

Trademark Issues

The current policy of providing ownership rights in Internet domain

names based on trademark registration is flawed for the following

reasons and should be abandoned:

1. Even the United States Patents and Trademarks Office (PTO) recognizes

that more than one entity can hold a trademark on the same word,

provided that the word is used in completely different business

categories. Failure to provide for multiple uses of the same

trademarked term have resulted in big companies beating up on small

companies using the same terms.

2. Even the PTO recognizes that registration of a trademark creates a

right in the term while the trademark is still pending. The Internet

policy does not recognized registered but still pending trademarks, and

small companies with pending applications have been bullied into giving

up their domain names by larger companies with similar trademarks.

3. There are other legitimate uses of words that have nothing to do with

trademarks that get and deserve legal protection yet are not recognized

by the current policy. For example, Kayvan Sylvan, a man who runs his

own computer consulting business, registered the name sylvan.com.

Sylvan Learning Systems, which holds a trademark on Mr. Sylvan*s last

name, threatened to take that domain name under the current policy. Mr.

Sylvan had to trademark his last name in a foreign country in order to

keep his domain name! There are other examples of nontrademarked, yet

legal, uses of terms. For example, the World Boxing Association might

want to use the domain name knockout.com, even though Hasbro toys has

trademarked the term knockout as the name of a game. This should be a

legally protected use of a domain name.

Many of these trademark disputes can be avoided with the creation of

several additional top level domains. The current policy of taking

domain names away and giving them to those who have the first trademark

gives an unfair advantage to large corporations that have been existence

for a long time, to the detriment of smaller companies who were on the

Internet first. The current policy is flawed and must be repaired.

Thank you again for giving us the opportunity to comment as you work to

formulate United States government policy in this important area. We

would be happy to meet with you and work to create a system that is in

the public interest. Please contact me at 301/375-8856 if I can be of

any further assistance.

Sincerely,

Shari Steele

Staff Attorney

Electronic Frontier Foundation

CC: NTIADC40.SMTP40("eff-board@eff.org","eff-staff@eff...

###

Number: 354

From: Bob Jakeway <Bob.Jakeway@postoffice.worldnet.att.net>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/15/97 6:57pm

Subject: domain comments

RE: Internet domain-

I agree with the concept that the government has no overriding need to become involved with this isssue. I do agree, however, with comments I read in the New York Times regarding ".XXX". This seems a legitimate method to help parents restrict access to sites they do not want their children to access.

Robert B. Jakeway

###

Number: 355

From: "Donald E. Eastlake 3rd" <dee@cybercash.com>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 3:38pm

Subject: RE: Docket No. 970613137-7137-01

Thank you for this opportuity to comment. Attached are my comments as a
Microsoft Word document and a text copy of RFC 1958 as reference material.
I wish I had had more time to polish and extend my comments but hope that
they are still of some use in their current rough condition.

Sincerely
Yours, Donald
================================================================== ===
Donald E. Eastlake 3rd +1 508-287-4877(tel) dee@cybercash.com
318 Acton Street +1 508-371-7148(fax) dee@world.std.com
Carlisle, MA 01741 USA +1 703-620-4200(main office, Reston, VA)
http://www.cybercash.com http://www.eff.org/blueribbon.html
WARNING, on 4 Sep 1997, +1-508 will change to +1-978.

CC: "Donald E. Eastlake" <dee@cybercash.com>

Memorandum of Comment

DATE: August 19, 1997

TO: Patrice Washington, Off. of Public Affairs,
(NTIA), Room 4898,
14th St. and Constitution Ave., NW,
Washington, DC 20230 USA

FROM: Donald E. Eastlake, 3rd
318 Acton Street
Carlisle, MA 01741 USA

RE: DEPARTMENT OF COMMERCE

[Docket No. 970613137-7137-01]

Request for Comments on the Registration and Administration of Internet Domain Names

CC: Internet Assigned Numbers Authority, Internet Architecture Board, gTLD-MoU Policy Oversight Committee

----------------------------------------------------------------------------------------------------------

Dear Department of Commerce,

Attached is my response to your above referenced request for comments. I hope that you find it helpful. In addition, I have attached a copy of Internet RFC 1958, Architectural Principles of the Internet, which I believe may provide a useful broad background, particularly as to its rapidly changing and interoperabilty based nature. Principles 4.2, 4.3, and 4.4 from this document apply particularly to the Domain Name System.

Sincerely Yours,

Donald E. Eastlake, III

A. Appropriate Principles

[The Government seeks comment on the principles by which it should evaluate proposals for the registration and administration of Internet domain names. Are the following principles appropriate? Are they complete? If not, how should they be revised? How might such principles best be fostered?]

While you have good principles, which I comment on below, missing from this section are any overarching principles as to why the Domain Name System (DNS) exists and what it is good for. I think it is necessary to always keep in mind, when considering any change, that the essential purpose of the DNS is to

(1) provide stable and (2) globally unique names

that are either

(3a) mnenomic or (3b) systematic.

Going into a little more detail on each on these:

(1) DNS names primary function is to provide a stable cyberspace identity. When your telephone number changes or your postal address changes or even you Internet Protocol (IP) binary address changes, people who know your cyberspace identity are, because of DNS, unaffected for cybersapce interaction. (Note: some types of cyberspace identity include a user or account name as well as a domain name but even these identities are, for some purposes, completely mapped in to a domain name with the user name as the initial label such as mapping user@host.org.nil to user.host.org.nil.) In fact, in order to provide aggregatable routing and avert the collapse of the Internet due to routing tables overflowing, IP binary address renumbering is becoming and increasingly common and frequent occurrence more or less eliminating the only plausible other globally unique identifier as insufficiently stable.

(2) DNS names must be globally unique to be of any use. To permit interpretation, when a name is referenced in a email address or a web link or any of the many other contexts where a domain is used, it must mean the same thing to different people. It is a witness to the general acceptance of this principle that even the most strenuous and, in some causes, fraudulent efforts of those attempting to set up rogue root zones and rogues top level domains (i.e., those not recommended by the Internet Assigned Numbers Authority (IANA)), have never managed to get more than about 0.5% of the Internet to recognize their names until recently where, for a brief period of time for a few names, they went beyond fraud and criminally interfered with the operation of computers owned by the US government and others in attempts to temporarily secure recognition.

(3a) In the areas of the DNS name tree that are of primary concern in this inquiry, it is very desirable for DNS names to mnemonic. People should be able to say them over the telephone, write them on a napkin, and have some chance of remembering them, at least for a brief period, without too much confusion. The mnemonic portions of the DNS tree are, at the current stage of Internet software development, frequently seen by people.

(3b) There are substantial and very important portions of the DNS name tree that are systematic. For example, there is in "*.in-addr.arpa" part of the tree, which is used to map IP version 4 binary addresses into the domain name system so pointers to their primary symbolic name can be found, and "*.tpc.int" which maps every telephone number in the work into the DNS. For example, my phone number of +1 508-287-4877 maps into 7.7.8.4.7.8.2.8.0.5.1.tpc.int (i.e., reverse the digits and add periods and .tpc.int.) This format permits convenient delegation at country code, area code, city code, etc., levels within telephone numbers. In addition, the "??." portion of the top level domain name space of DNS, that is, the two letter TLDs, map all ISO two letter country codes into the DNS and could be considered a systematic portion. The systematic portions of the DNS tree are frequently program generated and, except for the ISO country codes, not usually seen by people.

Portions of the DNS name tree can be classified as either mnemonic or systematic or divided into contiguous subparts with these characteristics.

As to the first sentence above, while competition is an important tool, I believe that the principle that should be encouraged is choice and variety between stable names and policies, not just competition. While the ISO country code TLDs and the special .edu and .int TLD provide some variety, it would be preferable to have some gTLDs shared among competitive registrars and some gTLDs run by entities under a regulated utility concept, to provide at least two varieties, rather than to put all of the gTLD DNS eggs in the basket of one system.

As t the second sentence above, while interoperability is the paramount virtue of the Internet, I wonder what you mean by "not be permitted"? There is every evidence that operators and users of portions of the Internet infrastructure realize this which is why they follow the recommendations of the Internet Assigned Numbers Authority, including IANA's recommendations as to the contents of the DNS root zone, and the Internet Engineering Task Force, to voluntarily achieve such interoperability. Certainly no one should be permitted to destroy the interoperability of other parties through crime, fraud, or abuse, but that is vastly different from an operator or user knowingly deciding to use a different DNS name space. Coercive attempts to prohibit this would probably have the opposite effect of turning a few adolescent rebels who are mostly ignored into martyrs.

Finally, interconnectivity is the result of voluntary cooperation. While the DNS should not be designed or used so as to gratuitously block connectivity, many organizations have enclaves containing local machines which they do wish to have external connectivity. While this does not interact with gTLDs, there is currently a draft proposal before the IETF to create a .local TLD to standardize the handling of such hidden areas (see draft reference in Appendix A).

Yes, and the private sector has made giant strides toward such mechanisms via the gTLD-MoU and the open process it was arrived at.

c. These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.

Yes and the gTLD-MoU allows for such evolution, but I would add that the governance mechanisms should also recognize regional differences in point of view.

d. The overall framework for accommodating competition should be open, robust, efficient, and fair.

It's certainly hard to disagree with any of these principles in those cases where competition is the correct tool.

e. The overall policy framework as well as name allocation and management mechanisms should promote prompt, fair, and efficient resolution of conflicts, including conflicts over proprietary rights.

This is a vital point. Since the most important characteristic to strive for in domain names is stability, at least on the time scale of the existence of the entities named, efficient and early resolution of conflicts is important, though there are other factors of significance. Unless care is taken, the cost of conflict resolution can dominate the costs of a system.

f. A framework should be adopted as quickly as prudent consideration of these issues permits.

A firmer foundation from domain name registration is desirable but there actually does not seem to be any immediately threat of collapse or the like.

General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name registration systems?

There are actually quite a variety of systems. Each ISO country code has its own, run as it seems fit and has the advantage for providing local autonomy and express with the disadvantage of possible instability and capriciousness. Network Solution Inc. approximates a loosely regulated utility and provides a large capable organization but currently has offices only in the US and while its prices are regulated by the NSF, it is not at all clear that effective oversight has been provided for its terms and conditions. The gTLD-MoU proposes a self-regulated competition model with a likely advantage of low cost and responsive service and the disadvantage of instability in the existence of registrars (but not in names if properly engineered).

2. How might current domain name systems be improved?

Probably many ISO Country Code systems could be improved but it's their business.

The NSI system could be improved by requiring the establishment of offices in other times world regions and changes to its trademark policies to give less weight to the artifact of registration and more to the reality of infringement.

The gTLD-MoU system is not yet functioning so operational improvements are not clear.

3. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

This ISO country code domains are properly administered by the corresponding country if it so wishes. In default of a statement by such country, IANA appears to have adequate policies.

If NSI or a similar regulated utility model is to continue, some strong regulatory body would be needed. This is a difficult questions as a national body such as the US FCC seems inappropriate but the office of IANA, for example, would have to be beefed up significantly to take this on.

The gTLD-MoU POC is a great start at a balanced administrative body. To account for regional differences, it should be augmented by voting representatives for RIPE and APNIC immediately, a non-voting representative for ARIN to change to a voting representative as soon as most of ARIN's governing body is elected instead of appoint (i.e., in two years according to the last by-laws I saw), and similarly representatives of the proposed African and Latin American NICs should also be added. These five regional representatives would provide closer feedback to the POC from those affected by its policies and a good balance between a homogenized global view and an overly parochial local view.

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

Domain name registration is really a new thing and it is not clear that any existing model would server well in setting policy for the creation or operation of TLDs. Regulated utility or competitive models may work to see that such policies are carried out, once set, but are not helpful in setting policies.

There are few if any governmental organizations, particularly at the international level, used to moving as fast as the Internet does. Nor is the situation stable. There is continuing engineering to improve the DNS, the possibility of a new generation DNS II being designed, etc. Many of the questions in this request for comments seem based on the idea that domain name to IP address translation is the only important function of the DNS, but it also provides mail redirection and may in the future provide other services of equal importance. Any decision making system needs engineering input to avoid political effort to fine tune an answer to some current problem from blocking the evolution and growth of the DNS to new uses.

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the .com registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

Deleting any top level domain name with a substantial number of users is a terrible idea that, in any case, would be almost impossible to implement. It violates the fundamental purpose of DNS because it violates that stability of existing names. It is exactly the sort of really destructive move that would cause whoever mandated it (be they governmental or private) to be a laughing stock and cause most of the Internet to switch to using unofficial and possibly inconsistent root zones instead of the official namespace. It is this difficulty in ever eliminating a TLD (generic or otherwise) that implies that great caution should be exercised in adding TLDs of any type. It would be possible to freeze new registrations in ".com".

There is no ISO country code TLD management question. It's up to the countries involved. gTLDs are a whole different matter. But even ISO and gTLDs are not the whole story. There is .edu and .int and I believe the evolving set of five regional NICs (ARIN, RIPE, APNIC, etc.) for the world should each have a TLD which would be rTLD or regional TLDs, etc.

6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

The primary controversies are social and political. Registrars and TLDs are inherently dependent on the root zone.

7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and coordinate?

For interoperability, everyone has to use the same root zone contents except that disjoint sets of root servers could be used. This means that at least from a logical perspective, a single central authority such as IANA must exist and almost everyone needs to follow their recommendations.

8. How should the transition to any new systems be accomplished?

In general, new systems should be created to use new TLDs. Only after they have actually shown themselves be to be not just operational and competent but significantly superior overall should the transition of any existing reasonably functional TLD to such a "new" system be contemplated. Even then, adjustments to improve the "old" system that has been shown inferior should be considered first. The benefits in terms of organizational robustness of multiple systems should be considered.

Of course, if the registration or, even worse, domain names service for one or more TLDs actually collapsed, it would be essential to their stability as identifiers to get the names back on the air and get registration services operational at least to the extent to be able to update the locations of servers and the like, so any transition might be an improvement.

9. Are there any other issues that should be addressed in this area?

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

There are many such considerations:

Technical Considerations:

There are reasons to avoid single letter TLDs. Should the number of TLDs become very large, solutions involving splitting the server load for root can be more efficiently and elegantly accomplished if the one letter TLDs are still available.

As the number of TLDs increases (regardless of whether they are gTLDs or other kinds), the load on servers grows but this is unlikely to be a substantial consideration unless insane policies are adopted (such as a free for all where anyone can grab any unclaimed TLD at no cost resulting in many computers hammering on some TLD allocation server allocating absurd numbers of TLDs and overloading root server memory).

Practical Considerations:

In the arena of gTLDs and other TLDs longer than two characters, mnemonicity should be preserved. Thus no two such TLDs should be easily confusable in the minds of users. Appendix A to my comments, below, suggests barring any gTLD that would be a SOUNDEX conflict with an existing longer than two letter TLD.

Because gTLDs can not practically be eliminated once they achieve significant use, they should be added with some caution, probably not more than 10% of the existing number of TLDs as an increment each year.

Policy Considerations:

Two letter TLDs are reserved for two letter ISO country codes, and so should not be used a generic TLDs.

New gTLDs should avoid violating existing rights in names. For example, gTLDs should not be globally recognizable country, geographic, or organizational names. (Such names might be made TLDs under the right circumstances where the named entity controlled them.)

11. Should additional gTLDs be created?

Yes, I believe that additional gTLDs should be created in order to provide more variety in policies of registration and variety in names.

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

In something changing as rapidly as the Internet, there are no guarantees but in general, the DNS was engineered to handle deeper names and smaller zones than are currently in use. Thus, for example, a policy of requiring two level names to register in a hypothetical not TLD, say .biz, would be beneficial. Then, instead o of things like unitedairlines.com, americanairlines.com, etc., you could have the delegation point be names like united.airlines.biz and american.airlines.biz. With reasonable controls on the proliferation of second level domain names in such a zone, it would be much more scaleable. If enough *.airlines.biz registered, the entity running .biz (either a coordination point for non-exclusive registrars or an exclusive registrar) could split out airlines.biz to separate server machine and run it as a separate zone.

13. Are gTLD management issues separable from questions about ISO country code domains?

gTLDs and country code TLDs are very different. The gTLD area is appropriate for one or more experiments in self-regulation. The ISO country code domains provide privileged short TLDs to each of the recognized political areas of the earth.

14. Are there any other issues that should be addressed in this area?

This section and this entire inquiry seem to be based on the premise that gTLDs (generic Top Level Domains) are the only question. While the settlement of this question may seem to be the most urgent, I do not believe that it is the whole story. For example, it would be reasonable to the five regional IP registries that exist or are being formed (ARIN, RIPE, APNIC, etc.) should be able to each administer a regional TLD (rTLD). Variety of user choice is the key.

D. Policies for Registries

15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

In as dynamic and rapidly changing environment as the Internet, perfect answers don't exist and providing choice between a variety of policies is probably best. Certainly exclusive and non-exclusive TLDs can exist. There may be a problem in maintaining ".com" because of its dominance but if new registrations in ".com" were frozen, to create a more level playing field, it would be reasonable from a policy point of view to have some gTLDs run exclusively under a regulated utility registrar model and others under a non-exclusive competitive model.

16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have? Who will determine these and how?

Users of domain names frequently build up value in those names and become dependent on their stability. Steps need to be taken to assure a reasonable level of stability. For a non-exclusive competitive model, it may be possible to get away without threshold requirements, by placing the burden on some exclusive central coordinator to maintain the names. For exclusive TLDs run under a regulated utility model,

17. Are there technical limitations on the possible number of domain name registrars?

If you are talking about registrars having exclusive control over a gTLD, there is no reason to believe there would be any more technical limitations that there are on, say, second level registrars in any particular TLD. For example, .com has shown that, with great effort and substantial pain, hundreds of thousands of such registrars can be technically accommodated, although having that many with different names produces so many names which are so similar that the resulting cesspool of confusions destroys mnemonicity.

If you are talking about registrars that are sharing one or more TLDs via some central coordinating mechanism, it appears from the operation of such mechanism in the United Kingdom that at least a few hundred registrars can be accommodates.

18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

There is little effect on the name space itself just because of the number of direct or indirect exclusive or non-exclusive registrars.

19. Should there be a limit on the number of different gTLDs a given registrar can administer? Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

There does not seem to be a reason for such a limit for non-exclusive TLDs. For exclusive TLDs, variety in service and polices would, to some extent, come from a variety of registrars which tends to imply limits on the number of TLDs they administer.

20. Are there any other issues that should be addressed in this area?

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-a-vis domain names?

This entire issue is grossly overblown with regard to trademarks and has not been sufficiently considered in relation to national / geographic indications. The mere use of a name at some level in a domain name, even if that name is used in commerce or registered as a trademark should not normally be considered infringement. The questions is, does the use cause actual confusion in people's minds.

Of course, the most important case would be to review new gTLDs to see that they do not have important conflicts. For new gTLDs, it seems practical to announce them months in advance. Of course, with the level of interest in and conflict concerning DNS, it would be expected that there would be some complaints against any TLD name.

Within ISO country code TLDs, it is solely the prerogative of the country involved to resolve such disputes although other national courts can issue orders concerning operations within their jurisdiction. For generic

It is not possible to prevent all trademark conflicts. People use names to mean different things. Commercial usage changes, language changes, conflicts can occur where there initially were none. The value of Aides trademark has been destroyed by the AIDS disease. So it goes.

It would probably be possible to nearly eliminate registered trademark conflicts by constructing a part of the DNS name tree with relatively artificial systematic names mapping all the jurisdictions that register trademarks and all the categories in which they register them. For example, foo.17.us.trademark for a trademark registered in category 17 in the United States. Creating such a system of domain names is fine, if they are used to provide more choices. But I think they would be unpopular.

Again, variety if the key. It would be good to have TLDs in which such basis is required and those where it is not. Those where it is required would be more expensive to administer but might provide more valuable names.

26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

I believe that the number and difficulty of trademark disputes will vary with the number of points of control and the number of policies. Thus, many registrars with exclusive control would tend to increase the problem.

27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

The question is misleading. It is only a very recent phenomena that organizations have actually started to use domain names, such as amazon.com, as trade marks. Where an entire domain name is a trademark, there is no possibility of direct conflict. The domain name is globally unique and so is controlled by and can only have been used in cyberspace by one entity.

Perhaps you means to ask if there is a technical solution to there being multiple users of the same "word", such as Apple, as a trademark in different product/service/geographic areas and the seeming desire of everyone to register as a second level domain name in .com. There really isn't much in the way of a technical solution to this problem. How would it help if you had apple.1.biz and apple.2.biz? The problem is potential confusion in users minds and I suppose apple.music.biz and apple.computers.biz would help but to judge who should be in what second level name would be an administrative nightmare since most businesses would just try to grab as many categories as possible.

28. Are there any other issues that should be addressed in this area?

Appendix A:

TLD "collision" and SOUNDEX

One method that could be used to determine if gTLD names are likely to "collide" in the minds of users and violate the mnemonicity goal would be to avoid the creation of any TLDs outside the two letter ISO country codes if their SOUNDEX Codes collide. Section A.2 below described how SOUNDEX is computed and section A.1 should that none of the domain names longer than two characters that exist or are proposed by the gTLD-MoU POC or that I have proposed in Internet Drafts collide.

Other methods are possible. For example, you could define a "step" as adding a character, deleting a character, or swapping two adjacent characters. You can then define the distance between two names as the minimum number of such steps needed to transform one into the other.

A.1 TLDs Longer than Two Letters by SOUNDEX Code

SOUNDEX Category Name

A610 existing ARPA

A632 new gTLD ARTS

C500 existing COM

E251 test/doc EXAMPLE

E300 existing EDU

F650 new gTLD FIRM

G100 existing GOV

I510 new gTLD INFO

I530 existing INT

L240 local LOCAL

M400 existing MIL

N300 existing NET

N400 test/doc NIL

N500 new gTLD NOM

O620 existing ORG

R200 new gTLD REC

S360 new gTLD STORE

T230 test/doc TEST

W100 new gTLD WEB

Category explanations:

existing = currently in the IANA recommended root.

new gTLD = the seven gTLDs being created by the gTLD-MoU POC.

test/doc = a TLD suggested for reservation for use in private testing and documentation by an existing Internet Draft (ID). This would avoid possible conflicts between names in actual use and those used for expository purposes and in private testing. The current ID is at <ftp://ftp.isi.edu/draft-ietf-dnsind-test-tlds-00.txt>. ID's are ephemeral documents that can lapse or be replaced at any time by later versions and are deleted when and if they finally issue as either and informational or standards track RFC.

local = the TLD suggest for use in identifying names of local scope. See Internet Draft <ftp:://ftp.cybercash.com/pub/dee/draft-eastlake-local-tld-00.txt>. See immediately above for information on Internet Drafts.

A.2 How to Calculate SOUNDEX

Example: Schmidtzel

1. First delete all spaces and non-alphabetic symbols. Delete all accents, umlauts, and other diacritical marks. Delete all H and W, unless one is an initial letter. (Scmidtzel).

2. The first letter of the name becomes the first letter of the Soundex. (Scmidtzel).

3. For the remaining letters of the name, write down the code number from the following table. (Scmidtzel=S25033204).

A, E, I, O, U, Y = 0

B, F, P, V = 1

C, G, J, K, Q, S, X, Z = 2

D, T = 3

L = 4

M, N = 5

R = 6

4. Combine all double numbers into one. (S2503204).

5. If the first number is the same as the code number for the initial letter, delete that number. (S503204).

6. Delete the zeroes. (S5324).

7. Retain one letter and three numbers (S532).

8. If there are fewer than three numbers, add zeroes to the end to make three. For example, Schell=S400.

NOTE: It is important not to delete the zeroes (vowels) until step 6, because double numbers separated by zeroes are not to be combined.

###

Number: 356

From: <kconnolly@evw.com>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 3:54pm

Subject: Response to Request for Comments

It is my honour to transmit the Response of the members of the Policy

Advisory Body ("PAB") established in connection with the Memorandum

of Understanding Respecting Generic Top Level Domains. As indicated

in the attachment (which is in DOS text format), the PAB's internal

procedures require that a document be circulated for at least two weeks

before cloture may be invoked to test the existence of a rough

consensus. While the attachment has been circulated and commented

upon for a longer time than two weeks, it includes revisions which have

been made to reflect its members comments. Accordingly, PAB is

constrained by its own fundamental documents from declaring the

attachment to be other than the statement of the members who have

subscribed to it. However, PAB will supplement this message in two

weeks' time to indicate whether, and the extent to which, this Response

may be viewed as an official statement of PAB. The supplement will be

e-mailed to the NTIA and otherwise appropriately publicized.

Very truly yours,

Kevin J. Connolly

PAB Member

CC: NTIADC40.SMTP40("pab@gtld-mou.org")

The rapid growth of the Internet has not gone unnoticed by the

self-governing bodies which have, with the assistance of the United

States, fostered the growth of the Internet to date. The International Ad

Hoc Committee was formed in response to an Internet-Draft circulated by

Dr. Jon Postel of the Information Sciences Institute of the University of

Southern California. The process through which Dr. Postel's proposal

has been discussed and refined through broader and broader public input

and consensus-building processes is in the best traditions of the Internet.

The Policy Advisory Body is proud to have been a participant in the

process of building consensus which has lead to the development and

refinement of the GTLD-MoU. On May 1, 1997, with the signing of the

Memorandum of Understanding relating to Generic Top Level Domains

("GTLD-MoU"), the International Ad Hoc Committee concluded its work

and passed the baton along to the bodies mentioned in the GTLD-MoU.

The Policy Advisory Body ("PAB") is one of the bodies established in

connection with the GTLD-MoU. Its role is to provide recommendations to

the Policy Oversight Committee regarding general policy matters,

amendments to the GTLD-MoU and to the Memorandum of

Understanding governing the Council of Registrars. Its role is analogous

to that of the Internet Engineering Task Force, in the sense that PAB

members are uncompensated, driven by a desire to participate in the

development and growth of the Internet, and it makes its

recommendations to a much smaller body (The Policy Oversight

Committee, which in this sense is analogous to the Internet Activities

Board) which is responsible for implementing policies which reflect the

input not only of PAB but of relevant experts and other relevant entities.

The members of PAB welcome this opportunity to comment on the

relevant issues as the United States Government develops its policy in

this crucially-important sector.

This document was prepared by a member of the Policy Advisory Board in

consultation with a number of its members. The document was then

circulated on an Internet Mailing List to all members of the Policy Advisory

Board for comments and suggestions. It was refined in light of those

comments and suggestions, as well as in light of the lessons learned at a

conference, held July 30-31, 1997 in Washington D.C. under the auspices

of the Interactive Services Association, the Center for Democracy and

Technology, and the Information Technology Association of America. The

document was then re-circulated for final comments and input. PAB's

internal ethic calls for members who disagree with a proposed action to

reply to the mailing list, and that an item-by-item poll of membership be

taken if ten or more of the 157 members request it. There have been no

such requests. PAB's internal ethic also calls for a two-week time period

in which members may call for a vote. However, because this document

has evolved in response to comments from members, some portions of

this document have not been exposed to comment by the membership of

PAB for the requisite two-week period. Accordingly, while the major

portion of this document meets PAB's criteria for stating the "rough

consensus" of its community, our adherence to the traditions of the

Internet requires at this time that the substance of this document be read

as the statements of the persons whose names appear below. Any

statement in this document such as "PAB believes" or "it is PAB's position"

should be understood as a declaration of the authors and supporters that

they believe these statements carry the rough consensus of PAB's

membership, ascertained in the manner indicated above. With your

permission, we would like to supplement this submission in two weeks'

time with a formal statement clarifying whether, and the extent to which,

after due opportunity for balloting, this statement can properly be read as

the collective position of the gTLD Policy Advisory Board.

Kevin J. Connolly, PAB Member and Principal Author

Eaton & Van Winkle

Antony VanCouvering, PAB Chair

NetNames USA, Inc.

Rick H. Wesson, PAB Deputy Chair

Internet Business Services

Robert F. Connelly, PAB Secretary

Procurement Services International

John Broomfield, PAB Member

SYSTEL

Dan Busarow, PAB Member

DPC Systems (with Addendum regarding Part E)

Stéphane Guerry, PAB Member

Interdeposit

Robert Nelson, PAB Member

The Internetwork Operating Company, Inc.

Ivan Pope, PAB Member

Netnames U.K.

Werner Straub, PAB Member

Axone Services & Development, S.A.

(with addendum) respecting B1, D15, D20

Alan Sullivan, PAB Member

Top Domain Registry, Inc.

(with addendum regarding B1, D15, D20)

This Response addresses all of the points raised in the Request for

Comments. It is enumerated in accordance with the Request for Comments.

A. Appropriate Principles

PAB believes that the principles set out in the Request for Comments are

generally appropriate. It believes that the principles of subparagraph (a)

should be expanded in one respect, by explicitly recognizing that the root of

the domain name system is a unique resource which must be administered

as a public trust in order to promote the objectives of interoperability, stability,

robustness, efficiency and fairness. PAB believes that explicitly recognizing

that the root of the domain name system is a public trust is the best, if not the

only, means of fostering these principles.

The principle of private sector leadership can be fostered by governmental

observation, comment, and support for proposals which hold reasonable

prospects for stability, consensuality, and self-governance. PAB welcomes

the participation of the United States government in seeing that the final

result is one which adequately defines responsibilities and maintains

accountability. PAB also believes that accountability can be fostered within

the existing framework through the existing functions of the Internet Society,

which is a democratically-operated organization of Internet-aware and -concerned persons worldwide. As stated in the documentation maintained by

the InterNIC which explains the Internet, "The ultimate authority for where the

Internet is going rests with the Internet Society, or ISOC. ISOC is a voluntary

membership organization whose purpose is to promote global information

exchange through Internet technology."

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name

registration systems?

The current system must first be identified. PAB draws a distinction between

(a) the ISO 3166 domains, such as the hierarchy of .us domain names, (b)

the generic top level domains (currently, .com, .net and .org) and (c) the

specialized domain names (.edu, .mil, .gov and .int). PAB is concerned only

with the generic top level domain names.

To the extent that there are problems in the ISO 3166 domains, we believe

that the relevant public authorities in each country should decide the extent

to which reform is indicated and the manner in which it should be

implemented. While PAB recognizes that a different structure of the .us

domain could have gone far to prevent the pressure on the .com, .net and

.org domains we are now experiencing, it also recognizes that the opportune

moment for including .com.us, .org.us, and .net.us as second level domain

names has passed. (It is also worth noting that a differently-configured .us

domain might have completely obviated the trademark disputes that create

so much concern today. Nevertheless, the .us domain has not been so

structured, and it does not seem likely that United States companies will flock

to that domain even if it is restructured today.) United States companies

perceive a high value in registering their domains under the generic

international top level domains, and the likelihood of convincing them to the

contrary is small.

Likewise, we do not perceive problems with the specialized top level

domains. Not every enterprise will be granted access to these registries; on

the contrary, only a limited number of entities will ever appear in these

registries; they are, and likely will remain, non-commercial; and there is scant

incentive for the private sector to become involved in their administration. At

the same time, PAB does not believe that a need exists for the creation of

additional special domains. It would be difficult to adjudge the merits of a

proposed special domain, to distinguish a bona fide special domain from an

attempt by a speculator to create a sinecure for profit. It is also hard to

understand why most, if not all, current proposals for special domains would

not fit comfortably within another top level domain.

2. How might current domain name systems be improved?

A principal advantage of the present system is its vertical integration. For

most registrants, InterNIC stands as a single-source font for domain name

registration, IP number allocation, and root name service. The single-source

aspect minimizes the opportunities for responsibility-shuffling.

The disadvantages, on the other hand, are considerable. Prospective

domain name registrants are confronted with a pricing structure that has

been imposed by unilateral fiat, without the benefits of competition. There

are considerable venues for non-price competition as well, but the

single-source nature of the existing system rules out non-price competition

as well. The contractor to whom InterNIC has delegated the operation of the

domain name system -Network Solutions, Inc.("NSI")- has continued to

charge registrants the full registration fee -covering two years of operation-

even though its contract with InterNIC expires in March 1998. NSI is now the

subject of a civil antitrust investigation by the United States Department of

Justice. Whether NSI has unlawfully exploited its position or not, its position

has considerable potential for monopolization. Moreover, NSI has adopted

policies and practices which PAB believes could never be supported in a

competitive marketplace. Its Domain Name Dispute Policy -discussed below

under Trademark Issues- has been widely, and, we think, soundly, criticized

as irrational and serving no purpose other than to insulate NSI from liability

for secondary or derivative trademark infringement. Other aspects of NSI's

policies and practices bespeak a bureaucratic inertia and unresponsiveness

which would seriously undermine its position in the market if there were an

alternative to using it.

3. By what entity, entities, or types of entities should current domain name

systems be administered? What should the makeup of such an entity be?

The most important single change which can be introduced into the domain

name system is competition. Competition in turn will require a degree of

cooperation in order to enable the public to connect seamlessly to content

providers who have chosen to employ different registrars. Competition does

no good if the Internet were to experience the fragmentation that

characterized telephone communications before the advent of the Bell

System. While it might be possible to commit the operation of the central

system to a council of registrars, the potential that coalitions would form and

take anti-competitive action against other registrars is too great to ignore.

The GTLD-MoU addresses this concern by committing policy decisions to the

Policy Oversight Committee, which is charged with administering the system

as a public trust. While the registrars will have a voice in the system, it will

not be loud enough to control the operation of the domain name system.

Further balancing forces are supplied by providing for a Policy Advisory

Body, which is based on the widest possible public input, and which

aggregates the competing demands to build consensus in the traditions of

the Internet.

4. Are there decision-making processes that can serve as models for

deciding on domain name registration systems (e.g., network numbering

plan, standard-setting processes, spectrum allocation)? Are there

private/public sector administered models or regimes that can be used for

domain name registration (e.g., network numbering plan, standard setting

processes, or spectrum allocation processes)? What is the proper role of

national or international governmental/non-governmental organizations, if

any, in national and international domain name registration systems?

An existing process that can serve as a model for guiding the future of the

domain name system can be found in the relationship of the Internet

Engineering Task Force (IETF) and the Internet Activities Board (IAB,

formerly known as the Internet Architecture Board, a component body of the

Internet Society). The "standards setting" process is one aspect of the

IETF/IAB relationship, but PAB does not believe that the extreme rigor which

characterizes the standard setting process needs to govern all aspects of the

domain name system reform. While there will certainly be some aspects of

the reform which demand that level of rigor, the development of solutions "as

quickly as prudent consideration of these issues permits" demands that the

standard setting process not be applied indiscriminately.

The network numbering system provides a model only to the extent that it

demonstrates that competitors can successfully share in the use of a fixed

and unique resource, i.e., the universes of telephone numbers and IP

addresses. That model also illustrates the benefits of permitting portability:

just as an 800/888 telephone number assigned to a user by one long

distance carrier can be retained when changing carrier, so, too, a multiple-registrar system makes it possible for a user to keep a domain name when

changing registrar. Spectrum allocation is not usefully analogous to the

domain name process either. While both processes deal with a unique and

fixed asset, the depth of namespace is so far greater than the

electromagnetic spectrum that the considerations are highly distinct.

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from

circulation? Should geographic or country codes (e.g., .US) be required? If

so, what should happen to the .com registry? Are gTLD management issues

separable from questions about International Standards Organization (ISO)

country code domains?

PAB believes that any system which would undermine the stability of the

.com registry (as well as .net and .org) would be irresponsible and should be

consigned to the scrap heap. Enterprises on the Internet have far too much

invested in the .com hierarchy for anyone seriously to consider retiring it.

PAB believes that upon the expiration of the cooperative agreement between

the National Science Foundation and NSI, the registration and maintenance

of the .com, .org and .net registries should devolve upon the regime which

administers generic top level domains generally.

The issues relating to ISO 3166 country codes are entirely separable from

those of generic top level domains. PAB believes that each country should

determine for itself, as a matter of national pride and self-determination, how

to administer the region of cyberspace which pertains to its national identity.

6. Are there any technological solutions to current domain name registration

issues? Are there any issues concerning the relationship of registrars and

gTLDs with root servers?

Technology-based solutions to current domain name registration issues,

while available, appear to be unattractive. The DNS namespace has turned

out to be much "flatter" than was originally expected. The designers thought

that third-level (and higher) domain names would be used much more than

has turned out to be the case. Thus it might have been expected that a

domain relating to the best-loved rodent on earth would be something like

mickey.disney.com. It has not turned out that way. The public has not used

the Internet in the way that the engineers would like it to be used. Ultimately,

the use of search engines to locate desired content on the Internet will

become second-nature to users. Whether that is implemented by educating

users or embedding search engine invocation in web browsers cannot

presently be predicted.

Another species of available technological "fixes" is available in the greater

use of geographic top level domains. However, the .com, .net and .org

domains have always been populated by predominately United States

domain name holders, while the .us domain has structured itself as

exclusively geographical. While there is a second level domain within .us

that would pertain to nationwide United States enterprises (.dni.us) it is

virtually unknown to the users of the Internet and content providers. Hence

while the technology certainly exists to relieve pressure on the domain name

system, it is unlikely to be used, and it cannot overcome the principal

shortcoming of the existing system -its lack of competition.

The crucial issues regarding the relationship between registrars and gTLDs

with root servers can be simply stated:

(a) There can be only one true root. Fragmenting the root, so that

calls by users for a domain will be resolved differently

depending on whom one asks will spell the end of the Internet.

While it is far from disastrous if a web browser is served the

wrong page, it is very serious if e-mail is sent to the wrong

address.

(b) The root will expand, but its rate of growth must be controlled.

Once a zone is added to the root, it will be as hard as kudzu

to eliminate. Far less harm can be done to the system by

restraining the rate of growth than by allowing the root to grow

in an uncontrolled manner.

7. How can we ensure the scalability of the domain name system name and

address spaces as well as ensure that root servers continue to interoperate

and coordinate?

First of all, the fundamental relationship between the root servers and the

TLD servers should not be changed. The root servers do not name the

decision whether to add new zones to the root. That decision has, up until

now, been made by IANA. If IANA delegates that decision making authority

to a different body, the root servers should not seize upon that delegation as

an invitation to make these determinations for themselves.

Second, the slow and deliberate growth of the root is crucial.

Third, the number of top level domains must be limited. There are technical

limits on the number of TLDs which are possible, but the number may be as

high as 30,000. The problem lies in the fact that once the door is opened for

registration of top level domains, what principled basis exists for allocating

the small number of domains among the competing applicants? Moreover,

how could the administrators of the system possibly police the creation of so

many top level domains? Keeping in mind that the consequence of allowing

the technical limits to be exceeded is the collapse of the domain name

system, a policy of allowing the root zone to grow rapidly and to large size

appears to be unwise and dangerous.

The practice of enlarging the root slowly is also consistent with the way the

Internet has evolved to date: the service is delivered, used, and then refined

in light of the incremental lessons derived from use. This sort of feedback

loop has served the Internet well in the past and no reason appears at this

time to depart from it. The alternative is to allow the root to grow with such

speed, and to such size, that its collapse becomes inevitable.

8. How should the transition to any new systems be accomplished?

The transition to any new system should not overlook the time-honored

technique for computer system implementation: design the new system;

implement it in parallel, so that both old and new systems operate

contemporaneously; and only then is the functioning of the system committed

to the new software.

It is worth noting that if the Internet evolves in the way the signatories to the

gLTD-MoU anticipate, NSI will be able to retain its existing customer base by

becoming a CORE Registrar. Alternatively, it may well evolve that NSI will

secure an extension of the cooperative agreement under which it has

operated the registration services provided by the InterNIC, or it may secure

a fresh grant of the franchise to provide registration services for the

.COM,.NET and .ORG. domains. The disadvantage in that transition model

is that the owner-developers of the second-level domains within .COM and

its cohorts will be required to abandon their domain names in order to secure

the benefits of portability afforded under the GLTD-MoU.

PAB believes firmly that NSI has no cognizable intellectual property rights in

the .COM, .NET and .ORG registries; that NSI can and should be required

to make the registry databases available to any successor registrar or group

of registrars; and that if NSI is in fact more competent than other registrars,

it can succeed in a competitive environment with other registrars.

9. Are there any other issues that should be addressed in this area?

a. Compelling NSI to surrender the fruits of the cooperative agreement, as

required by the terms of that agreement. At this time, NSI has asserted that

the fruits of its work under that agreement are its own intellectual property.

It has not agreed even to provide a comprehensive report of the work it has

done.

b. Compelling NSI to surrender the domain name database for .com, .net

and .org when its contract expires.

C. Creation of New gTLDs.

10. Are there technical, practical, and/or policy considerations that constrain

the total number of different gTLDs that can be created?

Technically, the DNS can support a fairly large number of top level domains.

The number most often bandied about is 30,000. However, there are several

valid policy reasons for restricting the number of TLDs:

(a) limiting the number of TLDs makes the task of policing intellectual

property rights against infringement more feasible.

(b) limiting the number of TLDs makes it feasible to police the zone files

against the insertion of spurious data, such as the infamous .PER zone.

(a) restricting the number of TLDs does tend to reduce the burden on the

root zone and enhance the performance of the DNS.

From a policy point of view, too, once the growth of the root is substantially

freed of restrictions, it is hard to establish a principled basis for stopping short

of the point at which the system is physically incapable of absorbing new top

level domains. The net result could well be a large number of

underpopulated domains, which benefits no one not the developers of

second level domains in the midst of these wastelands, not the owners of

trademarks who must surveil thousands of top level registries, and not the

:"owners" of the top level domains who will confront a market which will be

glutted by domains with tremendous incentive to engage in predatory

competition. Many of those domains will be beyond the reach of United

States antitrust laws, thus exacerbating the problem.

11. Should Additional TLDs be created?

PAB answers this question in the affirmative. First, there is no alternative to

creating new TLD's if the existing problems are to be redressed prior to

March 1998 (when the NSF/NSI cooperative agreement expires). Moreover,

there is no guarantee that NSI will surrender its dominance in the domain

name registration process when its delegation of the .COM, .NET and .ORG

domains lapses. NSI's SEC filing indicates that it asserts intellectual property

rights in and to the domain name databases, and it intends to continue

operating as the exclusive registrar of the .com, .net and .org domains.

There is no guarantee that NSI will be unable to block other registrars'

acceptance and processing of these registrations.

Consequently, the only reliable solutions to the perceived problems of the

domain name system include the creation of new TLDs. New TLDs will

expand the Internet Name Space to make it possible for companies to

establish a presence in cyberspace which is logically related to their existing

business names. The differentiation made possible by distinct top level

domains reduces the likelihood of confusing similarly-named enterprises in

different markets. Most important of all, the gTLD proposal makes it possible

for the owner of a second level domain to port the domain when changing

registrars. This compels registrars to focus their business efforts on service

delivery rather than the unassailable security that comes from being the sole

supplier of a prized commodity.

12. Are there technical, business, and/or policy issues about guaranteeing

the scalability of the name space associated with increasing the number of

gTLDs?

There is an interrelated set of technical, policy and business concerns, all of

which converge in favor of allowing the DNS namespace to grow slowly, not

precipitately. The Internet paradigm of "deliver . . . use . . . refine" argues in

favor of this approach. The fact that the DNS can be broken by too rapid or

extensive growth points in the same direction. Finally, simplifying the task of

policing intellectual property rights also counsels in favor of allowing the

domain name system to grow slowly and deliberately.

13. Are gTLD management issues separable from questions about ISO

country code domains?

The ISO domains are as distinct from the generic TLD's as both the gTLDs

and ISO domains are distinct from .gov, .mil, .edu, .int, and in-addr.arpa.

These are distinct delegations from the root, and they can be managed

independently of each other.

The ISO domains and the gTLDs overlap with respect to IP space

allocations. When IPv6 is adopted, the restrictions on IP space will become

a thing of the past. So far as ISO Domain Name operations are concerned,

we do not believe that the ISO 3166 domain system has significant problems.

To the extent that there are problems, PAB believes that the relevant public

authorities in each country should decide the extent to which reform is

indicated and the manner in which it should be implemented. While PAB

recognizes that a different structure of the .us domain could have gone far to

prevent the pressure on the .com, .net and .org domains we are now

experiencing, it also recognizes that the opportune moment for including

.com.us, .org.us, and .net.us as second level domain names has passed.

United States companies perceive a high value in registering their domains

under the generic "international" top level domains, and the likelihood of

convincing them to the contrary is small.

14. Are there any other issues that should be addressed in this area?

D. Policies for Registries

15. Should a gTLD Registrar have exclusive control over a particular gTLD?

Are there any technical limitations on using shared registries for some or all

gTLDs? Can exclusive and non-exclusive gTLDs coexist?

PAB believes that no registrar should have exclusive control over a generic

top level domain. At the same time, there is no impediment to the

maintenance of exclusive control over specialized top level domains, such as

.int, .mil and .gov. The number of specialized domains should be sharply

restricted to prevent their being used to conceal what are, in fact, exclusively-owned generic top level domains. The reason for eschewing exclusivity

relates to the maintenance of a free market and ensuring that the developers

of second level domains have full and free portability between registrars.

Intellectual property rights, including copyright and patent, are facially

anticompetitive, but they serve an economically useful purpose in

encouraging the creation of protected works. There is no such creative effort

involved in the establishment of a top level domain.

Moreover, each top level domain is, or should be, unique. There should be

only one .com registry. Having multiple registries for a top level domain

would mean fragmenting the root and would pose potentially insuperable

problems for interoperability. It does not assist in the development of the

Internet as a global communications channel if one user's call for a second-

or higher-level domain within a given TLD is not consistently resolved

irrespective of the user's ISP. The question of multiple registries is quite

different from providing multiple registrars for a shared registry.

The technical limitations on the use of shared registries are just that,

technical. The technology has existed for some time of allowing multiple

users access to a fixed and uniform database. Record-locking and queuing

algorithms exist which are quite capable of overcoming the technical

difficulties presented by a shared registration system.

The problem posed by coexistence of shared and exclusive gTLDs is a

severe one. The Internet has already seen the explosive growth of

registrations in the .com domain to such an extent that new registrants are

finding that most of the attractive domain names have already been

registered. In many cases, the names have been registered by companies

that have no bona fide use for the name other than to sell the right to use the

name, at an exorbitant fee, to those who do have a legitimate use for the

domain. This practice of "warehousing" or "Cybersquatting" is one which

exacerbates the shortage of available domain names and which the existing

registration system is loathe to deter.

The creation of additional top level domains dilutes the power of

cybersquatters by providing additional space in which a legitimate domain

user can register. The ability of the additional domains to diminish

Cybersquatting is enhanced by the adoption of a policy which is adverse to

the hoarding of domain names. Particularly important is the absolute

prohibition on registrars' claiming second level domain names for their own

accounts.

16. Should there be threshold requirements for domain name registrars, and

what responsibilities should such registrars have? Who will determine these

and how?

A threshold requirement is a necessity to ensure that a registrar has the

necessary material ability to provide the registration services which are the

core of its function. A registrar must be able to respond to customer inquiries

through conventional channels (telephone, fax and postal mail) as well as

electronic means. Moreover, since history has shown that the registrar will

serve as a conveniently-fixed target for intellectual property litigation, the

registrar must have sufficient financial depth to deal with the expense

attendant upon responding to lawsuits. Even though insurance may well be

available to some registrars to cover these costs, it is inevitable that some

lawsuits will be cast to fall outside of the coverage of the policies, some

registrars' coverage will lapse, and some registrars -most notably, those

located in Japan- will be unable to obtain such coverage in the near future.

PAB has attempted to foster these requirements by insisting that registrars

have at least $300,000 in capital, and a staff of at least five full time

employees. PAB also encourages those registrars who have access to

commercial general liability insurance including coverage for trademark

infringement to obtain such coverage and keep it in force. PAB is also

exploring whether such coverage could beneficially be maintained by the

Council of Registrars for the protection of the entire system. PAB believes

that these thresholds are set as low as they can be without endangering the

reliability of the domain name system. PAB particularly is concerned about

the loss of credibility and trust that would result if a registrar were to accept

money from domain name aspirants and then fail to remit the registration fee

to the central authority, causing losses to the domain name aspirant.

The core responsibility for a registrar is the authentic transmission to the

shared database of the information necessary to register and maintain a

given domain. The system should also take responsibility for ensuring that

names which are congruent to recognized trademarks are registered only by

those who are duly authorized by the trademark holders to do so. The

practical constraints discussed below, however, make it infeasible to pre-screen domain registrations. Accordingly, gTLD conforms to the present

practice of allowing registrations to proceed without checking them for

trademark conflicts, and leaves it to the trademark holders to assert their

rights after the domain has been activated. The responsibility for assuring

that congruence has been placed, under gTLD-MoU, with the Administrative

Challenge Panels, to be administered by WIPO That responsibility, under the

gTLD system, has been institutionalized by making it possible for trademark

holders to exclude or totally block certain domain names in advance. A

common misconception is that only a very few trademark holders will be able

to invoke this pro-active exclusion process. In fact, while it would be unusual

for a name to be excluded from all of the generic top level domains, exclusion

from the domains that bear relevance to a trademark holder's business is

expected to be generally available to holders of trademarks which are

internationally known. The gTLD system is intentionally and highly hostile to

Cybersquatting. See below, p. 13.

PAB also believes that the gTLD system fosters the secure maintenance of

the domain name system by facilitating the use of digital signatures in

connection with registration-related transactions.

Other responsibilities of Registrars will be those which are required by market

forces. Under the shared registry approach of gTLD, Registrars will be free

to develop and market enhanced services and functions. When and if it

becomes evident that certain of those functions are essential to the orderly

maintenance of the domain name system, the gTLD regime has the flexibility

to evolve so as to require those functions and services of all registrars. Short

of imposing the requirement, the competitive market approach taken by the

gTLD-MoU will encourage the provision by Registrars of these functions and

services on a voluntary basis.

17. Are there technical limitations on the possible number of domain name

registrars?

PAB believes that there are not. In theory, there could be an unlimited

number of domain name registrars, all of whom would share access to a

common database.

18. Are there technical, business and/or policy issues about the name space

raised by increasing the number of domain name registrars?

The technical issues relate principally to assuring that the system will remain

as stable with multiple registrars as it is today, with a single registrar. The

approach being taken under the gTLD-MoU preserves this stability by

maintaining a bright-line division between (a) the operation of the domain

name system for lookup purposes and (b) its operation for purposes of

creating new entries. The systems being implemented under the gTLD-MoU

will interface with the global network of nameservers through a single central

database, to be maintained by or for the Council of Registrars. It is

anticipated that the principal difference between the gTLD system and the

existing system operated by Network Solutions, Inc., will lie in the mechanism

by which new domains are entered into the domain name system, without

altering the way in which the domain name information is looked up by users

who wish to access or use the Internet.

The primary policy issues raised relate to assuring that all registrars have

technical equality, i.e., that no one registrar or group of registrars will have

preferential access to the registration database. This equality is essential to

secure to consumers the benefits which are expected to flow from free

market competition. PAB believes that the equality issue can be redressed

through the use of queuing algorithms, which prevent a registrar from

creating more than a single entry in the shared database until the other

registrars have been polled and given an equal opportunity to register. When

and if there are so many registrars that the time lapse in polling all of them

produces palpable delays in processing, the queuing algorithm can be refined

to work with a pyramidal structure of channels which will assure prompt

throughput while preserving the level playing field which is essential to the

fair and equitable operation of a shared database. The name space itself will

not directly be impacted by increasing the number of registrars. The process

of domain name resolution will continue to be distributed, with updates from

the master database on a regular basis. There is no reason to suppose that

the operation of the domain name system will be affected materially by

altering the registration process.

19. Should there be a limit on the number of different gTLDs a given

registrar can administer? Does this depend on whether the registrar has

exclusive or non-exclusive rights to the gTLD?

PAB believes that the question of exclusivity drives the question whether a

registrar should be limited in the number of domains it can administer. If

domains are exclusive to a registrar, so that a registrar is the sole source of

registrations in a given domain, it stands to reason that limiting the market

power of registrars, to prevent the development of monopolies, calls for

limiting the number of domains which a registrar may control. Such a limit

has numerous practical problems, since it could be circumvented through

corporate combinations, joint ventures, and a myriad other business

structures which can conceal the reality of the business relationships.

PAB believes strongly that every registrar should have equal and unimpeded

access to registrations in every generic top level domain. Only in this way

can the benefits of free market competition permeate the entire domain name

system.

20. Are there any other issues that should be addressed in this area?

The first issue is the establishment of a principled basis for distinguishing

between "specialized" top level domains -- which can be operated by a single

registrar without doing harm to the interest of free competition -- and "private

property" domains, which, if allowed to proliferate, could do significant harm

to the public interest. To put it another way, PAB concurs that there may well

be times when additional domains, over which the Council of Registrars will

not have operational control, can properly be added to the root of the domain

name system without seriously impairing the integrity of the Internet, but at

this time it finds a lack of consensus as to a measurement system which can

be identified in advance for ascertaining whether a proposal is an attempt to

circumvent the public trust nature of the generic top level domain namespace

or is instead a bona fide specialized domain. At present, there is a belief that

such issues will need to be explored cooperatively on a case by case basis,

but this belief has not yet crystallized into any sort of consensus. Neither has

a mechanism for weighing the bona fides of a proposal acquired substantial

support.

A second item, and one as to which, again, there remains a lack of

consensus within PAB, is whether the restrictions on the qualifications of

registrars which currently are applied are appropriate. There is substantial

support within PAB for reducing the threshold qualifications, on the theory

that the presence of unqualified participants in the registration process will

be a self-limiting problem: such registrars will be unable to satisfy the public's

demand for services and will therefore fall by the wayside. The initiative for

lowering of threshold qualifications has not developed sufficient support

within PAB to be proposed to the interim Policy Oversight Committee. A

substantial number of PAB members believe that the qualifications presently

set are needed to assure consistent service delivery and build up public

confidence in the gTLD system, which is, at present, an untried service

provider. The traditional Internet paradigm of "deliver . . . use . . . refine" may

well be the key to the evolution of the Council of Registrars' service delivery

policy. It is entirely possible that once the gTLD system has proven its ability

to deliver consistent domain name registration services, the threshold can be

lowered incrementally to enhance participation in the system. At present,

however, it appears that the threshold qualifications, discussed elsewhere,

will be left in place for at least the immediate future.

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law

trademarks, geographic indications, etc.) if any, should be protected on the

Internet vis-a-vis domain names?

At one extreme, the position has been taken that domain names have

nothing to do with intellectual property. That sort of position has appeal for

engineers and technologically-oriented but socially-challenged persons

("techno-geeks"), but it is out of touch with the reality that domain names'

association with intellectual property rights is what creates the economic

value in the domain name system itself.

The question posed in the NOI cannot be answered in isolation from the

threshold issue: by whom the Intellectual property rights in question will be

protected. The values to be protected will vary depending on the agency

which supplies the protection. There are three immediately-apparent

resolutions of the threshold issue, and they are not mutually exclusive:

(a) the domain name system itself;

(b) national court systems; and

The question also cannot be answered in isolation from the context of the

ongoing dispute over the relationship between domain names and intellectual

property rights. The relationship is not a simple one. Policy-makers in this

arena must distinguish among:

(a) use of a domain name by a business which happens to coincide

with a registered trademark in a different product or jurisdiction

(b) the registration of a domain name by an company for

"warehousing" purposes and

product or service for the purpose of confusing the consumers or disparaging

the product or service that is suggested by the domain name.

That the domain name system can and should stamp out the warehousing

of trademark-related domain names is explicit in the Memorandum of

Understanding relating to Generic Top Level Domains. The pernicious

practice of "Cybersquatting" adds nothing of value to the domain name

system. Likewise, the use of domain names by competitors in order to

confuse the public or disparage the goods and services that have a logical

connection with the domain name is contrary to the best interests of the

public and the Internet. When it is patent that this sort of unfair competition

was the motive for the registration of a domain name, there is no good

reason for the domain name system to await the adjudication of a court,

national or international; the domain name system can and should act in the

best interest of the public and the Internet.

However, the resolution of competing intellectual property claims is not

something which a domain name system is equipped to resolve. There can

be only one "united.firm," and there is no reason for the domain name system

to displace the first registrant of that name. Neither is it proper for a domain

name system to set itself up to decide whether a holder of Intellectual

property rights has sufficient rights for the recognition of those rights.

Whether a claimant has a common law trademark, or other intellectual

property rights deserving of recognition, is a matter for courts and legislators,

not for the domain name system.

The task is complicated by the fact that there is no authoritative international

trademark registry. PAB believes that the ACP Substantive Guidelines

balance the competing concerns as well as any non-governmental solution

can hope to achieve. It consciously refrains from the judicial function of

deciding whether a claimant enjoys common law Intellectual property rights

related to a domain name. It also consciously refrains from a mechanical

determination that the earliest trademark registration is dispositive of the

domain name dispute. Rather, it charges the Administrative Challenge Panel

with a commonsensical approach to ascertaining the strength of Intellectual

property rights and their proper area of relevance. Thus a trademark for

plush toys should not per se trump the right of an established business to

use its own name on the Internet. PAB specifically disapproves of the

manner in which Network Solutions, Inc. handled the Roadrunner.com

domain name dispute.

In short, the rights that should be protected by the Domain Name System

itself are those which are created or confirmed exogenously to the Domain

Name System. This leaves to nations (and to the extent applicable,

international organizations) the basic legal authority to create and confirm

Intellectual property rights. There does remain a danger of inconsistent

adjudications between different countries. This is not a new phenomenon in

international relations. It was with a recognition that conflicts might arise that

the gTLD regime sought to seat the domain name resolution process in the

Swiss Confederation, a country which has consistently demonstrated special

sensitivity to international relations while at the same time remaining aloof

from entanglements.

The rights that should serve as the basis for protection on the Internet, when

the tribunal is a court, remain a matter for each nation to determine. National

courts determine the obligations of the parties before them, and can affect

the operation of the domain name system indirectly. The Domain Name

System itself would obey the mandates of its governing national court, and

carry out the directives of other national courts only as implemented by the

actions of the domain name holders themselves. Presumably, a registrant

who was ordered by a competent court to surrender a domain name would

inform the registrar, and in that way, the DNS, acting through CORE, would

implement the national court decision, but only indirectly. This is an

imperfect solution, but until wide acceptance for an international convention

can be found, it is the only presently-workable solution.

22. Should some process of preliminary review of an application for

registration of a domain name be required, before allocation, to determine if

it conflicts with a trademark, a trade name, a geographic indication, etc.? If

so, what standards should be used? Who should conduct the preliminary

review? If a conflict is found, what should be done, e.g., domain name

applicant and/or trademark owner notified of the conflict? Automatic referral

to dispute settlement?

It is inappropriate to impose on a Domain Name System the

quasi-governmental obligation of reviewing a domain name to ascertain if it

conflicts with Intellectual property rights. There is no easy answer to the

question of how many nations' Intellectual Property registries would need to

be consulted, nor which nations' registries those should be. Neither is there

an easy or logical way to specify, in advance, what to do where the same

Intellectual property rights are vested in different entities in different

jurisdictions.

PAB believes that a prospective registrant should assume the responsibility

of reviewing the Intellectual property rights relevant to its proposed domain

name before investing in the development of the domain. It also believes

that a registrant should have the right --but not the obligation-- to expose a

proposed name to public comment before activating the domain. While such

pre-activation review should not be conclusive, or give the registrant rights

to the domain which are incontestable, the practice of voluntarily submitting

the domain name to public comment is a factor which should be taken into

account in deciding whether to allow the domain name to persist. It remains

one factor among many.

PAB believes strongly that any domain name policy should be drafted and

applied with the objectives of (a) preventing warehousing, (b) precluding the

use of confusing domain names to obscure the authorship or source of

material on the Word Wide Web and (c) encouraging the growth and

development of the Internet.

23. Aside from a preliminary review process, how should trademark rights be

protected on the Internet vis-a-vis domain names? What entity(ies), if any,

should resolve disputes? Are national courts the only appropriate forum for

such disputes? Specifically, is there a role for

national/international governmental/nongovernmental organizations?

As indicated above, the self-governance of the Internet is self-limiting. The

gTLD-MoU scheme recognizes this and sharply limits the role which can be

played by the Administrative Challenge Panels ("ACPs"). The objective is to

ensure that the ACPs do not do more than govern the internal operation of

the gTLD-MoU system. The ACPs cannot make decrees which establish the

existence of Intellectual property rights, or mandate the performance of an

act by a domain name registrant or challenger. The final and conclusive

resolution of disputes requires the application of positive law. Courts, in

particular, are limited by basic notions of due process in acquiring and

exercising jurisdiction. Once the process which is due has run its course, the

parties to the process are entitled to have the resulting decree respected and

obeyed. Ultimately, the question whether a domain name dispute has been

conclusively resolved will be determined by the courts which have jurisdiction

over the registrar. For example, a recent notorious domain name dispute

centered around the registration of the yahoo.com.tw domain. The decision

to shut off that domain was taken by the registrar responsible for the Taiwan

TLD. If the holder of the yahoo.com registration had chosen instead to

institute a court proceeding in a jurisdiction other than Taiwan, it would have

fallen to a Taiwanese court to determine whether that other court had

properly exercised personal jurisdiction over the holder of the Taiwanese

domain.

The foregoing states the minimum level of government involvement in

domain name disputes. Whether a greater level of governmental

involvement is called for, and, if so, what form that involvement should take,

is a remarkably complex question. PAB believes that the international

organizations whose briefs involve electronic communication --including ITU

and WIPO-- are the proper fora in which to consider the multifarious policy

considerations bearing on this question. PAB also believes that United

States policy in this respect should be based on recognition that the United

States, as one of the world's premier commercial markets, has a strong

interest in promoting the stability, predictability, and robustness of IP

protection.

It remains important, however, that the United States advance these

objective through international cooperation. The United States has the power

to play the role of a 350-kilo

gorilla with respect to domain names and the Internet generally, but that

would not be in the public interest. It could well lead to fragmentation of the

Internet, as other countries' national pride and the fierce independence of

Internet users generally would react unfavorably.

Accordingly, PAB recommends to the United States that its influence over the

domain name system be exerted through the existing international

organizations, particularly WIPO and ITU.

24. How can conflicts over trademarks best be prevented? What information

resources (e.g. databases of registered domain names, registered

trademarks, trade names) could help reduce potential conflicts? If there

should be a database(s), who should create the database(s)? How should

such a database(s) be used?

The prevention of trademark-domain name collisions is complicated by the

lack of a universal registry for trademarks, and the fact that identical

trademarks can be held by different entities in different jurisdictions or in the

same jurisdiction with respect to different goods and services. Because the

prevention of disputes does not have an easily-identifiable basis, PAB does

not believe that any sort of mechanical rule is appropriate.

While the issues involved here are analogous to those entailed in the use of

"trademark watch" services, PAB does not believe that a registrant should be

compelled to secure the services of an outside contractor or search firm as

a prerequisite to registering a domain name. PAB believes that the domain

name registration process should be as streamlined as possible, and does

not believe that the existing system --which places on the trademark holder

the burden of "policing" the marketplace-- should be revised or departed from

solely in the context of Internet domain names. When and if the international

trademark regime changes to remove or relieve that burden, then it would

naturally be incumbent on the domain name system to revise its policies and

procedures. Until that happens, and except to the extent that "pro-active"

exclusion, discussed in the next paragraph, is appropriate, it will and should

remain the responsibility of trademark holders to preserve and protect their

intellectual property rights.

At the same time, certain disputes can be prevented through the pro-active

recognition that certain marks are so strong that no domain name which is

identical or closely-similar to a given mark should be allowed. PAB believes

that the decision whether pro-actively to exclude certain domain names from

ever being registered can be made only on a case by case basis.

Even if a "perfect" database could be compiled and kept current, it could not

dispose of these problems. A standard of "confusingly similar" is not

susceptible of automatic or rote application. A lesser standard, such as

"identical," errs in the opposite direction by failing to afford protection to many

legitimate trademarks.

25. Should domain name applicants be required to demonstrate that they

have a basis for requesting a particular domain name? If so, what information

should be supplied? Who should evaluate the information? On the basis of

what criteria?

Part of the charm of the domain name system is the imaginativeness of

names which can be selected by registrants. PAB does not believe in

general that a barrier should be erected in all gTLDs. However, a gTLD

which is based on one's name (e.g., .NOM) can and should require a

showing of a connection between the name of the registrant and the domain

name. Likewise, where a domain has a particular charter (e.g., .museum) a

similar set of considerations applies. (In this context, "charter" refers to the

specific realm covered by a particular top level domain. A domain name

which would be offensive in a commercial domain, such as McDonald's.firm,

would not have the same problem in McDonald.nom, since the charter of the

.NOM domain is to allow individuals to use their proper names to identify their

own second level domains within that namespace.) However, the

requirement that a particular second level domain name be justified prior to

its registration would restrict the growth of the Internet by making the domain

name system subordinate to other sources of intellectual property rights.

By the same token, there is no harm in requiring a domain name registrant

to acknowledge, when requesting the activation of a second level domain,

that it has no claim to the name or reason for requesting it other than that he

likes the sound of the name. Such a policy would simplify the actions of an

ACP which is weighing the claim of a challenger that it has recognizable,

internationally-known intellectual property rights in and to the text that forms

the basis for the registered domain name. It is not expected that such a

consideration, standing alone, would be dispositive, but it can and should be

one factor among many.

26. How would the number of different gTLDs and the number of registrars

affect the number and cost of resolving trademark disputes?

Improperly handled, increasing the number of gTLDs promises to increase

the frequency of trademark disputes as well as to increase the cost of

policing and protecting the rights of trademark holders, by affording a larger

territory in which those disputes can take place. Likewise, an increase in the

number of registrars, if the registrars are charged with the resolution of

trademark-domain name collisions, promises to increase the cost of

protecting the rights of trademark holders as well as raising the spectre of

inconsistent determinations. While the gTLD-MoU system does not

contemplate that registrars will resolve trademark/domain name collisions,

that process may well be a feature of other domain name systems.

It is for these reasons that gTLD-MoU contemplates a system under which

the internal policy of the regime is administered consistently by a single

organization, separate and distinct from the domain name registration

process. These considerations also underlie the contemplated policy of

permitting pro-active exclusion of domain names based on the existence of

sufficiently-strong Intellectual property rights.

27. Where there are valid, but conflicting trademark rights for a single domain

name, are there any

technological solutions?

If the Internet and its domain name system were concerned solely with the

World Wide Web, then it would be reasonable and possible to implement

special domains which would subsume the disputed domains and permit

users to select among the contentious domains. For example, a request for

www.united.biz might refer the user to www.united.dis[putes], where the

system would serve a menu page with pointers to www.united-airlines.com,

www.united-vanlines.com, and www.united-video.com. The problem with this

approach is that the Domain Name System must also serve the needs of

other protocols, including telnet, rlogin, ftp, and e-mail. Those facilities

-especially e-mail-require automatic functioning rather than a human-driven

menu selection. Until such time as the Internet is redesigned on a clean

sheet of paper -permitting recognition of the World Wide Web as a mass

medium distinct from the other protocols of the Internet, there does not

appear to be a technological "fix" for the domain name system that will

preclude collisions with trademark and other Intellectual property rights.

28. Are there any other issues that should be addressed in this area?

PAB believes that a number of related issues are presented, but have been

addressed in the foregoing paragraphs.

**************************************************

Addenda

Addendum of Werner Straub

The need [for specialized domains] does exist and is probably

much greater than that for .rec, for instance.

You can define need as the difference

between cost avoided and cost incurred. A predefined TLD like

.air incurs no cost (as there are no conflicts). The absence of the

.air TLD is costly (airports need to fight for their names in the

unstructured name spaces and advertise their names

because there is no pattern). A predefined space like

air.int is fine, but it does not cost anything less than a .air, so

why bother millions of people with a useless suffix, knowing

that half of those who can remember jfk.air will not be able

to remember jfk.air.int? The same is true for any number of

other potential predefined or managed name spaces, be

it .post or .isbn or .barcode.

It would be much better to argue in favour of these to be

handled by the CORE registrars than saying that they

are not needed because they have not been addressed

by the gTLD-MoU. Moreover, managed name spaces

are a case for industry self-regulation just like the rest

of the domain names.

Addendum of Alan Sullivan

While some members of PAB recognize that a different structure of the

.us (allowing second level functional domain names) would go far to prevent the

pressure

experienced by the current .com, .net, and .org domains, Top Domain Registry

(TDR) does not agree that the opportune moment for including functional second

level .us domains has passed. TDR also believes that opening up the gTLD name

space in a controlled manner provides many more benefits than liabilities. TDR

does not know how US companies would react to functional second level domains

such as .com.us, .net.us, .or .org.us. since these domains have never been

available to US companies. TDR believes it would be healthy for the .us domain

to adopt a functional second level domain name structure, concurrently to opening

up the gTLD

name space.

Addendum of Dan Busarow

DPC generally supports this response. The only substantive difference

is that I do not believe that new gTLDs will ease name collisions.

Competition and portability are the saviors here.

###

Number: 357

From: Vince Wolodkin <wolodkin@digitalink.com>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 4:07pm

Subject: Docket No. 970613137-7137-01

Thank you for taking the time to include my comments on this important

issue.

Vince Wolodkin

(703)469-3128

wolodkin@digitalink.com [WORK]

wolodkin@erols.com [HOME]

What follows below are my answers to the questions posed in the NOI, docket

number 970613137-7137-01, Registration and Administration of Internet

Domain names. I am a private citizen of the United States, and the opinions

I put forth are not necessarily those of my employer.

I have been involved professionally with networking systems since 1989 and with

the internet itself since 1992. I was chiefly responsible for putting the

Federal Communications Commission on the Internet. When I left the FCC in

January 1996, I was acting Chief, Network Operations. I am currently employed

as the Senior Engr, Operations for washingtonpost.com. I have been active in

discussions regarding internet governance and domain naming for more than a

year now, and feel very strongly that the ITU approach currently touted by ISOC

is the wrong model for the future growth of the Internet. I believe that TLDs

should be nationalized, and that gTLDs should be removed, not expanded.

B.1) Domain registration is currently handled by a stable, pseudo-regulated

organization which is an advantage. On the other hand, the policies adopted

by NSI, presumably acceptable to NSF, seem illogical at best. The domain

dispute policy is biased and the mad domain speculation market has been

driven by an NSI who once able to accept money for registrations seemingly

stopped using any discretion when allocationg domains. Prior to charging a

$50 fee, NSI questioned organizations who wanted more than one domain name

and suggested that a third level domain was more appropriate.

B.2) By making them more like other systems world citizens deal with everyday.

At first glance, NSI has a world-wide monopoly on domain registrations, but a

closer look reveals hundreds of national registries for ISO country code

domains. Encouraging these ISO registries and minimizing or removing the current

gTLDs may yield good results. More on this in later sections.

B.3) Administration of domains should, at least for a time, be a regulated industry

in the US. Licensing for-profit registries under the .US domain would be an

acceptable solution.

B.4) I'd like to see Domain names be national systems. gTLDs necessarily

require international governance and yet, the internet is not ready for

such a governance. Natioanl name registration systems under country codes

could be a much better system.

B.5) gTLDs should be retired at such a time as the US can be prepared to

license registries under the .US or .USA domain(depending on a choice of ISO

2 or 3 character codes). At such time as registries are ready to operate

under .US, .com et al should be closed to new registrations and moved to

.com.us where all new registrations would go. Current registrants would

be allowed to make use of .com et al for a period of 5 years while

simultaneously using .com.us. THERE WOULD BE NO TRANSFERS ALLOWED during

this 5 year period. Anyone wishing to transfer a .com domain name would be

removed from .com and available only under .com.us.

In essence a gTLD only complicates issues of trademark and governance. There

is no clear ruling body. On the other hand national domains like .US and .UK

provide a clear direction for trademarks. The domain prince.co.uk would NOT i

conflict with prince.com.us. Both can have a valid mark in their own countries.

B.6) Currently, the only REAL problem with domain registration is that next

spring, the NSI/NSF agreement expires, and by that time a system needs to be

in place to continue registrations or the NSI agreement needs temporarily

extended until registrations under .US can take place. The issue of root

servers is important and will need addressed. I see no reason why the US

govt should continue to fund root servers for the world. In fact, under

a national domain system, each country could have their own roots. I

personally see no value multiple "Root Server Confederations". I cannot

grasp how such efforts would do anything but fragment the internet.

B.7) If systems are national, then we only need to worry about our own roots.

In fact, every licensed registry could be required to operate a root. Under

a national domain system, scalability will not be so much of a problem as it is

under a gTLD system.

B.8) NSI could conceivably stay in place to operate .com.us. I know there is

a great deal of resentment toward NSI in the community, though these are mostly

trademark lawyers and people who want to run .com themselves. NSI, if properly

managed, isn't that bad. Other registries would be allowed to serve things like

.guns.us or .butter.us.

B.9) It is very important to move away from gTLDs altogether. They are quite

frankly the wrong model for a global system. I can't think of any similar global

systems in existence today. All similar systems seem to deal with numbers or

frequencies and are more akin to the IP allocation problem.

C.10) Really, there are only policy considerations. There is no real

technical reason to limit the number of gTLDs. There is, however, no good

policy reason to have gTLDs. gTLDs require international governance that

the Internet is just not ready for. National TLDs, however, allow a more

measured approach.

B.11) No. Additional non business class or geography specific TLDs merely

exacerbate existing trademark issues.

B.12) Once again, there are only policy and business issues. Your average

everyday internet user couldn't care less what websites he visits are named,

it is the business, legal and registration interests that desire lots of

gTLDs. They are really not necessary.

B.13) No, they are not. As long as gTLDs are around, country codes will

not prosper because business(or should I say marketing) prefers catchy names

that end in .com. As I have said, Joe Average doesn't care whether he connects

to a website at ntia.gov or ntia.gov.us. In fact, it is quite possible to

allow users within the .us domain to connect to ntia.gov.us by merely

typing ntia.gov.

14. Are there any other issues that should be addressed in this area?

B.14) Just to re-iterate that gTLDs are the wrong model for the internet.

Internet users are typically used to dealing with international and national

boundaries. Businesses are also familiar with this construct as are courts.

Don't allow the ITU and WIPO to destroy the Internet with bad policy.

B.15) There should be NO gTLD registrars. But registrars under .US should

be allowed control over a particular SLD. Perhaps others would be structured

with a shared philosphy. Exclusive and non-exclusive can exist together under

a national TLDs. There should be no gTLDs.

B.16) Requirements should be determined by the NTIA in a seperate request

after the decision has been made to abandon the gTLD system and move to the

more logical national TLD system.

B.17) No. In fact, any ISP that registers domains for it's customers could

act as a registrar as long as there is a meta-registry that they work through.

This is similar to NSI's current implementation of channel-partners.

B.18) As long as it is done intelligently, this should be fine. People of the

United States have long dealt with system renaming and renumbering with phone

numbers and area codes, the interstate system and 911 street re-naming. People

get over it. There is no point in continuing a faulty system like the gTLD system

when changes can be affected to make it right.

B.19) gTLDs should not be created. gSLDs should be created under national

domain names like .US and .CA. gSLD registrars should obviously not be allowed

to have exclusive control over a boatload of SLDs, but what these limits should be

really depends on many other things as yet undecided. Perhaps, vanity SLDs like

.IBM.US and .MCI.US will be allowed. These might be private SLDs in which case

a single registrar might exclusively control many private SLDs.

B.20) gSLDs should be created under ISO country code domains(either 2 or 3 character)

and NO gTLDs should be created.

B.21) Country code TLDs provide for better trademark protection. If you

want trademark protection under the .US TLD then you need a US trademark.

We have seen trademark disputes regarding .com decided in other countries

against US citizens. .COM is a US registry and should only be subject to

US laws. Move .COM under .US.

B.22) I don't see why this review should be done. The application of

registrant for domain name name should merely state something to the effect

that applicant does not knowingly make use of a trademark not belonging to

him in the construction of this unique domainname.

23. Aside from a preliminary review process, how should trademark rights be

protected on the Internet vis-a-vis domain names? What entity(ies), if any,

should resolve disputes? Are national courts the only appropriate forum for

such disputes? Specifically, is there a role for national/international

governmental/nongovernmental organizations?

B.23) I believe that national courts are the only appropriate venue for

these types of decisions. Should domain holders be required to seek an

international body to rule in domain disputes. The gTLD model is the

wrong model.

B.24) It seems that should a database be created that it would be in the best

interest of WIPO to create it and make it freely available to potential

domain holders to search. Perhaps, domains could be checked against the

database and an email sent to the registering party such as the following:

"During the registration process for your requested domain XYZ.COM.US, we have

discovered that XYZ is a trademark of the following companies. If you wish to

proceed with this registration, please respond affirmatively to this message"

B.25) No, domain requesters should not need a trademark to have a domain

name. In addition, if a mark holder registers the name XYZ.COM, then first-come

first-serve should dictate that the name goes to the first holder. People

who use a domain and choose not to register it as a trademark should get what those

throughout history have gotten who fail to trademark their product names.

B.26) gTLDs will only make the matter worse as more and more trademark

disputes will arise. On the other hand, an intelligent system of

classification such as by SIC code, would go a long way to removing

disputes altogether.

B.27) No.

B.28) Just to re-iterate. gTLDs are the wrong model. .COM should be closed

to new registrations and the US govt should take up the task of licensing

registrars under .US and place all of .COM, including foreign registrations,

under .COM.US. Foreign registrations are currently a very small percentage

of .COM. If we close the TLDs(other than .INT) then the world will follow.

It's the right thing to do. If you value the future of the Internet you

will make the move to national domains. gTLDs have no long term future and

will only exacerbate the current situation. I don't think it's necessary

to make things worse before we make them better.

Thank you for your time,

Vince Wolodkin

###

Number: 358

From: Al Gilman <asgilman@access.digex.net>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 4:40pm

Subject: Commenting on DNS reform [Docket No. 970613137-7137-01]

This email is in response to:

> X-URL: http://www.ntia.doc.gov/ntiahome/domainname/dn5notic.htm

> DEPARTMENT OF COMMERCE

> [Docket No. 970613137-7137-01]

> Request for Comments on the Registration and Administration of

> Internet Domain Names

> A. Appropriate Principles

> The Government seeks comment on the principles by which it should

> evaluate proposals for the registration and administration of Internet

> domain names. Are the following principles appropriate? Are they

> complete? If not, how should they be revised? How might such

> principles best be fostered?

Comment: This statement implies an over-broad Government role in making

the final determination. It would be more consistent with the Government's

stated intention to say "The Government seeks comment on the principles

by which it should identify the Government interest in the allocation

of Internet domain names and current and proposed practices for the

maintenance (registration and administration) of this namespace."

There is a missing principle: The Government should pursue continuing

clarification of concepts for the concerns (rights and performance

parameters) where it has a lead or follow role to pursue, and to

clearly distinguish these from the concerns where it is most important

for the Government to get out of the way.

Fairness in trade is a concern where there is clearly a Government role.

There is no way we can say that the Government should play _no_ role in

policing the use of names in Internet communication particulary as it

increasingly becomes a platform for commerce.

The conceptual framework exposed by the set of questions asked

here does not line up perfectly, yet, with the boundaries between

concerns where the Government should lead, follow, or get out of

the way. Inadequate assertiveness with regard to proper

Government missions is as evident here as meddling in network

internals that should not be a matter of Government concern.

Nobody has all the answers yet. Thank you for listening.

> B. General/Organizational Framework Issues

> 2. How might current domain name systems be improved?

New top-level domains should be introduced. These domains should

be administered under a variety of business rules. In

particular, experiments should go forward for commercial TLDs

where the principles of operation, beyond a very limited set of

services guaranteed for all domains in the system, would be

determined by the proprietor of the name-service venture vending

the service of using these names. At least three such domains

should be created, licensed for finite periods to licensees

selected by auction. These top-level-names should have mnemonics

that one can remember, but which _do not_ imply any functional

distinction between the different competitors. Functional

pigeonholing should be practiced for governmental and NGO

addresses, but the commercial ventures must distinguish

themselves by service features outside the domain name.

For a simple example of how the service might vary in this

system, there could be analogs of 800 numbers which are free long

distance calls within a restricted service region. The flatness

of service with regard to everything that looks like a current

DNS name should not be assumed.

Flat-access name spaces should be maintained and the

what-the-market-will-bear, fee-for-service domains forced to

create a service so attractive that it draws market share away

from the current vanilla grade of service.

> 3. By what entity, entities, or types of entities should current

> domain name systems be administered? What should the makeup of such an

> entity be?

The current reform initiative should be supported. [Process started

by the ad-hoc committee.]

> 4. Are there decision-making processes that can serve as models for

> deciding on domain name registration systems (e.g., network numbering

> plan, standard-setting processes, spectrum allocation)? Are there

> private/public sector administered models or regimes that can be used

> for domain name registration (e.g., network numbering plan, standard

> setting processes, or spectrum allocation processes)? What is the

> proper role of national or international governmental/non-governmental

> organizations, if any, in national and international domain name

> registration systems?

There should be different classes of TLDs within which the

decision processes and authorities are different. The ISO

national TLDs should be administered by national

authorities. There should be a clearly identified domain (Top

level or short tree) for inter-governmental and non-governmental

bodies whose principal line of business is global, such as the UN

and the WTO, ISO, IEC and IUCN. To get one of these names, you

should have to meet credentials. Likewise, in the US there is a

recognizable class of non-profit enterprises which should have

access to a restricted namespace. But we should not try to

retread .ORG to make it pure in this sense. This should be done

by an enterprise of the US Government because the _defacto_

discriminant for this class of enterprises is tests defined and

applied by the tax authorities of the Federal Government. So

this namespace should be a .foo.US namespace for recognized

non-profit organizations that want to take advantage of their tax

status in connecting with Internauts.

> 5. Should generic top level domains (gTLDs), (e.g., .com), be retired

> from circulation? Should geographic or country codes (e.g., .US) be

> required? If so, what should happen to the .com registry? Are gTLD

> management issues separable from questions about International

> Standards Organization (ISO) country code domains?

Generic TLDs should be preserved and geographic naming should be optional.

On the other hand, under the commercialization strategy outlined above,

the service available under the commercial new TLDs will eventually be

so good that the demand for the generic TLDs will plateau or decline, and

their rapid growth of late will not be a continuing problem.

The management issues for country-code domains can be _partly_ separated

from those of gTLDs or domaines mondiales. The healthy model is that

there is a progressively diversified class structure for names where

all domain name spaces are derived from a root class, and end-user services

and governance-doctrine gradually diverges as one gets farther from the

root level.

> 6. Are there any technological solutions to current domain name

> registration issues? Are there any issues concerning the relationship

> of registrars and gTLDs with root servers?

Yes. See above. Yes. But not Government Issues.

> 7. How can we ensure the scalability of the domain name system name

> and address spaces as well as ensure that root servers continue to

> interoperate and coordinate?

The Government should get off the issue of root servers. Root servers

should be retired in favor of redundant hyperreliable administration

systems, but the Government should be a distant follower in this process.

> 8. How should the transition to any new systems be accomplished?

Gradually, under constant evaluation.

> 9. Are there any other issues that should be addressed in this area?

Taxation of for-profit namespace ventures to support the whole.

> C. Creation of New gTLDs

> 11. Should additional gTLDs be created?

Yes.

> 12. Are there technical, business, and/or policy issues about

> guaranteeing the scalability of the name space associated with

> increasing the number of gTLDs?

You have to take some risks. Guaranteed scalability is a perfection

we probably cannot afford.

> 13. Are gTLD management issues separable from questions about ISO

> country code domains?

See above. Progressively varigated namespace is the model that both

integrates and distinguishes these subclasses of domains.

> 14. Are there any other issues that should be addressed in this area?

> D. Policies for Registries

> 15. Should a gTLD registrar have exclusive control over a particular

> gTLD? Are there any technical limitations on using shared registries

> for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

The government should focus on isolating service characteristics to be

defended and leave the algorithmic solution to the providers.

In other words, I believe that this question is too "inner" to the

realization of Internet service to be a proper concern of the public

or Government.

> 16. Should there be threshold requirements for domain name registrars,

> and what responsibilities should such registrars have? Who will

> determine these and how?

Some trustability should be required. How this is articuated is TBD.

Does not necessarily constrain _who_ the registrar is if there is a

viable audit plan for checks and balances.

> E. Trademark Issues

> 21. What trademark rights (e.g., registered trademarks, common law

> trademarks, geographic indications, etc.), if any, should be protected

> on the Internet vis-a-vis domain names?

We need to continue to grope our way forward on this one. What

rights should pertain is actually a function of what services and

operations are possible. And we don't know that yet.

There are rights on both sides that cannot be made entirely clear

all at once. Eventually, it is possible that parties making

commercial use of the Internet should for that reason have to

make certain disclosures that will facilitate the prosecution of

claims against their business unit. But this is not reduceable

to final writ at this time.

> 25. Should domain name applicants be required to demonstrate that they

> have a basis for requesting a particular domain name? If so, what

> information should be supplied? Who should evaluate the information?

> On the basis of what criteria?

In general, no. In the specialized cases outlined above, yes. In the

commercial-enterprise domains, this is to be answered by the entrepreneurs.

> 26. How would the number of different gTLDs and the number of

> registrars affect the number and cost of resolving trademark disputes?

Minimally.

> 27. Where there are valid, but conflicting trademark rights for a

> single domain name, are there any technological solutions?

There are possible solutions. These are a blend of technology and

business negotiation. No pure technological instant fixes.

###

Number: 359

Before the:

U.S. DEPARTMENT OF COMMERCE

National Telecommunications and Information Administration

Washington, DC 20230

In the Matter of :

REGISTRATION AND ADMINISTRATION OF

INTERNET DOMAIN NAMES

Docket No. 970613137-7137-01

Comments of Paul M. Kane,

Director,

Internet Computer Bureau Plc, London.

paul.kane@icb.co.uk

I, respectfully submit the following comments in this proceeding.

I have been a Director of the above company since July 1996, having worked in the computer industry as a systems analyst/designer since 1983.

I have participated as an invited attendee in at nine public and private conferences at national and international level on this issue.

Paul M. Kane

15th August 1997.

Summary of Comments:

Internet Governance:

Top Level Domains.

1. The DNS architecture is monopolistic by nature and consequently the number of gTLDs should be increased, with competition amongst shared and exclusive Registries.

1. The March 1998 deadline has focused many entrepreneurs minds' as all proposals ultimately result in a single entity having exclusive power over a specific gTLD namespace.

1. Evaluate the other gTLD proposals over a longer time frame, exclusive and share gTLDs can co-exist providing a firm basis on which to stimulate competition in the Domain market.

Consumer Protection

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name registration systems?

ISO 3166 Registration System

Advantages: Disadvantages:

Efficient

Uncomplicated Controlled monopoly

Regionally Accountable

Frequently cost recovery basis

Distributed Processing

InterNIC Registration System

Advantages: Disadvantages:

Efficient high profit /cost ratio

Uncomplicated Monopolistic

Not regionally accountable

Processing failure/delay affects

global network

2. How might current domain name systems be improved?

DNS system

3. By what entity, entities, or types of entities should current domain name systems be administered?

What should the makeup of such an entity be?

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)?

What is the proper role of national or international governmental/non-governmental organisations, if any, in national and international domain name registration systems?

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation?

Should geographic or country codes (e.g., .US) be required?

If so, what should happen to the .com registry?

Are gTLD management issues separable from questions about International Standards Organisation (ISO) country code domains?

6. Are there any technological solutions to current domain name registration issues?

Are there any issues concerning the relationship of registrars and gTLDs with root servers?

1. From a political perspective operating CORE as a distributed model enables registry contractors to be spread over a number of countries which fosters a truly international Domain.

7. How can we ensure the scalability of the domain name system name and address spaces as well as ensure that root servers continue to interoperate and co-ordinate?

1. More gTLDs - both distributed and monopolistic

1. Navigation requires effective directory services to identify the site in the virtual world known in the physical world.

8. How should the transition to any new systems be accomplished?

Are there any other issues that should be addressed in this area?

1. Increasing the gTLDs name space should be considered supplementary name space rather than an Internet governance issue.

1. IAHC seeks to build a gTLD reputation for .STORE, .WEB, .PER, .ARTS, by determining the parameters of allocation within the designated name space.

1. The Internet is driven by private enterprise. Satisfying and protecting customer interests on the Internet, should be the role of governance.

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

1. Yes.

2. Technical: - the caching rate of the DNS System, when gTLDs are in the tens of thousands.

2. During July consumers were confused when identical names in different gTLDs were used with one service provider emulating the other. Consumers were duped into paying twice the normal price.

11. Should additional gTLDs be created?

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

2. The crux is to be able to identify the shop known to the consumer in the physical world in the virtual world, irrespective of its Domain Name.

3. The technology exists but is not being applied, because the payback time to recover the investment is too long for most private enterprises.

13. Are gTLD management issues separable from questions about ISO country code domains?

14. Are there any other issues that should be addressed in this area?

D. Policies for Registries

15. Should a gTLD registrar have exclusive control over a particular gTLD?

Are there any technical limitations on using shared registries for some or all gTLDs?

Can exclusive and non-exclusive gTLDs coexist?

1. Yes, and their viability will depend on the reputation, quality of service and price of each gTLD.

16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have?

Who will determine these and how?

17. Are there technical limitations on the possible number of domain name registrars?

1. No. The more Registrars the more complicated the process for the Registry to monitor progress.

18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

19. Should there be a limit on the number of different gTLDs a given registrar can administer?

Does this depend on whether the registrar has exclusive or non-exclusive rights to the gTLD?

20. Are there any other issues that should be addressed in this area?

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-à-vis domain names?

23. Aside from a preliminary review process, how should trademark rights be protected on the Internet vis-à-vis domain names?

Preliminary review inherently requires discretionary judgement. This could be contention and lead to litigation. In addition it is futile, please see 22 above.

What entity(ies), if any, should resolve disputes?

71. Mediation, Arbitrators and the Courts.

Are national courts the only appropriate forum for such disputes? No

Specifically, is there a role for national/international governmental/nongovernmental organisations?

72. Yes

24. How can conflicts over trademarks best be prevented?

73. Use technology to marginalise the counterfeiters and fraudsters. We have an operational model and procedures in place today!

What information resources could help reduce potential conflicts?:

74.iii) trade names - difficult to assimilate, and of questionable value in the domain registration process due to many identical names.

75. If there should be a database(s), who should create the database(s)? How should such a database(s) be used?

25. Should domain name applicants be required to demonstrate that they have a basis for requesting a particular domain name?

78. Yes, to obtain additional support from the Registry if the Domain Name is subsequently challenged.

If so, what information should be supplied?

Who should evaluate the information?

80. The Domain Name Registry not the Registrar.

On the basis of what criteria?

81. One domain name per person or trading entity or brand.

26. How would the number of different gTLDs and the number of registrars affect the number and cost of resolving trademark disputes?

27. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

83. Yes. That is what this company has been working on since I attended the WIPO meeting in May. We now have a working model !

END of submission.

###

Number: 360

From: Kaye Caldwell <kaye@ix.netcom.com>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 4:47pm

Subject: Domain Name Comments [Docket No. 970613137-7137-01]

Attached as both a MSWord v. 7 file and a PDF file are comments from

CommerceNet on The Registration And Administration Of Internet Domain Names

[Docket No. 970613137-7137-01].

In the event that there is a problem reading the files as transmitted via

e-mail they are also available on the Internet at:

http://www.softwareIndustry.org/commercenet/dncomments.html

Thank you,

Kaye Caldwell

4005 Miranda Avenue
Suite 175
Palo Alto, CA 94304
415.858.1930

fax: 415.858.1936
E-MAIL: info@commerce.net
http://www.commerce.net

Comments Filed In Response To

Department Of Commerce

Request For Comments On The Registration And Administration Of Internet Domain Names

[Docket No. 970613137-7137-01]

August 18, 1997

CommerceNet is the leading industry consortium dedicated to accelerating the growth of Internet commerce and expanding Internet markets. Launched in California's Silicon Valley in April 1994, CommerceNet's membership has grown to over 150 leading organizations in areas such as banking, electronics, computers, online services, information service industries, as well as major end users. Together, we are transforming the Internet into a global electronic marketplace.

The Domain Name system is a significant component of the Internet. It is an essential part of making the Internet user-friendly in that it converts a technical numeric-only Internet location indicator to an indicator which is frequently easy to determine simply by educated guessing. Unfortunately, the increasing popularity of the Internet has led to disputes over registration and ownership of specific domain names. CommerceNet appreciates the Department of Commerce's interest in this matter and further appreciates this opportunity to comment on the administration of the Internet Domain Name system. We will restrict our comments to the U.S. Domain Name system and its interaction with the domain names systems of other countries. We will not address the domain name systems of non-U.S. countries. In our comments we will use the numbering system provided by the Notice of Request for Comments, skipping those numbers for which we have no input.

A. Appropriate Principles

CommerceNet agrees in general with the statement of appropriate principles as expressed in the notice of request for public comments. We note however, that the principles as stated may not sufficiently address two issues which we suggest should be added or incorporated into existing principles:

1.) The domain name registration system should enable control of national domain name systems within each country while ensuring international interoperability of domain names.

No business, organization, or individual should be required to deal with a foreign entity or international entity in order to obtain a domain name. If that were required it would lessen the ability to obtain redress if an entity fails to ensure the easy availability of domain names.

2.) Care must be taken to ensure that the Internet remains accessible, as a medium for making information available, to entities of all types and sizes.

As the Administration's Framework for Global Electronic Commerce points out, "Citizens of many nations are finding additional outlets for personal and political expression. The Internet is being used to reinvent government and reshape our lives and our communities in the process." This aspect of the Internet must be preserved. While competition in domain name registration systems is an excellent goal, the danger may exist that individuals, non-profits, local government agencies, and other small non-commercial entities will receive a lower quality of service, or no service, as compared to large well-funded commercial enterprises. The government of each country, and in particular the U.S. Government, should ensure that such a result does not transpire. The value of the Internet is enhanced by the wide variety of uses it is amenable to. As in many network-type entities, the value of the network itself increases with each additional use or user the network gains. While CommerceNet is dedicated to accelerating the growth of Internet commerce and expanding Internet markets, we believe that the growth of Internet commerce and expansion of Internet markets is enhanced by the availability of non-commercial activity on the Internet.

CommerceNet also notes that principle (e) may be especially difficult to achieve. In resolving conflicts the rights of existing domain name holders must be balanced against the proprietary rights, particularly trademark rights, of others. We do not believe the correct balance has been reached. We also note that current existing policies regarding resolution of disputes over name allocation and management have not been subject to public participation in their development. We recommend that this oversight be corrected.

B. General/Organizational Framework Issues

1. What are the advantages and disadvantages of current domain name registration systems?

While obtaining a domain name through an ISP or other service provider is sometimes quick, convenient, and inexpensive, it is reportedly much more difficult to obtain a domain name directly. This needs to be resolved. In addition, continuous maintenance of a domain name can also be difficult. Problems have arisen regarding proof of payment, inactivation of domain names resulting from inadequate payment tracking and notification, and inactivation of domain names, without due process, resulting from claims of trademark infringement by other parties.

2. How might current domain name systems be improved?

Certainly better record keeping and notification procedures need to be put in place, indeed this may already have occurred. Name dispute mechanisms should be developed with public input and should balance the rights of existing domain name owners with the rights of trademark owners. Additionally, specific policies for treatment of trademarks must be developed. For example, those policies should NOT merely assume that a trademark holder has the right to "trademark".com or "trademark".org. As detailed further below (see response to question #21), we recommend that a trademark specific space be created in which trademark owners have specific rights and obligations within that space, and that those rights and obligations be isolated from non-trademark-specific spaces to the extent possible.

1. By what entity, entities, or types of entities should current domain name systems be administered? What should the makeup of such an entity be?

The entity or type of entity is not as essential as ensuring that the entities comply with the principles stated.

1. Are there decision-making processes that can serve as models for deciding on domain name registration systems (e.g., network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (e.g., network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

We suggest that the government processes for obtaining public input to rules and regulations are one model which could be used to ensure that domain names registration systems are fair.

5. Should generic top level domains (gTLDs), (e.g., .com), be retired from circulation? Should geographic or country codes (e.g., .US) be required? If so, what should happen to the .com registry? Are gTLD management issues separable from questions about International Standards Organization (ISO) country code domains?

In order to provide continuity, existing generic top level domain names should be preserved. However, serious consideration needs to be given to whether the continued issuance of such names should either be discontinued or based on a policy other than "first come first served." In particular, adding to the domain name space could resolve some controversy. See our response to question # 21 below for additional details.

1. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

Technological measures should be undertaken to ensure that domain names do not become inactive due to the failure, either technological or commercial, of a registrar.

9. Are there any other issues that should be addressed in this area?

The domain name space should be recognized as belonging to the public trust. All governments should take steps to guarantee the due process that will ensure the fair and unbiased availability of domain names.

C. Creation of New gTLDs

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created?

The total number of different gTLDs should be constrained by the need to provide a logical domain name space - for example, there should exist ONE logical space for a given company to inhabit as its trademarked domain name. The situation in which a commercial company feels that it must obtain all possible domain names which might possibly contain its trademarks should be avoided.

1. Should additional gTLDs be created?

CommerceNet believes that the created of additional gTLDs may be necessary in order to serve the need for a better classification system and possibly to resolve the trademark issues.

D. Policies for Registries

1. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

It is difficult to envision exclusivity and competition co-existing and for that reason exclusivity should be considered very carefully.

16. Should there be threshold requirements for domain name registrars, and what responsibilities should such registrars have? Who will determine these and how?

Registrars should adhere to a common set of principles, with a common oversight body. Service to non-commercial entities must be ensured.

20. Are there any other issues that should be addressed in this area?

Insofar as the domain name space belongs to the public trust and a domain name belongs to owner, a registrar must not be permitted to claim any intellectual property rights with respect to its database of domain names.

The question of registrar liability should also be addressed. Provided that an appropriate mechanism is developed for public input into policy development, and provided that a registrar follows that stated policy, then registrars should be protected from liability arising from disputes over domain names.

E. Trademark Issues

21. What trademark rights (e.g., registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-a-vis domain names?

Trademark rights vis-a-vis domain names should follow similar principles to trademark rights in the physical world, in particular, trademarks vis-a-vis domain names should NOT be stronger than in the physical world. If the domain name system needs to be modified in order to achieve this, then that modification should be undertaken. CommerceNet particularly supports the concept of creating a trademark domain name space with rules that are similar to the physical world's treatment of proprietary rights. For example, a mechanism could be devised to allocate domain names within a product-type classification system and to further allocate domain names geographically within the trademark domain name space. With such a separate trademark space, non-commercial use of common words as domain names could simply exist in a different domain name space. Thus Sun Microsystems and Sun Tea could inhabit the trademark space, while an educational web page on the sun could exists at sun.info, for example. Inside the trademark domain name space, the registrar should resolve name allocation disputes, while outside the trademark domain name space the resolution should be outside the dispute resolution policy of the registrar, in the courts or following other procedures appropriate to a particular country.

22 & 23. Should some process of preliminary review of an application for registration of a domain name be required, before allocation, to determine if it conflicts with a trademark, a trade name, a geographic indication, etc.? If so, what standards should be used? Who should conduct the preliminary review? If a conflict is found, what should be done, e.g., domain name applicant and/or trademark owner notified of the conflict? Automatic referral to dispute settlement? 23. Aside from a preliminary review process, how should trademark rights be protected on the Internet vis-a-vis domain names? What entity(ies), if any, should resolve disputes? Are national courts the only appropriate forum for such disputes? Specifically, is there a role for national/international governmental/nongovernmental organizations?

Unless a trademark specific domain name space is set up, preliminary review will only result in delay for non-commercial entities as well as commercial ones. CommerceNet recommends a trademark specific domain name space, with review immediately subsequent to allocation and a short time period after which disputes must adhere either to a dispute settlement mechanism or resort to the court system. However, in no case should a domain name holder be deprived of the use of its domain name without due process. Care must be taken to ensure that any agreed upon dispute settlement procedure is truly fair - either the mechanism must be developed with adequate public input or competition in dispute settlement mechanisms must be ensured by ensuring competition in registries (or both).

No, no basis should be required. Requiring a basis would limit free expression.

Conclusion

In conclusion, CommerceNet believes that accessibility to the Internet, as a medium for making information available, by a variety of entities is essential, whether those entities are commercial, individual, governmental, or any other type of entity. Rights of existing domain name holders must be balanced against trademark rights - and that balance may best be achieved by creating a separate domain name space for trademark holders. Furthermore, competition in domain name registries is a laudable goal, but in achieving competition universal access to the Internet, as a medium for information dissemination, must not be lost in a commercially competitive environment.

Submitted by:

Kaye K. Caldwell

Policy Director

Direct Dial: 408-479-8743 Direct Fax: 408-479-9247

E-MAIL: KCaldwell@Commerce.net

###

Number: 361

From: "Daniel J. Weitzner" <djw@cdt.org>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 4:48pm

Subject: CDT/ITAA/ISA Filing in DNS Request for Comment

August 18, 1997

Ms. Patrice Washington

Office of Public Affairs

National Telecommunications and Information Administration (NTIA)

Room 4898

14th St. and Constitution Ave., NW

Washington, D.C. 20230

Dear Ms. Washington:

In response to the July 1, 1997 Request for Comments on the Registration

and Administration of Internet Domain Names [Docket No.

970613137-7137-01], the Center for Democracy and Technology (CDT), the

Information Technology Association of America (ITAA) and the Interactive

Services Association (ISA) respectfully submit a report from the July

30-31, 1997 Forum on Internet Domain Names.

Report from the July 30-31, 1997 Forum on Internet Domain Names

The July 30-31, 1997 Forum on Internet Domain Names, which was sponsored

by CDT, ITAA and ISA, brought together over 100 representatives from

industry, government, international organizations, and users from around

the world to exchange perspectives on a range of issues including:

* the nature of domain name management and its reform

* trademark issues in domain name management

* governance issues for the domain name system; and

* operational and administrative concerns in considering transitions

to different domain name systems.

The Forum served as a means by which many varied perspectives on domain

names were aired and discussed. The enclosed Report contains a summary

of the Forum proceedings. For each issue covered, a short discussion is

included along with a list of the major areas of consensus among

participants and areas of divergence or identification of questions

still to be resolved.

Together, CDT, ITAA and ISA represent more than 11,000 members and

sponsors in the U.S. and the concerns of Internet users. Our members are

commercial and non-commercial providers and users of Internet services, all

of whom depend on the reliable, competitive, fair and accountable

administration of core Internet resources such as domain names.

We appreciate the opportunity to comment on this important issue and are

committed to working with the U.S. Government as well as other

organizations in working toward resolution of the complex issues

surrounding domain names.

Sincerely,

Jerry Berman, Executive Director, CDT

Harris N. Miller, President, ITAA

Jeff B. Richards, Executive Director, ISA

===============================================================

[MS Word File attached, ASCII text follows in next message]

CC: Harris Miller <hmiller@itaa.org>

Report from the

Forum on Internet Domain Names

Washington, D.C.

July 30-31, 1997

Sponsored by

Center for Democracy and Technology <http://www.cdt.org>

Interactive Services Association <http://www.isa.net>

Information Technology Association of America <http://www.itaa.org>

Introduction: Forum on Internet Domain Names

On July 30-31, 1997, a group of over 100 representatives of industry, government, and the public interest community met in Washington, DC to discuss the increasingly contentious debate over Internet domain names. In two days of discussion sessions participants exchanged perspectives on a range of issues including:

the nature of domain name management and its reform;

trademark issues in domain name management;

governance issues for the domain name system; and

operational and administrative concerns in considering a transition to a different domain name system.

This report contains a summary of the proceedings of the Forum on Internet Domain Names. For each of the issues covered, a short discussion is included along with a list of the major areas of consensus among participants and areas of divergence or questions still to be resolved. As the entire conference was off the record (on background for the press involved), this summary is not intended to be attributable to any specific individuals but rather to characterize the conversation more broadly.

In addition to this report. further background material including a "virtual briefing book" with pointers to most of the materials distributed during the Forum, can be obtained on the World Wide Web at:

http://www.cdt.org/dns

The organizers of the Forum thank the participants for their candid and well-motivated contributions to this debate. The Forum was created in the hope of providing a productive means of gathering the disparate interests involved to discussing the contentious issues . We look forward to contributing to a process of addressing this issue so critical to the long-term well-being and potential of the Internet.

Session 1: Domain Name Management and Its Reform

Moderator: Dr. Robert E. Kahn, Corporation for National Research Initiatives

Panelists: Don Heath, Internet Society, IAHC

Gabriel A. Battista, Network Solutions

Jay Fenello, Iperdome

Respondents: Brian Kahin, Office of Science and Technology Policy

Ken Fockler, CA Net

The domain name system was designed to meet the need of network users to address any public node on the network by means of a alphanumeric name lookup system rather than with an unintelligible number. At one time this mapping of names to addresses was simple to maintain through a centrally controlled and widely stored "hosts" file. Today the databases matching "domain names" with actual network addresses have grown into a widely distributed database system maintained through a series of "top level domain" (TLD) root servers. In addition to TLDs for each country (based on ISO 3166 codes such as .us, .ca (Canada), or .sa (South Africa)), particular attention had been paid to the so-called "generic" TLDs (gTLDs) -- .com, .org, .net. These gTLDs are not geographically specific, and have seen particularly high demand worldwide; they are currently administered by the Virginia-based company Network Solutions Inc. (NSI) under contract to the U.S. government.

It is widely accepted that a unified approach is needed to the Internet domain name system. A common approach is essential to maintain the stability and accessibility of the network. So far, it is not clear that there is any consensus on what that common approach should be.

Pressure to reevaluate the current system for allocating domain names within the gTLDs has come from several emerging tensions in the current system:

Congestion in the domain name space -- There are only a finite number of useful and convenient names. This has led to "collisions" between those wanting the same names.

Trademark conflicts -- The process for resolving domain names that correspond with trademarked names remains an open question. A single domain name may be sought by organizations with similar names each holding a legitimate trademark in the name.

Governance concerns -- The growing worldwide importance of the gTLDs has led to concerns about the monopoly over domain name allocation currently granted to NSI through its contract with the US government -- and with the level of US control over global domain name allocation.

These issues have become particularly acute with the expiration of NSI's contract with the US government in March 1998.

Several alternatives for DNS governance were discussed at the Forum's opening session:

IAHC model -- A detailed system for international, independent governance of the gTLDs as a public trust. Emphasizes the "public" ownership of gTLDs, and the need for a self-regulating system independent of national governments.

NSI model -- Support for a competitive, market-driven system developed with a strong role for the U.S. government. Emphasizes the need for a global solution with stability and reliability.

AlterNIC/eDNS -- Support for the ad hoc development of alternative TLDs and registries without central control. Emphasizes a decentralized, private sector approach, as typified by the AlterNIC and other ad hoc registry efforts.

There was no consensus expressed at this forum for any of these plans, but there is emerging and significant overlap in the rationale behind these plans.

Major Areas of Consensus:

All of the participants agreed that there is a need for an ongoing process to resolve this issue. While there is some disagreement as to how open the process has been to this point, all acknowledge that further discussion is needed to reach a critical mass of consensus on the solution.

The domain name issue is of critical importance -- both for the continued well-functioning of the Net as well as for promoting the decentralized approach to Net governance that has fostered the growth of electronic commerce and democratic participation in the information infrastructure.

There is an overriding interest in preserving the stability of the domain name system, and in avoiding fragmentation of the Internet.

Self-governance of the Internet is the right direction in order to promote these goals. No one government can or should exercise complete control over the establishment of the DNS.

Competition is welcomed by all as a useful tool to maximize consumer welfare and create efficient, self-regulating systems.

Areas of Divergence/Questions to be Resolved:

Is there a domain name crisis? To what extent are the tensions over name congestion, trademark law, or NSI control critical at this time? How long can the current system stay in place and remain stable?

What is the appropriate proposal for change in the system?

Arguments for the IAHC proposal contend that the current system is a fundamentally flawed monopoly, and provides too narrow a name space. At the same time, fully open privately owned gTLDs are risky for consumers as well and could create technical instabilities. There is, however, some optimism that a carefully constructed self-regulatory framework supported by robust technology would yield a stable and flexible DNS.

Arguments for the current NSI system are based on the need for a government role in ensuring the stability of the network such as the role the Federal Communications Commission has played in establishing the North American Numbering Council. Broadly sharing large numbers of new gTLDs may risk the stability of the network.

Supporters of the AlterNIC approach argue that a freely open domain name space is the most market-driven, least monopolistic, and most flexible -- and that technical challenges in maintaining network stability can be met.

"How do we decide how we decide" the future of the domain name system? The global, shared nature of the Internet -- as well as its track-record for self-regulation -- present new and open questions concerning the structure and legitimacy of institutions and processes for governing the domain name system.

Session 2: Trademarks and Internet Domain Names

Moderator: Mark Hellmann, American Intellectual Property Law Association

Panelists: Kevin Connolly, Policy Advisory Board

Michele Farber, AT&T

John Wood, Prince Plc.

Tim Casey, MCI

Doug Wood, Coalition for Advertising Supported Information

and Entertainment (CASIE)

Respondents: Lynne Beresford, U.S. Patent and Trademark Office

Susan Anthony, International Trademark Association

John Kamp, American Association of Advertising Agencies

Bill Burrington, America Online

Albert Tramposch, World Intellectual Property Organization

Trademark issues are a central problem in the debate over domain names. Trademarks identify a vendor or organization; they are names or other marks by which people can recognize goods and services. Consumers rely on trademarks to identify sources of information or products. Vendors rely on trademarks to differentiate their goods and services and promote good will for their products. Trademark law is designed to protect and promote this valuable investment in identifying names or other marks. Trademarks are especially important on the Internet, where traditional business relationships may not otherwise exist, geographical separation may be great, and reputation thus takes on greater import. At the same time, the traditional balance of trademark protection must be preserved to avoid chilling the competitive environment.

Trademark law varies widely internationally. American law is based largely on the use of a mark. Other countries, and recently the U.S. as well, based trademark on registration. The World Intellectual Property Organization (WIPO) has been involved in creating international standards for trademarks, but many of these issues remain unresolved even outside of the Internet and domain name context.

Emerging tensions in the current domain system include:

Limited space of useful names -- Leading to conflicts between companies or individuals seeking the same domain name within a given TLD.

Trademark inherently geographically based, while the Internet is global -- Two companies might easily do business under the same trademark name in New York, California, and Paris. On the Internet, however, they might all seek the same .com address.

Trademark based on avoiding consumer confusion, yet potential for consumer confusion is much greater on Internet where the entire world has access and consumers have little context for identifying the vendor or product. (For example, does united.com mean United Airlines, Van Lines, Video Sales -- or some other United?)

Companies have an interest in protecting trademarks throughout the TLDs: For example, AT&T might want to register att.com, att.firm, att.net, att.inc, and any other potentially confusing att.* addresses.

Need for clear rules as domain names are marketed -- Current lack of clarity in rules makes it difficult for companies to invest in marketing their web sites; it is "a hostile environment for business."

These tensions have created at least two major classes of problems: 1. "Bad actors" who seek to preemptively register a domain name in the expectation of reselling it at a high price to a more legitimate user. 2. Mark owners with legitimate interests in the same mark (for example, Prince Sports Equipment in the US and Prince plc in the UK, both seeking the prince.com domain name.)

Proposals for the reform of trademark treatment under the domain name system include:

Addition of new TLDs in order to create a larger namespace -- For example, under the IAHC proposal different companies could register under xxx.com, xxx.firm, xxx.net, etc.

Creation of dispute resolution procedures -- NSI and the IAHC proposal offer arbitration and other legal and extra-legal routes for handling conflicts over domain names.

Creation of screening process -- including waiting periods and notification process to allow for challenges to perspective registrations.

Isolation of bad actors seeking to ransom domain names -- Some have suggested using the expense of registration, or possible statutory measures, to discourage the registration of names solely for later resale.

Random domain names -- Some have suggested random assignment of names or numbers, like phone numbers.

The gTLD MOU proposal combines a slightly larger namespace (seven new gTLDs) with a new dispute resolution system to be administered through WIPO.

Major Areas of Consensus:

Need for stable domain name registration and trademark rules in order for businesses and others to invest in and maintain the value of their domain names.

Recognized growing importance of domain name identities to both business and consumers.

Need for a process, or at least greater collaboration within existing processes, to create greater consensus.

Need for a global, not-U.S.-centric, policy that is not reliant on the legal framework in any one country.

Areas of Disagreement/Issues to be Resolved:

Expansion of domain space -- Creates problems as well as benefits. While providing more names, it requires trademark owners to police more domains. It may also increase consumer confusion; for example, how do consumers differentiate united.com, united.firm, united.inc, etc.

Appropriate structure of the dispute resolution process -- How to best combine non-binding and binding mediation and arbitration, along with potential recourse to litigation. Choices of jurisdiction and arbitrator also remain at issue.

Appropriate notification and challenge waiting periods -- Proposals for 30- to 90- day dispute resolution systems may clash with electronic commerce imperatives for speed.

Session 3: Role of Governments and Multilateral Organizations

Moderator: Henry Perritt, Center for Information Law and Policy,

Chicago-Kent School of Law

Panelists: Robert Shaw, International Telecommunication Union

Albert Tramposch, World Intellectual Property Organization

Roger Cochetti, IBM

Respondents: Richard Beaird, U.S. Department of State

William Black, Nominet U.K.

J. Beckwith Burr, National Telecommunications

and Information Administration

Harold Feld, Domain Name Rights Coalition

The appropriate role of governments in Internet regulation as a whole has been hotly debated, and the domain names system is one of the focal points of that debate. While many favor self-regulatory, self-governing, and market-based models for Internet governance, there is also a sense that some role for government may be necessary even within those models. Governments have an increasing interest in maintaining the stability of major infrastructures, including the DNS; to the extent that the DNS is structured on private contractual relationships, government may need to play a role as the enforcer of those contracts. At the same time, the inherently global nature of the Internet calls into question the old models of government involvement. In addition, government involvement comes with a degree of caution; in a highly decentralized network, the DNS represents one of the few points of central control remaining to allow nations or multilateral organizations to exercise their control over the Internet -- for better and for worse.

Traditionally, only sovereign governments can make, apply, and enforce laws. This has traditionally meant nation-states with defined borders, a defined population, and the physical power to exercise coercive compliance with the law. In the context of Internet domain names, many of these definitions break down. There is no legitimate sovereign with the coercive authority to enforce rules on the Internet. This tension has led many to explore the possibilities of private self-governance as a model for domain name and broader Internet rule-making.

The private self-governance model relies on a contract-based system for domain name governance, based on market agreements between users, service providers, domain name registries and domain name registers. The global nature of the Internet makes the alternative a potential patchwork of conflicting national law both less desirable and less effective. Moreover, the interactive nature of the Internet creates new possibilities for effective contract-based choices. Yet such private self-governance systems face certain challenges, such as:

Who defends and enforces contracts?

How are disputes with different decision-making bodies handled?

How is compliance ensured?

How is basic fairness ensured among large and small entities with unequal bargaining power?

There may be additional reasons to expect an active role for governments. Government tends be good at addressing potential market failures, such as development of public goods and infrastructures, protecting due process, or promoting the interests of consumers and individuals. Government also may retain an added stamp of legitimacy based in their own local sovereignty.

The debate about government roles has focused on a continuum of models for domain name control, including:

"Private self-governance" based largely on contracts and market-based systems for allocation and registration of domain names with minimal governmental involvement.

"Voluntary multilateralism" -- Typified by the IAHC MoU, with features such as its WIPO-administered alternative dispute resolution system, and an enabling MoU signed by companies and NGOs.

"International legal framework" -- Independent from national governments, but with due process, perhaps using existing organizations (such as WIPO or the World Trade Organization), and perhaps based on existing frameworks (such as the FCC North American Numbering Council).

A more visible role for the U.S. government, beginning with its activities regarding the Department of Commerce Request for Comment.

Within this spectrum of governance models, the overriding question facing the Internet community remains one of process and legitimacy: "How do we decide how we decide?"

Major Areas of Consensus:

The global nature of the domain system dictates that any successful approach will be outside the control of any single government. This has both good and bad effects.

There is a shared desire to use private self-governance systems to the extent possible. (Though there is real divergence on what exactly that extent is.) To that end, there is a need to act quickly so that increasingly concerned national governments do not find ways to take over the process. The private sector needs to get on with working out a private contractual framework.

Many agree that the Department of Commerce Request for Comment process has the potential to create a process for further achieving consensus behind a private contractual framework.

Areas of Disagreement/Questions to be Resolved:

To what extent is traditional government needed to support self-governance systems for the domain name system?

To what extent should national governments retain an interest in the domain name system based on their traditional interests in the safety and well-being of their citizenry, the well-functioning of the economic system, and their consequent interest in the stability of the domain name system and the appropriate redress of claims by their citizens?

What is the appropriate role of "voluntary multilateralism"? Will it sufficiently protect all of the important national and international interests, such as those of consumers and the public?

Level of consensus behind MOU process -- There remains some controversy about the openness of the process.

Appropriate role of U.S. government -- Does the U.S. government retain any special interests in gTLDs, and the Internet in general, based on its role in the creation of the ARPAnet and its continued maintenance of the gTLDS?

How can these governance structures -- particularly non-governmental structures -- be designed to be responsive to the needs of individuals or consumers, who tend to be less powerful market players?

Session 4: Administrative and Operational Issues -- Getting From Here to There

Moderator: John Curren, BBN Planet

Panelists: Dave Crocker, Internet Mail Consortium

William L. Schrader, PSINet

Respondents: Ted Wolf, Jr., Dun & Bradstreet

Alan Sullivan, Top Domain Registry

Jay Adelson, DIGITAL Internet Exchange

Administrative, operational, and other technical issues have a real impact on the debate on domain name policy. As many have noted, the domain name system is a running service, used by millions every day. It is also a sensitive system, where a small mistake can cause a lot of damage. There is a paramount interest in maintaining the technical stability of the system in transition.

There remains some ambiguity about how, when, and even if a transition to a new system would happen. Some have noted that it may be possible to run a "shared registry" model (such as the IAHC model) and the "old system" (the NSI model) at the same time on "parallel tracks." Such a possibility would depend on the policies according to which rootservers are operated. In any event, there remains great concern in some corners about the lead time needed for transition from, for example, NSI to the IAHC CORE -- and a great concern over ability to get technical specifications as soon as possible.

Some of these technical issues directly implicate the larger set of issues surrounding the IAHC MoU and other competing models for domain name governance. For example, the technical ability to separate domain registrar from "ownership" of a domain registry might make it possible to break up what many fear is a natural monopoly. While there is no technical limit on the number of registrars for a given registry, there may be a practical limit. More discussion is needed about the processes and technology required to best implement shared registries.

Major Areas of Consensus:

There is a widespread commitment from all parties to the stability and robustness of the domain name system and the Internet as a whole.

Care must be taken to assure that necessary processes are developed and in place to maintain the stability of the DNS in a changing environment.

Areas of Disagreement/Issues to be Resolved:

There still remains a good deal of concern over the IAHC MOU, and what sort of operational concerns are raised by its potential enactment in advance of the March 1998 expiration of NSI's contract with the NSF.

There may be a need for more outreach and discussion before enactment of MOU.

Further exploration of the potential for separation of registry and registrars my prove helpful in examining the larger questions concerning governance models for the domain name system.

Conclusion

It is clear from the Forum's sessions that there are significant areas of agreement among most of the interested parties involved in the debate over domain names. Among the points of consensus are:

The need for some process to resolve the ongoing issues regarding the future of the domain name system.

Desire by all parties to do what's best for the Net -- and to avoid fragmentation of the Internet.

Self-governance and self-regulation are widely regarded as desirable models for the Internet.

Continued commitment to maintain the stability of the domain name system, even in a time of transition.

At the same time, there remain significant open questions about the appropriate model for the domain system moving forward from the March 1998 end of NSI's contract, the process used to decide upon that model, and the ultimate role of government in the domain name system. There is a particular need to address these issues if the Internet is to maintain its tradition of self-governance; otherwise, national governments and multi-lateral organizations are almost certain to step in, perhaps in ways that many fear would be damaging to the unique potential of the decentralized Internet. To that end, it is essential that we keep the Internet community talking.

Report prepared by Daniel J. Weitzner <djw@cdt.org>

Alan B. Davidson <abd@cdt.org>

For more information contact: Sheila O'Neill, <soneill@itaa.org>

Brian O'Shaughnessy <briano@isa.net>

Daniel J. Weitzner <djw@cdt.org>

Internet Domain Name Forum

Presenters and Respondents

Jay Adelson

DIGITAL Internet Exchange

Susan Anthony

International Trademark Association

Gabriel A. Battista

Network Solutions

Richard Beaird

U.S. Department of State

Lynn Beresford

U.S.Patent and Trademark Office

Jerry Berman

Center for Democracy and Technology

J. Beckwith Burr

National Telecommunications and Information Administration

William Black

Nominet U.K.

Bill Burrington

America Online

Tim Casey

MCI

Roger Cochetti

IBM

Kevin Conolly

Policy Adisory Board

Dave Crocker

Internet Mail Consortium

Michele Farber

AT&T

Jay Fenello

Iperdome

Ken Fockler

CA Net

Don Heath

Internet Society

Mark Hellmann

American Intellectual Property Association

Brian Kahin

U.S. Office of Science and Technology

John Kamp

American Association of Advertising Agencies

Harris N. Miller

ITAA

Jeff B. Richards

ISA

William L. Schrader

PSINet

Robert Shaw

International Telecommunication Union

Alan Sullivan

Top Domain Registry

Albert Tramposch

World Intellectual Property Organization

Ted Wolf

Dun & Bradstreet

John Wood

Prince, Plc.

###

Number: 362

From: Antony Van Couvering <avc@netnamesusa.com>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 4:51pm

Subject: NetNames Response to Request for Comments

NetNames is pleased to present a statement in regard to the establishment,

administration, and maintenance of Internet domain names in response to the

Department of Commerce's Request for Comments.

The response is attached in Microsoft Word format. The text can also be

found at http://www.netnamesusa.com/gtld-policy.html.

NetNames has also signed a response submitted by some members of the Policy

Advisory Body ("PAB") established in connection with the Memorandum of

Understanding Respecting Generic Top-Level Domains. We submit this

NetNames Policy Statement in addition because it presents a different focus

on the issues.

With thanks,

Antony Van Couvering

President, NetNames USA

Version 1.4

August 15, 1997

http://www.netnamesusa.com/gtld-policy.html

Note: since this statement was written, eDNS has been

disbanded, and has split up into different groups.

NetNames Policy Statement

on New Generic Top-Level Internet Domains

Contents

Introduction and Summary

Who is NetNames?

Background

Three Competing Models for Creating New gTLDs

Network Solutions Plan

The eDNS Plan

The gTLD-MoU Plan

Conclusion

Sources

Introduction and Summary

Domain names and related Internet addressing schemes lie at the very heart of the Internet, both commercially and technically. Insuring the smooth and stable growth of these critical resource is in the interests of everyone who uses the Internet.

NetNames is a commercial, global domain name registry whose first-hand experience in registering domains worldwide for over two years has given us a perspective which we would now like to make public.

Having witnessed and monitored first the clamor for, and then the development of, new generic Top-Level Domains (gTLDs), NetNames endorses the process begun by the Internet Ad Hoc Committee (IAHC) and continued with the signature of the generic Top-Level Domain Memorandum of Understanding (gTLD-MoU) over two other competing plans sponsored by Network Solutions, Inc., and the confederation known as eDNS.

Three models for introducing and managing new (gTLDs) are now being debated:

1. The gTLD-MoU proposal signed in Geneva May 1, 1997 by NetNames and other companies. This is a controlled-growth shared registry model.

2. The eDNS model now being implemented in scattershot fashion by members of the eDNS. This is a multi-gTLD monopoly model with different owners of different domains.

3. The proposal from Network Solutions published in rebuttal to the GTLD-MOU proposal. This is a multi-gTLD monopoly model, or a hybrid monopoly/shared model, with Network Solutions in either case owning .com, .net, and .org, the domains they currently control.

NetNames considers the following to be the paramount factors in our decision to support the GTLD-MOU proposal and to sign the generic Top-Level Domain Memorandum of Understanding (gTLD-MoU) promulgated by the IAHC:

A shared registry model is the the best way to assure the competitiveness in the domain name registry business, and for guaranteeing the stability and security of customers' domain name registrations. The GTLD-MOU model assures this, the other two plans do not. Domain name portability is the key benefit for customers in the shared registry model. The customer must be able to feel secure that the resources invested in advertising, marketing, and otherwise promoting his/her domain name will not be lost. Also, the consumer should be protected from unconscionable fee hikes.

A shared registry model permits this by allowing a customer to move to a different registrar without the change being visible to customers, as is now the case with 800 telephone numbers. The GTLD-MOU model allows for this; the customer can easily lose his/her entire investment with the other two models.

An open and participatory process for developing gTLDs is crucial for the smooth and rapid growth of the Internet. There are many interests with stakes in the development of gTLDs: individuals, governments, legal and trademark interests, commercial interests, Internet service providers, public policy groups, and others. The process of making changes to gTLDs must be open to interested parties. The GTLD-MOU model creates a Policy Advisory Body (PAB) open to any interest; there is no parallel body in either of the other two plans.

Some measures toward the protection of intellectual property is imperative. The worldwide domain space is being flooded by duplicate registrations from companies attempting to pre-emptively protect their branded equities and company names. The GTLD-MOU model goes some way toward this by the creation of Administrative Challenge Panels to be administered by the World Intellectual Property Organization (WIPO). The eDNS group vaguely suggests that "market forces" will sort out the problems; Network Solutions' proposal doesn't even mention it -- under their plan, users will continue to be at the mercy of the secretive and unaccountable mechanisms now in place.

Who is NetNames?

NetNames is an international Internet Domain Name Registry, a company that works exclusively with domain names, based in New York and London. We register domain names for clients in every top-level domain in the world, including all of the country-based Top-Level Domains (TLDs), e.g., .de for Germany, .my for Malaysia, and so on.

NetNames has extensive experience in this area. Not only do we deal with more TLDs on a day-to-day basis, with experience of every kind of cultural and technical difference imaginable, but Ivan Pope, the Managing Director of NetNames, Ltd., is a director of Nominet, which administers the TLD the for United Kingdom, and has been since its inception.

NetNames has the unique ability to view domain name issues from three perspectives: from the point of view of a TLD, from the point of view of the customer, and from the point of view of a commercial registrar of domain names. Our clients include 3M Corporation, Citibank, Goodyear Tire & Rubber, Microsoft Corporation, Netcom, Netscape Communications, and many other large multinational corporations. We have also registered thousands of domains for smaller companies.

What is NetNames' vested interest in this debate? From a financial point of view, we are neutral. We are a signatory to the gTLD-MoU, and NetNames will be filing an application to become a Registrar under the gTLD-MoU plan, so we have an interest in seeing this plan go forward. We are also a Preferred Partner of Network Solutions and receive special treatment from them, including a guaranteed turnaround of one day for our domain registrations, and publicity from Network Solutions. We have also been the owner of the .EUR domain, an eDNS domains, and in the short term we would almost certainly benefit financially from the fracture and fracas their plan would create. We have been keeping abreast of the developments in all three camps. Recently, Antony Van Couvering, President of NetNames USA, has been elected Chair of the Policy Advisory Board.

We feel that now is the time to make our voice heard concerning the disputes now raging about the gTLD-MoU signed May 1, 1997 in Geneva, and the competing plans put forward by Network Solutions, the current operators of the InterNIC in Virginia, and the group known as eDNS.

Background

In 1995, in part because of a perceived lack of available domain names, but also due to a chorus of complaints from end-users and registrars about the operating procedures of Network Solutions, which holds a contract to run the InterNIC (the registrar for the common .com, .net, and .org domains), pressure was put on the Internet Assigned Numbers Authority (IANA) to create new gTLDs. After intense discussion on relevant mailing lists, the IAHC was formed from a variety of respected Internet bodies with a mandate to produce a proposal for the introduction and governance of new gTLDs.

In response to community concern, the IAHC had as its main task solving several interrelated problems that had begun to plague the Internet namespace:

1. Domain hoarding, pirating, and "cybersquatting"

2. Trademarks and their relationship to domain names

3. Dispute resolution procedures for "name collisions," i.e. two or more parties desiring the same domain name

4. Secrecy, non-accountability, and lack of community involvement in the domain registration and dispute resolution policies developed by Network Solutions, the monopoly registrant of gTLDs.

5. Cost of domain name registration

6. The shortage of "good" domain names

For many months the IAHC floated proposals and solicited ideas, both on Internet mailing lists and from private submissions.

After some months of discussion it became apparent that a consensus existed that:

1. New top-level domains should be created as a way to alleviate the crowding of the domain space and to reduce the viability of "cybersquatters."

2. An open, accountable, long-term process for resolving future problems should be created

3. A "shared registry" model, as opposed to a "monopoly registry" model, should be created so that Internet users would no longer be beholden to the unilateral policies of a single, unaccountable entity. The key benefit to the domain name holder is "domain portability", or the ability to find another provider, without losing the investment in the name itself. A comparable model is found in the 800 telephone number market, where a customer can change long-distance companies without having to get a new phone number.

4. The shared registry model should be commercial, and that registrars should compete on price and services

5. Some attempt should be made to deal with the problem of trademarks and domain names

6. The solution should be global, not "U.S.-centric"

On February 4, 1997, the IAHC issued a press release summarizing its final proposal, and shortly thereafter released the text of the gTLD-MoU. Since that time, persons and companies have voiced objections to the gTLD-MoU, often upon the demonstrably false ground that the IAHC was not an open process. Other objections have been made on the grounds that the IAHC was not legally constituted, a red herring, since Internet working groups in general are not legally constituted. Very few substantive objections have been made.

The sticking point appears to be the shared-registry model, which had widespread support when first proposed on the mailing lists. The shared-registry model directly threatens two groups that have hopes of windfall profits based on the monopoly ownership of gTLDs:

eDNS, a loose confederation of small companies with claims to certain gTLDs of their own invention (e.g., .earth, .xxx, .mall), had its roots in opposition to the monopoly held by Network Solutions on the .com, .net, and .org domains. Since the gTLD-MoU, however, eDNS has found a common position with Network Solutions due to their common desire for the creation of monopoly gTLDs.

Network Solutions initially supported the IAHC process, but just before the signing of the gTLD-MoU in Geneva, issued a statement critical of the gTLD-MoU, asserting, in essence, that Network Solutions could do the job better. We would like to compare the three models -- Network Solutions, eDNS, and the gTLD-MoU plan -- based on their merits and flaws.

Three Competing Models for Creating New gTLDs

NetNames believes that competition within a stable framework is essential for the continued development of the commercial Internet. With that as our criterion, we went to Geneva skeptical about the gTLD-MoU, with several significant reservations about it. We actively participated in the discussions in Geneva with three days of very frank questions, and we were satisfied by the answers, and--almost more importantly--by the open nature of the process. We therefore decided to sign the gTLD-MoU, and we would like to share with you our decision process.

We would like to discuss first the plan put forward by Network Solutions, then the eDNS plan, and finally the IAHC plan.

Network Solutions Plan

NetNames strongly rejects the rebuttal paper ("Secure Internet Administration and Competition in Domain Naming Services", April 17, 1997) issued by Network Solutions in response to the gTLD-MoU proposal. We have read their plan closely and have found it to be self-serving and misleading. It is important to remember that the formation and work of the IAHC came in response to the public outrage at the poor customer service, the closed decision-making process, and the high-handed and ill-considered dispute resolution mechanisms of Network Solutions. Network Solutions has made improvements in the customer service area recently, but it continues to be a closed and protective monopoly.

We are left with the unmistakable impression that Network Solutions is attempting to wrest control of "root" (or "dot" or ".") from the IANA by raising the specter of a "crisis of credibility" which didn't, until the issuance of the Network Solutions paper, exist. IANA, an informal body headed by Jon Postel, has successfully run "root" for many years and its leadership in regard to domain name issues has helped insure the smooth growth of the Internet. As noted above, IANA was forced to initiate the process which resulted in the IAHC being formed in response to a real crisis of credibility at Network Solutions. We believe that IANA should probably acquire a legal status to protect it from any future threat of usurpation.

The Network Solutions proposal contains many untrue assertions and semi-truths that to our mind undermine their credibility in this arena. In particular, their objections to the gTLD-MoU plan seem flimsy and self-serving.

1. Network Solutions says that the gTLD-MoU proposal does not recognize the need for market branding and that legitimate corporations will not invest in "shared brands." This is a misstatement of the problem as well as a misinterpretation of the shared registry model. In fact the gTLD-MoU proposal is much stronger than other plans with regard to "branding" because not only does a company have control of its brand, it has security as to its continued existence as well, since under the shared-registry model the company can change providers without losing its name. NetNames consults with major corporations on this issue every day and we can say without hesitation that this assertion has no basis in fact.

For instance, if a company registers "company.web" with one registrar in the shared registry model, it does not share the brand of "company", in fact it has exclusive control over that name. BUT it has a choice of who it wants to pay to register and maintain that name. NetNames has taken advance orders for the new domains from many large American corporations and over 1000 others.

2. Network Solutions says that the gTLD-MoU plan is too complex to succeed. This has already been proven untrue in the event. Nominet, the UK registry, has auniversally acclaimed shared-registry model which was created out of whole cloth, immediately attracted over 100 registrars, now has over 350, and is completely scalable. It is simple in operation and, working on a cost-recovery model, has already dropped its prices by 20%, with another 50% (estimated) drop scheduled for later in 1997. The assertion that the international aspect is problematic is unsubstantiated and almost certainly untrue. NetNames deals with countries all over the world every day and sees all the country TLD templates, which are with few exceptions standardized and closely follow either the RIPE or InterNIC template forms.

3. Network Solutions says that the gTLD-MoU approach is too bureaucratic. We find this assertion hardly credible coming from Network Solutions, whose bureaucratic high-handedness prompted the whole re-examination of gTLDs in the first place. It is also untrue. The gTLD-MoU proposal closely parallels the Nominet model, which has a permanent staff of less than 10 and deals easily with a exponential growth, and (let us stress again) has nearly universal customer satisfaction. Contrary to what Network Solutions says, the shared-registry model encourages competition -- competition based on service, not on ownership.

4. Network Solutions says that the gTLD-MoU proposal is too narrow. The mandate of the IAHC was narrowly tailored, correctly so. There are a number of other initiatives (e.g., ARIN) to deal with IP address allocation. IAHC involvement in these issues would have been confusing and counterproductive. We feel that this is a non-issue and a red herring.

5. Network Solutions says that the name dispute resolution framework erected by the gTLD-MoU is unworkable. Indeed, there are problems with the gTLD-MoU proposal. We would submit, however, that many of these problems are close to insoluble -- trademark law, in over 100 years, has been unable to come up with an international solution. We would note, however, that Network Solutions' assertion that there is any technical problem involved in time-stamping registrations from around the globe is spurious. The technology exists and is being used every day with great efficiency. We would also note that Network Solutions' own dispute resolution policy, which even today is widely resented, was a major impetus for the creation of the IAHC.

6. Finally, Network Solutions claims that IAHC has no legal authority. Since IAHC was constituted under the authority of IANA, Network Solutions is really saying that IANA has no legal authority. Here, we feel, is the motivation for the Network Solutions proposal -- an attempt to wrest control of "root" from IANA. Network Solutions, whose own legal status in regard to the domains it controls is hardly clear, has written a proposal to muddy the waters and confuse people who don't understand the Internet in an attempt to jettison IANA and gain control of "root". It is true that IANA is an informal arrangement; it was formed in a time when the Internet was not worth as much money as it is today, and working methods were far more informal and cooperative. IANA is nonetheless a well-respected group in the Internet community and its directives are applied without question by domain administrators and other standards bodies worldwide.

NetNames feels that Network Solutions objections to the gTLD-MoU plan are flimsy and unwarranted, and serve primarily to enhance Network Solutions' own prospects for profits and control. To have the key to the entire Internet controlled by a private company would be a disaster, and we strongly oppose any attempt by Network Solutions to gain control of "root".

The eDNS Plan

The eDNS proponents are a loose assemblage of companies and individuals who propose to allow anyone to create their own TLD and run it as they wish. NetNames feels that this approach is dangerous and destabilizing on many counts:

1. The eDNS model would undermine the world Domain Name Service (DNS), which was designed to work on a hierarchical model with a limited number of top-level domains. While a large number of TLDs may ultimately be feasible, and should be at least considered as a policy goal, we don't see any point in possibly breaking the system on a whim.

2. eDNS domains are not portable. A customer who invested in promoting his eDNS domain would be stuck with one service provider, namely the eDNS operator. If this operator went out of business, raised its rates through the roof, or simply cheated the customer, the customer's investment would be lost.

3. eDNS domains are competitive only on the vanity of the name, not on service or goods. So a customer might desire to register his or her company under the gTLD ".biz", but find that the service was so abysmal that it was not a good business decision. On the other hand, service might be very good at the TLD ".scum", but because the name is so appalling the company would decide that it was undesirable to register there.

4. eDNS has no domain name dispute resolution policy. From this decision a horrific scenario arises: hundreds of new TLDs, and hundreds of companies trying to register in all of them as the only way to protect their names. NetNames already has clients who have contracted us to register all their important trademarks in every possible domain. The eDNS proposal would exacerbate this problem exponentially.

The gTLD-MoU Plan

While we have some reservations with this model, NetNames believes that it offers the best plan now and, far more importantly, the best process for improving the domain name space in the future.

First, we would like to highlight our PROBLEMS with the gTLD-MoU plan:

1. The administration as outlined in the gTLD-MoU and related documents allowed far too little input from outside bodies. The Policy Oversight Committee (POC) is stacked with members from the so-called "I" organizations -- IANA, ISOC, etc. We feel that, as presently constituted, there is insufficient representation on POC from either the Policy Advisory Board (PAB), which is open to all, or from the Council of Registrars (CORE), or from some other important interests.

BUT NetNames received assurances in public that PAB would have representation on the POC. This reassured us considerably: first, that PAB would have representation; second, that the IAHC was committed to an open process.

2. We felt that using a lottery process for choosing registrars was abominable. The IAHC members expressed their distaste for it also, but said that they had no workable alternative. NetNames felt that the lottery process kept interested parties out and had the potential for creating a many-headed monopoly with a vested interested in keeping additional registrars out.

BUT as of this writing the interim Policy Oversight Committee (iPOC), who are the successors to the IAHC, has eliminated the lottery and made registrar status open to all qualified applicants. We take this as another indication that the process works and outside voices are listened to.

3. We were very impatient with the idea that there were technical problems with having a large number of initial registrars. We hotly contested the validity of this idea.

BUT we were told that actually the reason for keeping the number of registrars to a low number was a logistical consideration. We rejected this as an important concern, and since the lottery has been discarded there is now no limit on the number of registrars.

We would now like to share our reasons for SUPPORTING the gTLD-MoU Plan:

1. Open Process -- Our interactions with the IAHC and now the iPOC have convinced us that they are committed to dealing with the issues at hand in a fair and open manner. We were initially very skeptical on this count, but found that our fears were unfounded. The vociferous complaints about how the IAHC ran a closed process we now ascribe either to personality conflicts or to professional posturing.

2. Competition Based on Service, Not Vanity -- The gTLD-MoU plan encourages competition among registrars based on the quality and value-added features of their service, not simply on the value of the gTLD in which they are registering. eDNS relies on the vanity appeal of a given gTLD, and Network Solutions has relied on having the only gTLDs available.

3. Domain Name Portability -- The singlemost important aspect of a domain name, from a customer's point of view, is that he or she can continue to use it and promote it publicly without fear of losing it due the incompetence or unethical practices of their registrar. In other words, investment in the brand equity is secure. This is possible with the gTLD-MoU plan, because a registrant can have the name of their choice as well as the registrar of their choice. This can hardly be stressed enough.

4. Shared Registry Model -- The shared registry model, a feature of the gTLD-MoU plan but not of the other two, allows for domain name portability and competition based on services. NetNames knows, from its close experience with Nominet in the UK, that a shared registry model works, and works well.

5. Domain Name Dispute Resolution Procedures -- For many of NetNames' larger clients, this is the most exciting part of the proposal. Although it is inevitable that the process would be flawed -- you cannot map a fluid and flexible name categorization such as trademarks onto a hierarchical unique namespace such as domain names -- we believe that the establishment of the Administrative Challenge Panels will bring significant relief for holders of trademarks without unnecessarily restricting registration by others. Most important in this regard is the "exclusion" process whereby holders of famous marks can petition to have their name(s) protected from registration by others.

6. Open to Non-industry Participation -- NetNames believes that a wide range of companies and persons have valuable perspectives on the Internet namespace. These include registrars, of course, but also ISPs, lawyers, trademark holders, governments, and public policy bodies. The gTLD-MoU, through its establishment of a Policy Advisory Board (PAB), has implemented a procedure whereby these disparate voices can be heard. The iPOC (formerly the IAHC) has already implemented the overwhelming consensus from the PAB that the lottery system of registrars should be abolished.

7. Wide Support Within and Without the Industry -- NetNames has had the unique opportunity to listen not only to the loud voices on various mailing lists, but also to its own customers. We have answered hundreds of questions, and fielded many opinions, about the establishment of new gTLDs. Most of these were from non-informed persons who would nonetheless bear the brunt of any changes. We can attest that the gTLD-MoU plan is overwhelmingly supported by our customers.

8. Wide support From the Country TLDs -- NetNames works every day with the country TLDs, and again, the reaction from these groups (who have good knowledge of many of the problems) is in favor of the gTLD-MoU plan. Domain authorities from the NICs of Argentina, China, Guernsey, Japan, Jersey, Korea (South), Norway and Switzerland have either already signed the gTLD-MoU or have announced their intention to sign. This underlines the international character of the gTLD-MoU plan.

Conclusion

NetNames recommends that any person or organization interested in seeing the Internet develop with stability and competition join the plan proposed by the IAHC by signing the gTLD-MoU, which entitles the signer to a place on the Policy Advisory Body and a chance to make your voice heard on this very important step in the development of the Internet.

We would be pleased to discuss our positions or ideas with any interested person.

Ivan Pope, Managing Director of NetNames Ltd., can be reached in London on +44 171 224-2017, or via email at ivan@netnames.com

Antony Van Couvering, President of NetNames USA, can be reached in New York at +1 212 627-4599, or via email at avc@netnamesusa.com

Sources

For the gTLD-MoU Plan and accompanying documents: http://www.gtld-mou.org. Anyone wishing to sign the gTLD-MoU can find the form and instructions at http://www.itu.int/net-itu/dnsmeet/mousign.htm

For the Network Solutions Plan: http://www.netsol.com/papers/internet.html

For the eDNS Proposals: http://www.eDNS.net

NetNames: http://www.netnames.com

###

Number: 363

From: "Richard J. Sexton" <richard@vrx.net>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 4:54pm

Subject: REGISTRATION AND ADMINISTRATION OF INTERNET DOMAIN NAMES

Before the

U.S. DEPARTMENT OF COMMERCE

National Telecommunications and Information Administration

Washington, DC 20230

In the Matter of )

)

REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01

INTERNET DOMAIN NAMES )

Comments of Richard J. Sexton

Richard J. Sexton

Maitland House

Bannockburn

Ontario, CANADA

Self

18 Aug 1997

------------------------------ break ------------------------------

TABLE OF CONTENTS

Summary

A. Appropriate Principles

Principles a-f

Other principles

B. General/Organizational Framework Issues

Questions 1-9

C. Creation of New gTLDs

Questions 10-14

D. Policies for Registries

Questions 15-20

E. Trademark Issues

Questions 21-28

F. Other Issues

Existing TLD Applications

------------------------------ break ------------------------------

SUMMARY

This document provides some information on the structure of the

names in the Domain Name System (DNS), specifically the top-level

domain names; and on the administration of those domains.

The day-to-day responsibility for the assignment of IP Addresses,

Autonomous System Numbers, and most top and second level Domain

Names are handled by Internet Registries (IR).

The Internet is partly being driven by commercial market forces

using domain names to identify corporate business units which are

known to the public as "brand names"). This document covers some of

the framework necessary to define the purpose, function, delegation,

and use of new top level domains.

Several factors need to be addressed such as why the TLD exists in

the first place, who accepts registrations for the TLD, and what

special purpose (if any) the TLD serves.

------------------------------ break ------------------------------

Before the

U.S. DEPARTMENT OF COMMERCE

National Telecommunications and Information Administration

Washington, DC 20230

In the Matter of )

)

REGISTRATION AND ADMINISTRATION OF ) Docket No. 970613137-7137-01

INTERNET DOMAIN NAMES )

Comments of Richard J. Sexton

1. Richard J. Sexton respectfully submits comments in this proceeding.

Richard Sexton is actively involved in top level domain name issues.

Richard

J. Sexton has been seeking a resolution to the perceived problem of

Network Solutions unilateral policy decisions since their inception,

favoring an internet community based solution. Richard Sexton

has written and spoken a great deal on this subject in Ottawa,

Washington DC, Wall Street, to educate users, industry and

government to the issues today concerning the Internet Domain

name system.

------------------------------ break ------------------------------

A. APPROPRIATE PRINCIPLES

a. Competition in and expansion of the domain name

registration system should be encouraged. Conflicting

domains, systems, and registries should not be permitted to

jeopardize the interoperation of the Internet, however. The

addressing scheme should not prevent any user from

connecting to any other site.

A.a.1. This is a generally accepted principle.

b. The private sector, with input from governments, should

develop stable, consensus-based self-governing mechanisms for

domain name registration and management that adequately

defines responsibilities and maintains accountability.

A.b.1. This is a generally accepted principle.

c. These self-governance mechanisms should recognize the

inherently global nature of the Internet and be able to evolve

as necessary over time.

A.c.1. This is a generally accepted principle.

d. The overall framework for accommodating competition

should be open, robust, efficient, and fair.

A.d.1. This is a generally accepted principle.

e. The overall policy framework as well as name allocation and

management mechanisms should promote prompt, fair, and

efficient resolution of conflicts, including conflicts over

proprietary rights.

A.e.1. This is a generally accepted principle.

f. A framework should be adopted as quickly as prudent

consideration of these issues permits.

A.f.1. No. Consensus is more important than speed.

------------------------------ break ------------------------------

B. GENERAL/ORGANIZATIONAL FRAMEWORK ISSUES

1. What are the advantages and disadvantages of current

domain name registration systems?

B.1.1 The Internet Domain Name System ("DNS") is a hierarchical,

decentralized database providing a one to one mapping of domain

named to numerical Internet Protocol ("IP") addresses. Every

entity that operates a "name server" is a part of this infrastructure.

This system was created a decade ago and it's evolution has

not been addressed until very recently, when the Internet

community has matured to the point where it is now beginning

to understand the issues central to the problems associated with

it.

B.1.2 Originally the nomenclature of the DNS was divided into

geopolitical entities (the ISO 3166 TLDs such as .us, .ca

and so forth) and international TLDs: .com, .net, .org, .mil,

.gov, .int, .ARPA. .uk was created as an anomaly; it was

in use and popular before the ISO3166 domains were established

and owing to this achieved a de facto legitimizaiton over the

strictly correct and largely unused .gb moniker for Great Britain.

B.1.3 Several events have caused problems.

B.1.3.1 The .COM "zone" is too flat, that is a disproportionate

number of computers are named "something.com". Whereas the DNS

was supposed to be a nomenclature to use arbitrary identifiers

to evenly spread the job of name to address lookups over a

large number of computers, the name www.something.com is now

seen as a cultural identifier, not technology. Businesses

and individuals alike have been fooled by this.

B.1.3.2 The geopolitical ISO 3166 TLD's have been underutilized.

The greatest reason for this is poor management, service and policy

from the ISO3166 registrars.

B.1.3.3 There is a collision between the "purists" to designed

and implemented the DNS nomenclature and commercial interests

to want to share in the recent revenue stream created by

instigation of fees for the most popular top level domain - .com.

There is little interest outside the most ardent of Internet

engineers to operate names such as ".nic, for network

information centers" that are not deemed to have wide

market appeal. The .com domain is most popular for several reasons.

B.1.3.3.1 The performance of the .com registry: "if the

name is not taken, you may use it". By keeping the amount

of domain registration application checking to its minimal

state, domains are delegated as quickly as possible, an

important concept as internet intellectual infrastructure

grows by the second, and any impediment to the information

based economy of the Internet hinders the development of

this global resource.

B.1.3.3.2 The rise in electronic commerce. The growth

of the DNS is quite simple. .com grew as entities believed

the nomenclature of RFC1591 that states .com is for commercial

entities. Thus, the nomenclature has been accepted universally

adopted.

B.1.3.3.3 The desire to participate in a global marketplace. There

is strong resistance by all entities to be identified by the

national country of origin.

B.1.3.3.4 The Internet is self organizing. The ISO geopolitical

nomenclature, as a top level organization of the Internet has

not flourished despite over a decade of advocacy. A

fundamental principle of Internet architecture is that names must

be self descriptive, and when naming computers, it is more important

to know that the computer does, than where it is. Internet technology

makes every other computer on the Internet seem to be directly

attached to any computer that a user chooses; geopolitical

boundaries are artifacts of the preinformaiton age, and serve

no great purpose to the Internet as a whole, other than to

drive policy, to an extent, for physical components of the

Internet infrastructure.

B.1.3.4 The fact that the Internet is self organizing is

both a "bug AND a feature". While this has allowed the best

and the brightest to develop meaningful taxa, others have

created some truly unfortunate names to the detriment of the

average user. Poor understanding and desire for commercial

visibility have driven this.

2. How might current domain name systems be improved?

B.2.1. By consensus. Anarchy will erupt otherwise.

3. By what entity, entities, or types of entities should current

domain name systems be administered? What should the

makeup of such an entity be?

B.3.1. That entity is, by definition, the Internet. Anybody

who wished to contribute and has, demonstrably, something

to contribute, should be able to. This is the fundamental

principle that has fostered the growth of the collaborative

collective that is the Internet.

4. Are there decision-making processes that can serve as

models for deciding on domain name registration systems (e.g.,

network numbering plan, standard-setting processes, spectrum

allocation)? Are there private/public sector administered

models or regimes that can be used for domain name

registration (e.g., network numbering plan, standard setting

processes, or spectrum allocation processes)? What is the

proper role of national or international governmental/non-

governmental organizations, if any, in national and

international domain name registration systems?

B.4.1. No. Not even close.

5. Should generic top level domains (gTLDs), (e.g., .com), be

retired from circulation? Should geographic or country codes

(e.g., .US) be required? If so, what should happen to the .com

registry? Are gTLD management issues separable from

questions about International Standards Organization (ISO)

country code domains?

B.5.1. Policy and "appearance" of a TLD drive it's acceptance. Each

TLD has a policy mandate that defines who can use it, and what it's

used for. For ISO 3166 TLDs this is usually a national sovereign

appointed agency, although the .us and .uk registries are delegated

dnd decentralized with varying measures of success. Migration away

from .com will occur if a cheaper or better alternative, or both,

arises. Some ISO 3166 TLDs have been delegated by national sovereigns

to commercial entities. The lesson here is clear: any entity that

can successfully operate a registry according to it's charter

mandate a policies, should be allowed to do so. If somebody has

a better way to run a registry, this contribution should not be ignored.

"Better" is determined by the Internet community as "rough consensus

and running code.

6. Are there any technological solutions to current domain

name registration issues? Are there any issues concerning the

relationship of registrars and gTLDs with root servers?

B.6.1. Yes, in theory, one single computer can completely automate the

domain registration process. Because there is no process by which

this could have been achieved owing to conflicting contractual and

sovereign interests, this has not happened. Privatization of the

domain registration process without contribution to the intellectual

infrastructure is seen as self serving and harmful.

B.6.2 The set of all name server owner operators determine what

content (names) those machines will carry. The registration of

a name is a very trivial data entry operation that the WWW makes

even easier. Front line support for people who lack an understanding

of the nature of the process, or who have run afoul of the system

due to an error made by the applicant or registry is the only issue

disparate from operational name service.

7. How can we ensure the scalability of the domain name

system name and address spaces as well as ensure that root

servers continue to interoperate and coordinate?

B.7.1. By consensus. Disenfranchised entities do not contribute

to, and harm, global stability and nomenclatural growth and innovation.

8. How should the transition to any new systems be

accomplished?

B.8.1. On a consensual basis, and very, very carefully. There is only

one chance, for all practical purposes, to do this.

9. Are there any other issues that should be addressed in this

area?

B.9.1. Yes, besides domain names, contact information (such as

"nic handles") and the whois database contain more information than

domain name and ip address. This had the potential to be a global

directory service, but proprietary interests, which are the

antithesis of the open standards based Internet diminish or

defeat this possibility.

------------------------------ break ------------------------------

C. CREATION OF NEW gTLDs

10. Are there technical, practical, and/or policy considerations

that constrain the total number of different gTLDs that can be

created?

C.10.1. Not in general. Some code changes are required on all name servers

somewhere below 10,000 TLDs. The total number of TLDs proposed is under

1000 in the most liberal case. There is no evidence that this number

needs be is a hard and fast rule, rather it reflects contemporary

operational practices.

11. Should additional gTLDs be created?

C.11.1. Currently, no "gTLDs" have been implemented. There are

several international TLDs. If the current problems in their

administration can be resolved and consensus on an improved

global nomenclature can be met, then implementation of new TLDs

would be appropriate.

12. Are there technical, business, and/or policy issues about

guaranteeing the scalability of the name space associated with

increasing the number of gTLDs?

C.12.1. No.

13. Are gTLD management issues separable from questions

about ISO country code domains?

C.13.1. No. Administration of a TLD according to it's method

and purpose bears no relationship to the name or meaning

implied by the name. This holds true for geopolitical names

and therefore ISO 3166 TLDs.

14. Are there any other issues that should be addressed in this

area?

C.14.1. No.

------------------------------ break ------------------------------

D. POLICIES FOR REGISTRIES

15. Should a gTLD registrar have exclusive control over a

particular gTLD? Are there any technical limitations on using

shared registries for some or all gTLDs? Can exclusive and

non-exclusive gTLDs coexist?

D.15.1. That depends on the charter for the TLD. "Shared registry"

is an artificial abstraction. "Sharing registry" is more accurate.

The database of domain names and ip addresses is monolithic. Just

what entities have read/write access to it defines the extent

to which the registry is "shared". Exclusive and non exclusive

TLDs already exist and present no known operational problems.

16. Should there be threshold requirements for domain name

registrars, and what responsibilities should such registrars

have? Who will determine these and how?

D.16.1 Yes. a registry needs to be able to fulfill the obligations

of the charter of the TLD it serves. Traditionally empirical

qualification of this process, not mandated bureaucracy has

been the way the Internet does this. Consumers of domain

names determine success of a model; their complaints are

aired on the Internet, most succinctly in the newsgroup

comp.protocols.tcp-ip.domains.

17. Are there technical limitations on the possible number of

domain name registrars?

D.17.1. No. Any entity that can follow a defined process can

currently automated, fees for service is seen arbitrary in the

case where the registrar performs an action that could be

done as well if not better, by a sufficiently knowledgable

domain applicant, who provides the front line support that

is, in effect a fee based education as to how the process works.

18. Are there technical, business and/or policy issues about the

name space raised by increasing the number of domain name

registrars?

D.18.1. The policy and technical issues are constant with 1 to

N instantiations of TLD's.

D.18.2 The root servers have always been subsidized by benevolent

entities and volunteers from the Internet community. If the registration

process were run in such a manner and the "price of admission" to

be able to register a domain were raised to the height of being

able to operate a robust and stable name service, there would

be no need to ask questions like this, and no need for "registrars"

per se. Every name server operator must by definition understand the

DNS to be able to program their name servers data records. There is

no special knowledge required to program the root zone name servers.

D.18.3 The business issue is only an issue of late, when the

cost of subsidizing the registration process outgrew the revenue

stream provided for it. The Internet community was never given

a decent chance to rectify this situation, nor encouraged to

do so. Ironic in a culture that had bootstrapped itself into

the most important communications network in the world on the spirit

of free cooperative collaboration. Prior to two years ago

there were no "business issues" and whether there should be

any or not is an unresolved question. There is no technical

reason why this needs to be a "business" at all.

19. Should there be a limit on the number of different gTLDs a

given registrar can administer? Does this depend on whether

the registrar has exclusive or non-exclusive rights to the

gTLD?

D.19.1. This is dictated by the charter of the TLD. If we agree

the USG should "own" .gov, it is appropriate they should be the

exclusive registrar.

D.19.2 A registrar could register into one or many TLD's depending

on TLD charter and technical competence.

20. Are there any other issues that should be addressed in this

area?

D.20.1. No.

------------------------------ break ------------------------------

E. TRADEMARK ISSUES

21. What trademark rights (e.g., registered trademarks,

common law trademarks, geographic indications, etc.), if any,

should be protected on the Internet vis-a-vis domain names?

E.21.1 Trademark law is a mature and venerable set of rules to

share names among different geopolitical entities for different

products and or services. Domain names do not fit well into

this model, they do not for the most part recognize geopolitical

delineations, nor do they apply to a consistent product or

service.

E.22.2 Any trademark, whether granted at the state, national or

international level, that prohibits that mark to be excluded

from the free pool of domain names is an unnatural expansion

of trademark law outside it's scope of providing protection

in a certain geopolitical arena and specific area of product

or service, and inhibits innovative growth in favor of

commercial protectionism.

E.22.3 The international trademark issues are nowhere near

resolved. Only the ISO country code domains provide reasonable

relief by constraining the problem to the laws of that sovereign

nation. Network Solutions handling of this problem, in an attempt

to not be the target of unfair litigation has been bad. The IAHC

plan is worse.

22. Should some process of preliminary review of an

application for registration of a domain name be required,

before allocation, to determine if it conflicts with a trademark,

a trade name, a geographic indication, etc.? If so, what

standards should be used? Who should conduct the

preliminary review? If a conflict is found, what should be

done, e.g., domain name applicant and/or trademark owner

notified of the conflict? Automatic referral to dispute

settlement?

E.22.1. No. The slight advantage this gives trademark holders

in prosecuting violators is damaging to the greater good of

the Internet community to be able to deploy sensibly named

information to the Internet. It wold be a shame to bog that

process down due to commercial interests that did not build

and do not understand the domain name system.

23. Aside from a preliminary review process, how should

trademark rights be protected on the Internet vis-a-vis domain

names? What entity(ies), if any, should resolve disputes? Are

national courts the only appropriate forum for such disputes?

Specifically, is there a role for national/international

governmental/nongovernmental organizations?

E.23.1 Trademark right do not automatically scale to domain

name rights! The International Business Machines corporation

has the domain name ibm.com. They registered it before anybody

else did. ibm.org is owned by and operated in a manner as not

to cause confusion in the eyes of the consumer. Certainly IBM

as a company could have easily, through legal machinations,

obtained or shut down ibm.org, and to it's great credit, it

did not. Instead it sends a clear signal that it understands

the Internet and trademark law, and where to intersect and

where they do not.

E.23.2 Merely registering a domain name the same as a trademarked

term (and it is to be noted that if all current terms were to

be scrupulously reviews, some trademarks should never have been

granted. To extend a flawed system into the Internet would be

wrong) does not imply a trademark violation, it merely means

the same arbitrary string of characters has more than one application

when examined in a global arena. At such time as that arbitrary

string is used in such a way as to formally and technically violate

a trademark, the two entities in dispute have legacy systems to

resolve that dispute. The must be held immune from these disputes

in the nominal case where no malfeasance was involved.

24. How can conflicts over trademarks best be prevented?

What information resources (e.g. databases of registered

domain names, registered trademarks, trade names) could help

reduce potential conflicts? If there should be a database(s),

who should create the database(s)? How should such a

database(s) be used?

E.24.1. Trademark violation occurs when a name is used in a way

that causes confusion in the mind of a consumers regarding a particular

product or service, and this is not clear until the domain name is

put to use, so there is no way of a priori determining trademark

violation in domain space.

25. Should domain name applicants be required to

demonstrate that they have a basis for requesting a particular

domain name? If so, what information should be supplied?

Who should evaluate the information? On the basis of what

criteria?

E.25.1 In the case of a private TLD - if we presume the Internet

community delegates .gov to the USG, it is clear the owner of

a TLD can set any requirement it likes. In the case of a public

TLDs, no requirement other than the technical validity of the

data being entered into the database (correctness of IP addresses

of the name servers serving that domain, live operational status

of these name servers) is warranted.

26. How would the number of different gTLDs and the number

of registrars affect the number and cost of resolving trademark

disputes?

E.26.1. The more entities involved the higher the cost and slower

the speed of resolution.

27. Where there are valid, but conflicting trademark rights for

a single domain name, are there any technological solutions?

E.27.1. No.

28. Are there any other issues that should be addressed in this

area?

E.28.1. No.

------------------------------ break ------------------------------

F. Other Issues

F.1 The IANA has served well at editing lists of names and numbers

used in Internet architecture, but has proved abominable as

leadership and setting policy of the DNS.

F.2 The Internet Society, by means of attempting to provide an

umbrella to the various "I*" organizations - IANA, IAB, IETF and

so forth, represents a concentration of power inconsistent with

organize evolution of the Internet.

F.3. International bureaucracies such as the ITU and WIPO have

not "offered to help" fix the problems in todays DNS, instead

they are trying to dictate law. If you want to kill or stifle

the innovative spirit of the Internet, this is how you would do

it.

F.4 The internet community, particularly name server owner/operators

must be encouraged, and helped, to find solutions to the problems

and realities of the current DNS crisis. Without consensus,

there will be a breakdown and the net will splinter.

F.5 Before any expansion of the name space is undertaken, the first

principle of the Hippocratic oath must be very carefully considered:

"do no harm".

------------------------------ break ------------------------------

###

Number: 364

From: Karen Brickley <KarenB@Prince.co.uk>

To: "'dns@ntia.doc.gov'" <dns@ntia.doc.gov>

Date: 8/18/97 5:28pm

Subject: NOI response from John Wood, Prince plc (in MS Word format)

A. Appropriate Principles

The Government seeks comment on the principles by which it should evaluate proposals for the registration and the administration of Internet domain names. Are the following principles appropriate? Are they complete? If not, how should they be revised? How might such principles best be fostered?

Yes, to the extent that they are consistent with the responses to questions below. Please note that more comprehensive principles are suggested below. See also the "Principles" document.

A. General/Organizational Framework Issues

Background

The current system may be summarised as follows:

The Domain Name System (DNS) is hierarchical and unitary in structure, distributive in method and highly scaleable. The DNS is defined by Request For Comments (RFC) 920 and 1591 authored by Jon Postel (USC).

A domain name is an alphanumerical representation of a unique IP address that corresponds to each host/node connected to the Internet, for example, PRINCE.COM corresponds to 194.200.207.65. Domain names can only contain the letters A thru Z and/or the numerical digits 0 thru 9. Also, there can only be up to 24 characters. The domain name hierarchy is read from right to left within the domain name space. Levels are delineated by dots. Therefore, the top level domains (TLDs) - .EDU, .ORG, .NET, .GOV, .MIL, .INT and .COM, and the country level domains - two digit country code extensions based on ISO 3166, for example .UK; are found in the extreme right of the domain name space. The US does have a country code ((CC), .US), and the .US domain name space is defined in RFC 1480. To the immediate left of the TLD/CC extension can be found the second level domains (SLDs), for example, .CO and .AC. There is also the ability to create a third and even fourth level domain, for example, CBT.PSYCHOLOGY.UCL.AC.UK.

The Internet Assigned Numbers Authority (IANA) has overall authority for IP addresses and domain names. The administration of the central database and directory containing all of the IP addresses and domain names is carried out by the InterNIC (Internet Network Information Center). The InterNIC was initially made up of AT&T, Network Solutions, Inc. (NSI), and General Atomics. European IP administration and technical co-ordination is carried out by Reseaux IP Europeens Network Co-ordination Centre (RIPE NCC). Internet IP registry services for the Asia Pacific region are handled by APNIC.

The TLDs are managed by NSI under a five year co-operation agreement with the National Science Foundation (NSF) due to expire on March 31, 1998. The country codes are managed by their respective registries, as are their corresponding SLDs. The .US domain name space is managed by the Information Science Institute (ISI) based at the University of Southern California.

The Berkeley Internet Name Domain (BIND) software resolves the domain name against its corresponding IP address. Versions 4.9 and above of BIND are produced by Internet Software Consortium (Mr. Vixie etc.). BIND is defined in accordance with RFC 974, 1034 and 1035. The latest release of BIND is version 8 (May, 1997). It is believed that as many as four of the root servers now run on version 8.

The server architecture follows a tree structure receiving replicated records from servers above it and caching the results that in turn can be accessed by servers below. The servers at the top end of the tree are called the root servers. The server at the very top of the tree is called the alpha server and it contains the central database including the records for all of the TLDs. It in turn replicates down to the other root servers on a daily basis all new records to its central database. Below the root servers sit approximately 35,000 name servers. Name servers are located in domains, that are in turn divided up into zones. Each domain has to have one primary server and for practical purposes at least one secondary server to act as a back-up.

There are thirteen root servers, and all but two are located in the US, (the NORDU server is in Sweden and the RIPE server is in Holland. NSI manages two root servers (A, J), ISI manages two root servers (B, L) and the US Defence Department manages two root servers (G, H). These root servers are managed primarily by volunteers.

The resolution process is as follows:

A DNS client requests via TCP port 53 of the name server (either in its domain or pointed at) the resolution of a domain name against an IP address. The name server queries the zones in its domain for the address. Each zone contains all of the domain names from its point downward. If the name server cannot locate the address in its domain it queries the root servers. The root servers track all of the TLDs.

The registration process is as follows:

An entity or person requests of an Internet Service Provider (ISP) a domain name. The ISP allocates an IP address from its block allocation of IP addresses from (ultimately) IANA (in Europe RIPE and Asia APNIC). If a prospective domain name is to be associated with a TLD. For example PRINCE.COM, then a request as to use and availability is made of NSI. Assignment of a domain name is done on a first-come-only-served-basis.

Where the domain name is to be associated with a particular country, a request is made of that specific registry. If the domain name is available then the IP address and the domain name are associated, placed on the database of the respective registry that ultimately reports the record to the central database on the Alpha server. Interestingly enough, there is nothing to prevent someone from requesting a domain name from any one of the 200 or so registries in the world, for example, in the UK, NOMINET. Nor of the registrar within their registry coming up with innovative second/third or fourth level domain name designations, for example .COM.US!

1. What are the advantages and disadvantages of the current domain name registration systems?

Advantages

The current DNS is simple and it works. There is an obvious limit on the number of domain names that can exist within a 32 bit domain name space. But, if we look at availability within for example the very popular .COM domain name space, there are still an available 37²² possible combinations of domain names.

To those that would say that the system is breaking down due to the explosive levels of demand for the Internet, their attention should be drawn to the Inverse Network Technology, Inc. report of June, 1997. The report indicated that the last quarter saw a drop in the number of failure rates, an improvement in the number of connection rates and even a speeding up of transmission of data over the Internet.

It is important to recognise the invaluable service provided by NSI with regard to the current DNS. There is no substitute for hands-on experience and a tried and tested infrastructure. NSI has created from scratch a scaleable, robust, and stable solution. It also provides an efficient cost-effective service. In fact, according to the OECD, NSI is the least expensive registry of the top 20 registries in the world. NSI for its part claims that, on average, 92% of domain name registrations are processed within 24 hours. This is no mean feat when one realises that 18 months ago it took NSI an average of 4 weeks to process an average of 50% of the applications received. NSI without a doubt represents the back office of the Internet.

Disadvantages

The registering body for the top level domain names (TLDs) including the .COM domain name space, is a private, for profit American company, based in Herndon, Virginia. As such it is accountable only to its shareholders for its actions.

NSI registers domain names on a first-come, first-served basis. Its operations are legitimised through a co-operation agreement with the National Science Foundation (NCR-921842). The initial agreement was for five years and envisaged a not-for-profit administration of the TLDs by NSI for and on behalf of the InterNIC. At a peer review in 1994 it was determined that NSI had provided an excellent service, but due to the demand for TLDs in general and the .COM domain space in particular, a charging model was now required. This finding led to Amendment 4 to the Co-operation Agreement (dated September 13, 1995). The amendment provided that as of September 14, 1995, NSI could charge an initial fee of $100 for two years and a bi-annual renewal fee of $50. The amendment also provided that 30% of the fees collected would be placed in an Intellectual Infrastructure Fund. Consequently, as of March 31, 1997, there was $20,560,000 in the Intellectual Infrastructure Fund.

Currently, 88% of all domain name registrations with NSI are for the .COM domain name space. NSI is registering an estimated 3,000 new domain names a day at $100 a registration. It has already registered over 1.4 million domain names.

RFC1591 had not envisaged a registrar being embroiled in legal disputes over domain name rights. However, the Juris, Inc., MTV, Hasbro Inc. cases, held that domain names were subject to trademark law. The growing problem of domain name hoarding/piracy by individuals and entities in anticipation of a pay-off from well-established trademark registrants unwittingly drew NSI into legal disputes between domain name holders and trademark registrants. Consequently, in July 1995, NSI released a guideline document for disputes entitled "The New Domain Name Dispute Resolution Policy."

With the increase in the number of cases and their increased complexity, NSI felt compelled to revise its policy and did so on September 9, 1996. Consequently, under the revised policy, upon the mere furnishing of proof of a trademark registration by a trademark registrant, the burden of proof moved to the domain name holder to show within thirty days:

The existence of a corresponding trademark registration to its domain name; or

Evidence of litigation in any court of competent jurisdiction; or

An indication of a willingness to be assigned a new domain name in substitute for the domain name in dispute.

Thereafter, NSI would determine whether or not to place the domain name on hold and in effect end the commercial viability of the corresponding Web site. Issues of due process and breach of contract immediately come to mind until you realise you have already signed your rights away! How? The guidelines are incorporated by reference into your contract with NSI. All you have is a two year right to use the domain name unless NSI sees it otherwise.

NSI's Dispute Resolution Policy does not take account of the fact that the domain name holder may have either a state trademark registration and/or common law rights in a domain name arising out of use. All that the policy does do is highlight the disparity between domain names and trademarks; namely, that currently there can only be one global domain name. The holder of the domain name is then in the enviable position of "winner takes all." This is to the detriment of all other users of the name regardless of whether their use is related to different goods or services and occurred only in, for example Arizona or Afghanistan.

On the other hand, under US trademark law, intellectual property rights may arise out of use and be protected by state registrations or federal registrations or the assertion of common law rights. However, trademark law is created, developed and enforced on a territorial basis. Wherein, organisations and individuals are able to co-exist through the provision of an international trademark classification system.

Thus, in the recent dispute between Prince plc versus Prince Sports Group, Inc. over the domain name PRINCE.COM, Prince plc being a purveyor of computer training and technical services could co-exist under trademark law with Prince Sports Group Inc., a producer of tennis rackets in trademark classes 41 and 28 respectively.

The effect of the Dispute Resolution Policy is to encourage registrants to assert their rights in a manner that would not otherwise be afforded it by a court. Consequently, it has led to an increased number of innocent bona fide domain name holders being placed under attack by aggressive trademark registrants who in effect attempt reverse hijacking of a domain name. For example, Prince plc had enjoyed two years of unopposed use of PRINCE.COM prior to Prince Sports Group attempting to use the Dispute Resolution Policy to get the domain name.

The private nature of the Dispute Resolution Policy Guidelines has enabled NSI to change and vary both the content and interpretation of the meaning of the guidelines. This has led to a great deal of speculation as to the meaning of the actual policy. For example, section 7 provides for litigation to be in "any court of competent jurisdiction in the United States". This has been interpreted by NSI to mean any court that is held to be competent by the US judicial system. Accordingly, Harrods and Prince plc were able to bring actions in defence of their .COM domain names in the UK courts. One should observe that the number of legal disputes directly involving NSI only amounts to some 30 or so cases, this figure pales into insignificance against the daily registration average of 3,000 or so new domain name holders. With a total to date of 1.4 million domain names.

The scope of contractual liability of the ISP and/or the registrar such as NSI is another area of great concern. Issues such as the extent of the agency role of the ISP for and on behalf of NSI is not clear. There is a need for a standard indemnification clause in the ISP contract, that specifies that the domain name holder indemnifies the ISP from liability in any infringement action brought by a trademark registrant against the domain name holder. Further, there needs to be a disclaimer placed within the ISP contract explaining that the domain name holder does not own the domain name, but merely has right of use. On the latter point, it is interesting to note that NSI has claimed the intellectual property rights in the central database. That is to say, it believes NSI has the intellectual property rights relating to the .COM domain name space. This raises really serious issues as to the ownership and use of the existing .COM domain names after the co-operation agreement between NSF and NSI lapses on April 1, 1998.

Whereas, the number of legal questions pertaining to domain names are legion and seem to multiply exponentially by the day, the amount of case law is myopic and the experience of the judiciary with regard to domain name disputes is very limited. This can only lead to a great deal of uncertainty and chaos in the short term, prior to there being a body of mature case law and statute in order to foster a satisfactory legal framework for domain names.

The recent investigation by the Department of Justice into alleged anti-trust violations by Network Solutions, Inc. has raised the issue of why was it that the US Government created a contractual monopoly in the first place?

The first point is that the NSI monopoly only relates to the TLDs and that there are some 200 registries around the world that administer the country code domains and their respective SLDs. Therefore, strictly speaking, NSI does not enjoy a monopoly within the administration of the Domain Name System per se. The second point is that the Domain Name System is hierarchical and as such requires the existence of an alpha master server, i.e. the creation of a natural monopoly. The third and most important point is the fact that the monopoly concerns only arose when NSI started to make money for, in particular, the registration of .COM domain names. Ironically, NSI states in the documents it filed in support of its IPO that it lost money in 1995 and only broke even in 1996. In other words, it is only now beginning to make a profit.

The events of the early part of the morning of July 18, 1997 highlighted some of the other problems with the current Domain Name System. Due to human error at Network Solutions, Inc., a corrupted database was replicated down within the zone file to the root servers. NSI recognised the problem and rectified it within less than five hours of its occurrence. However, it transpired that perhaps no more than two other root servers had any software in place to detect corrupt files. This vulnerability of the root servers is further exacerbated by the fact that the vast majority of people administering the Domain Name System are volunteers. They come to the job not with a certificate of competency or a pay cheque, but with a love for the Internet. In most cases this would not be critical. However, as the Internet becomes a significant commercial market place, it is essential that its DNS administration is fully professional and adheres to the highest standards possible, in order that there is a stable, robust and online market place in cyberspace. It should also be noted that although we are moving towards BIND Version 8. Currently very few of the root servers are using BIND 8.

2. How might current Domain Name Systems be improved?

The current Domain Name System may be improved by the implementation of the following principles: evolution not revolution, the creation of a Domain Name System that is representative, accountable and responsible to all of the stakeholders in the Internet. The solution must be international in scope, commercial in nature, and definitive in its legal framework.

International i.e. representing all of the countries that use the Internet.

Representative of all the stakeholders: governments, legal institutions, commercial entities, technologists and users.

The DNS must make its legislators, administrators and managers accountable to the treaty member governments of any new DNS and to its consumers.

Responsible such that all of their actions, decisions and new initiatives must be open, capable of change and scrutiny where necessary.

The DNS should be a commercial resource set within the context of being in the public trust.

There needs to be a single, integrated Internet.

There should be the ability to have shared access to the central database by the registries.

All entities involved in the administration and governance of the Internet require a proper legal standing and well-defined funding mechanism. Accordingly, a treaty is needed to establish a new Domain Name System. This treaty should enable there to be a partnership between government and industry, whereby, government provides oversight and legitimacy and industry provides a self-governance model and technical infrastructure to the new DNS.

NOTE: It is preferable that reform occur sooner rather than later, in that the system is currently small comprising approximately 1.4 million domain name holders, whereas by the year 2002 there maybe as many as 600 million domain name holders.

3. By what entity, entities, or types of entities should current Domain Name Systems be administered? What should the makeup of such an entity be?

There should be a secretariat and policy making body. The secretariat should support the policy making body and oversee the DNS. The policy-making body should develop policies, rules and regulations for the Internet including the DNS.

3.1 Administrative framework

Registries

Each country should have a registry maintained by a registrar to administer the domain names under the country code and respective SLDs. A domain name should only be issued to an individual or entity who resides in the country represented by the particular country code. Thereby maintaining the integrity of the designation of origin status commonly associated with addressing and trademark law. With regard to the TLDs, they should be issued in principle to entities that operate networks, non-commercial and commercial operations in more than one sovereign country, i.e. are international in character.

The registries should be supported by fees from their ISPs for domain name registrations. Each registrar of a registry should submit once a year an operations report to the policy making body of the DNS. The operations report should provide statistics on registration numbers and revenues derived, information on current operational resources and future plans to upgrade resources and train staff.

Regulations should be put in place to ensure quality and standards of registry performance that ensure a stable, robust and scaleable DNS. All employees who administer the Domain Name System should have to meet certain minimum competency standards determined by a practical and theoretical examination. Also, their equipment should meet specified minimum standards set by the policy-making body. Also, an inspectorate should be established to ensure that standards defined by the policy making body are adhered to by the registries as to the quality and competency of their staff, equipment and facilities. All registries should be inspected at least once a year.

Every registry should have adequate security and virus checking software, back-up facilities and a 24 hour help desk. Quality assurance and service are paramount.

There should be a central registry in place to protect the domain names of Internet users around the world. Therefore, at no time would a domain name holder lose his domain name even if its country registry went down.

Root Servers

The central database facility and root servers should be governed by similar principles and supervisory systems as the registries. For example, root server administrators should have to pass competency examinations, root servers should have security and virus detection software on them, and the alpha server administrator should be able to ensure coverage for any root server should it go down. The importance of the root servers and central database is such that they should be financed out of fees levied on all ISPs.

Registration Procedure

A prospective domain name holder would apply for a domain name space via an ISP and have to pay a fee for the service. The ISP would interrogate the country registry in order to check for the availability of the domain name space applied for. If it appeared to be available, the ISP would: file an application, take the domain name holder's fee, notify the domain name holder that such an individual / entity did not own the domain name, but merely had a right of use bequeathed upon it by the registry, and have the domain name holder indemnify the ISP against any threat of trademark infringement litigation.

If the application was successful, the ISP would issue a contract evidencing the arrangement between the ISP and the domain name holder. The latter would have three days to decline the terms of the contract. The contract should state that the ISP would be indemnified from any trademark infringement action that may be brought by a third party with respect to the domain name. It would also state in clear terms the responsibilities and liabilities of the ISP to the domain name holder and the extent to which the ISP could hold itself out as being a representative or agent of the registry.

The registry, upon receipt of the registration and application fee, would conduct a domain space and trademark search. The trademark search would be carried out by a competent examiner capable of carrying out a search of the appropriate Federal/National/State Trademark registry databases as well as those of the respective Secretary of States' Corporation database (within the US). If it appeared to the examiner that the domain name space was available, then the domain name space applicant would have the application published on a notice server for sixty days. In the opposition period, a third party may file an opposition to the domain name holders application. Grounds for such an opposition would include but not be limited to the third party having pre-existing trademark rights; the domain name holder's application / registration being obtained through fraud; and/or the legal incompetence of the domain name applicant to obtain a domain name registration. Also, the examiner may deny an application on grounds not only of unavailability, but also on the basis that the issuance of the domain name space would not be in the public interest.

Finally, certain domain names would be restricted to certain domain name spaces. For example, all sites related to content of a sexual nature, should be restricted to sites that have reference in their domain name to the word "sex."

Domain name status

Domain name registrations should be for ten years, renewable in the last six months of the period for further periods of five years. However, domain names that have not been used continuously for two years may have their registration revoked upon the bringing of a challenge by a third party. If a revocation petition is filed by a third party, the registrar shall give the domain name holder registrant sixty days from the time of notice of the challenge to appeal against the prospect of revocation. Revocation should not be granted if the domain name holder demonstrates that there has been continuous use or that within the next three months the said holder intends to actively use the domain space once again. The latter defence by a domain name holder shall be accepted at the discretion of the examiner.

THE POLITICAL FRAMEWORK

1. The Self-Governing Body: introduction

The urgency of the call for a commercial self-governing body should be understood. There is a need to define how governance would be conducted and what checks and balances would be provided to make the system accountable and not subject to abuse. What follows is a detailed structure of the proposed governing bodies. While some may consider the proposal to be a complex solution, we consider it essential to have a structure that will be able to effectively manage the Internet during its evolution into a mature system of governance, able to accommodate the explosive levels of growth that the Internet is undergoing both now and during the next five years. As such it provides the appropriate legal framework.

The self-governing body comprises a policy making body and a secretariat.

The procedures of the policy making body should be dynamic and flexible in order to mirror the dynamic environment of the Internet, wherein a year is not a calendar year, but in a most cases a four month cycle.

The policies of the policy-making body should be administered and supported by the Secretariat. The Secretariat should comprise of five departments: government, administration, legal, technological, and consumer.

3.21 The Policy Making Body

The policy-making body would comprise of two chambers - Upper and Lower Houses. All its members serving for renewable two year terms.

Upper House

The Upper House would consist of two elements, a Permanent Council and a legislative body.

The Permanent Council would consist of one government representative from each of the ten largest Internet markets (in terms of use). Each permanent council member would have veto power over legislation and the findings of any investigative committees set up in the Upper House. The permanent council would be able to initiate "fast track" policy, rules and regulation on the basis of a two-thirds vote of all of the permanent council members in favor of such action. Fast track means it shall, where necessary, be able to promulgate policy, rules and regulation within sixty days. However, all acts initiated under the "fast track" mechanism would have to be subsequently ratified by a two-thirds majority of both houses.

The Legislative Body of the Upper House would be on the basis of one seat per country signing the Treaty constituting the new self-governing framework. By a vote of 75% or more of the membership of the Upper House, other countries may be added to the body subsequent to the initial constitution of the Upper House. All proposed changes to the rules, regulations and policies of the Internet should require a two-thirds majority vote.

The Lower House shall be of equal status to the Upper House. Where a dispute over proposed regulations, rules or policy arises, there shall be a committee of both houses to resolve the differences and to formulate the new rules, regulations and policies. The committee will comprise five permanent council members of each of the Houses and will be expected to approve findings and/or prospective legislation on the basis of a simple majority.

A quorum for the purposes of meeting shall mean not less than 34% of the members of the Lower House are present within the House assembly when a roll call is made by the speaker of the Lower House. A quorum for the purposes of voting shall mean not less than two-thirds of the members are present at roll call.

Lower House

The Lower House shall comprise of a Permanent Council and a Legislative Body.

The Permanent Council shall comprise of the commercial representatives from the ten largest Internet nations. It shall operate in a similar manner to that described with regard to the permanent council of the Upper House. Except that a veto by a member of the permanent council of the Lower House may be overridden by a simple majority of the members of the Upper House permanent council. Why the latter? Ultimately, public trust should prevail over commercial expediency.

The Legislative Body shall comprise of delegations from each of the countries whose member governments ratified the treaty constituting the self-governing body. It shall operate in a similar manner to that described for the Upper legislative body, for example, rules regarding quorums for meetings. All votes on policies, rules and regulations effecting the Internet shall require a two-thirds majority.

The representatives of the Upper House are appointees of the respective governments, the delegates of the Lower House will come from nationally recognised and legitimate bodies that represent the stakeholders of the Internet, for example, the American Bar Association, the Information Technology Association of America, the American Management Association. There should be representatives (no more than 3 per caucus) from the legal, technological, consumer, and commercial communities within each country delegation. The head of the commercial community caucus shall act as the spokesman for the delegation.

Committees

Ad Hoc Committees

Ad hoc investigative committees may be convened by the Upper or Lower House. All such committees require a simple majority vote by their respective house to approve their convening, the composition of their committee and the scope of their investigation. Their findings and recommended actions shall require the approval of a simple majority of the house in order for them to be acted upon.

Departmental Committees

The Upper and Lower House shall convene committees to oversee and direct the five departments of the secretariat. These committees shall consist of ten house members, five from the upper and five from the Lower House. Committee resolutions shall be by a simple majority. Committee members shall serve for renewable terms of one year. All findings by such committees shall be submitted to the two houses for action. The chairperson of the committee shall come from the stakeholder community that the committee is ultimately following. For example, if the committee is overseeing the legal department, then a member of the legal caucus should chair the committee. Committee membership shall be by secret ballot and held within the legislative body of the two respective houses. With regard to a governmental committee it shall be the exclusive domain of the Upper House. Similarly, the commercial committee shall be the exclusive domain of the Lower House.

Miscellaneous: rules and procedures and constitutional considerations

It is suggested sessions/timetables of the new upper and Lower House should follow the pattern of the sessions/timetable of the United Nations.

Each House should elect a speaker who shall oversee and maintain the order, procedures and protocols of the respective houses. Where there is a tied vote within a house the speaker shall have the determining vote.

All votes shall be by secret ballot.

Upon a resolution by a House passed by simple majority an independent council may be appointed to initiate impeachment proceedings with regard to a member. Upon the conclusion of a independent council's investigation, a recommendation shall be submitted to the House that initiated the proceedings whether to do nothing, censure or expel the member. Such action will require a simple majority of the House.

Any changes to the self-governing body shall require two thirds majority of both the Upper and Lower House in order to be ratified and become part of the constitution of the self-governing body.

3.22 Secretariat

Government Department

The Government Department should liase and co-ordinate the activities of the legislative bodies with the respective treaty member states governments. It shall submit annually to all treaty member state a report of the activities of the self-governing bodies. It shall draft all prospective legislation. It will also provide the logistical and functional support to any investigation sanctioned by the legislated bodies.

Administrative department

The administrative department shall be responsible for the accountancy function, legal representation and issues arising out of activities by the secretariat and legislative bodies, marketing, public relations and sales aspects of the self-governing body. It shall ensure the smooth running of the self-governing body.

The Legal Department

The Legal department shall provide research and counsel for all the activities of the secretariat and legislative bodies. It shall create and run working groups investigating the following areas of law, international conflicts of law, intellectual property law, contractual law, constitutional law, telecommunications and computer law, and any other law impacted by the Internet. It shall submit annual reports on developments in these respective areas of the law to the legal committees of the upper and Lower House. It shall also have an office of independent council who shall investigate any breaches of the constitution and/or laws by members of the self-governing bodies. Finally, it shall be responsible for the constitution of the self-governing body and ensure that its policies and procedures are current and modified as directed by the legislated bodies.

The Technology Department

The technology department should administer and maintain the hardware, software and telecoms systems pertaining to the Domain Name System. It should also set up working groups with regard to the respective Internet standards and liase with pre-existing Internet standards bodies including, but not limited to, the Internet Engineering Task Force (IETF), the Internet Advisory Board (IAB), the Internet Assigned Numbers Authority (IANA) and the American Registry for Internet numbers (ARIN). It shall ensure that all registries and administrators for the self-governing body comply with all the standards relating to the Domain Name System, for example RFC2010.

Commercial Department

The Commercial Department shall support the activities of commercial entities wishing to do business on the Internet. As well as the activities of Internet service providers and other IT organisations involved with the infrastructure of the Domain Name System and/or the Internet. It will oversee and draft regulations and rules promulgated by the legislated bodies with regard to commercial activity on the Internet. For example, rules relating to advertising on the Internet. By virtue of the fact that it will be the commercial entities who finance the new Domain Name System, the commercial department that shall include a financial arm which shall oversee the distribution of funds designed to ensure the maintenance and development of the Domain Name System.

Consumer Department

This department will have a mandate to protect the rights of individual consumers who use the Internet. It shall have an Ombudsman who can investigate complaints, impose regulations and sanctions upon offending entities. The fees charged by the registries and Internet service providers will also be subject to review by the Ombudsman.

The underlying registries shall pay a fee to the self-governing bodies. The registries will be permitted to act for profit and therefore charge Internet Service Providers within their registry territory. Within the administration of domain names. In turn the Internet service providers may charge end-users for domain names.

4. Are there decision-making processes that can serve as models for deciding on domain name registration systems (eg network numbering plan, standard-setting processes, spectrum allocation)? Are there private/public sector administered models or regimes that can be used for domain name registration (eg. Network numbering plan, standard setting processes, or spectrum allocation processes)? What is the proper role of national or international governmental/non-governmental organizations, if any, in national and international domain name registration systems?

1. Are there decision-making processes that can serve as models?

The FCC in its trials and tribulations over the telephone numbering issue

(1-800) clearly came up with a structure and mechanism in the North American Numbering Plan (NANP) which would be useful as a model or a case study of how the reform of the Domain Name System may be undertaken. It is clear that where there is a resolve and a determination to find a solution, it is never out of reach. It is to be noted that one of the issues that arose out of the 1-800 issue was that of ensuring that an organisation not only got the requested 1-800 number, if it was available, but also, those numbers which were immediately surrounding that number. In light of the NASA.ORG and NASA.COM incident recently it is perhaps important to consider whether a prospective domain name holder may be able to get other domain names that are considered as being in close proximity to the domain name.

4.2 The Role of Government

1. Provides legitimacy

There is no doubt that there needs to be government involvement, but it must be well-defined and have a basis for its legal standing. Therefore, either an international organisation such as the United Nations, or an internationally recognised body with the delegated authority to confer effective legal standing upon a self-governing body will be required. Consequently, a treaty signed at inter-government level seems the appropriate mechanism to legitimise any self-governing body.

1. Provides oversight

The danger of a self-governing body becoming "a law unto itself" requires that it be held accountable to those treaty member governments that constituted it in the first place. Accordingly, there is a need for some form of oversight by government. Only government is likely to protect all the vested interests in the new Domain Name System and the rights of parties both large and small who may use the Internet in the future.

5. Should generic top level domains (gTLDs), (eg .COM), be retired from circulation? Should geographic or country codes (eg., US) be required? If so, what should happen to the .COM registry? Are gTLD management issues separable from questions about International Standards Organisation (ISO) country code domains?

5.1 TLDs

Historically speaking .NET referred to network providers ie NIC and NOC computers, administrative computers and network node computers, whereas .ORG was for "organizations that did not fit anywhere else" (RFC 1591). Due to the self evident vagueness of these definitions, a divergent range of entities have obtained these TLDs. Accordingly, it would seem appropriate to either redefine .NET and .ORG, or to close them to further additions. Because some rather notable entities have already utilised these domain name spaces it would seem ill-advisable to retire them from circulation. Thus, the .NET should refer to a server directly involved in the DNS infrastructure. The .ORG domain name space should refer to any non-commercial entities. The .MIL, .EDU and .GOV domain name spaces should remain and be administered individually within their respective communities.

5.2 Country codes

To the extent that is practical, entities who do business exclusively in the US should be required to use .us and the country code as is required of all other domain name holders around the world. The clear exception to this rule should be the .COM domain name space. The country codes provide a key to issues of venue and jurisdiction defining where a dispute arises between a domain name holder and a trademark registrant, in that the prevailing law should be that of the domain name holder.

5.3 However .COM remains

The .COM domain name space should be retained because:

The .COM domain name space could be used:

1. To support existing famous trademarks and domain names which represent multinational companies.

2. Entities that exist in countries or industrial sectors that necessitate anonymity. For example, a corporation in Israel may wish to do business globally, but is precluded from so doing due to the embargo.

3. Existing companies that have incorporated into their marketing strategies reference to their .COM domain name space and have incurred large expenditures in the support and development of their existing Websites.

4. The .COM domain is seen as a global designate for commercial activities. The word "Com" can be transliterated as an abbreviation for commerce in many languages and as such, further supports the global positioning. Consequently, in the commercial world, the proposal to scrap the .COM domain name space, for example, in favor of use of .co, .us, would be viewed with outright hostility. In particular, in light of the fact that compliance with the move from the .COM domain name space to the .US domain name space would be costly in both time and money. For example email post office administrators would have a huge and complex task in order to reconfigure their post offices from .COM to .Co or .US. (Please note that a global/international business should be deemed as one which derives or intends to derive at least 10% of its business revenue outside the country in which it is registered as its principle place of doing business).

The .COM domain names could coexist:

The .COM domain names space could be expanded by allowing the browser, upon selection of a domain name, to provide a drop down menu. All entities which share that domain name space eg PRINCE.COM would have a drop down menu which would identify that there is a corporation producing tennis rackets in the US and a corporation providing IT services in the UK. The user could then select the appropriate domain name and folder. Thereafter be taken off to the appropriate Website. The drop down menu would reflect in its methodology the international trademark classification system and ISO codes in its determination of the different entities that share the same domain name space.

5.4 TLD and ISO management are separable

TLD management occurs in the US. ISO management arises in and out of the host country. The sticking point is the management of the .COM domain names base in that its management and perceived value has been global in scope and demand. Companies with global business aspirations and/or actual business coupled with those desiring to penetrate the US market view the .COM domain name space as a vital entreé. Consequently many businesses around the world have both the .COM and their corresponding commercial country code domain name space designation. Interestingly enough, the use of the .COM domain name space has created problems with regard to commercial email. For example, where is jsmith@microsoft.COM actually located? Thus there does seem to be a good case for those organizations using .COM to also have a correspondingly commercial domain name space at the country code level so that people know where they are.

6. Are there any technological solutions to current domain name registration issues? Are there any issues concerning the relationship of registrars and gTLDs with root servers?

There have been cases where a dispute has arisen between a trademark registrant and a domain name holder where the court has directed that the parties share the same domain name space through either a drop down menu or icons. Upon selection of the domain name space by a user, the user is offered a choice as to which of the respective web sites the user wishes to visit.

Registrars, gTLDs and Root Servers

There is a need to separate the powers and administrative tasks with regard to the DNS. Whereas, the operational management of the DNS that occurs at root server level should remain non-profit, the registration of the TLDs and SLDs should be on a for profit basis. The rationale of this approach is that the DNS architecture requires administrators who merely manage the system and ensure its day-to-day continuance. Such persons or entities who make determinations as to who should be on the database and who should not be on, should not be motivated by money. The determination should be made merely on the basis of availability.

On the other hand, there is no doubt that the TLDs and SLDs have become electronic brand names and as such warrant commercial value.

Clearly, the party providing the domain name space owes a duty of care and has a contract to ensure the availability and maintenance of what is offered as a domain name space to the end user. However, the liability of the administrator of the root server should exclusively begin and end there.

7. How can we ensure the scalability of the Domain Name System name and address spaces as well as ensure that root servers continue to interoperate and co-ordinate?

We could attempt to rationalise the disposition of current IP addresses. Recover delegated but unused IP addresses if not already used - perhaps develop the class D and E addresses.

However in the long term DNS needs to take account of IP version 6 and BIND version 8. For example IP version 6 will move from a 32 bit Internet address value to 128 bit, alleviating enormously the demand for IP addresses. Finally, there will be a need, when IP version 6 is made available, to rationalise the process by which blocks of IP addresses are given to ISPs in order to ensure the most efficacious use of this resource and avoid a repeat of what has happened up to now.

Although the domain name system allows up to 24 channels including 0-9, A-Z and hence delivering 37²² permutations of the .COM domain name system, the real issue of domain names is that multiple organisations cannot share the same domain name.

As long as this is the case there will always be competition for a source resource.

8. How should the transition to any new systems be accomplished?

There is a need to form a joint government commercial task force to:

Ensure that all the relevant stakeholders have been informed of issues surrounding the reform and privatisation of the Domain Name System and give an opportunity to respond and provide input into the process through a series of worldwide summits.

Formulate a plan based on all current proposals and comment on the Domain Name System.

Post on the Internet and in all relevant media for a period of 30 days the proposed draft plan.

Formulate and post on the Internet for comment for 30 days. The technical specifications for the new system to the IETF.

Approve of the technical specification and creation of the appropriate software. Also conduct a pilot test of the software and then roll out to a control group the complete solution for analysis and evaluation.

Purchase of relevant hardware and software, buildings and hiring of personnel with regard to the new system.

Select and elect members and delegates to the legislative bodies and appoint the secretariat.

Creation of rules, regulations, procedures, constitution and all enabling laws in treaty member states to enable the system to go live.

Select and appoint registries and approve their respective administrators.

On completion of the above the Domain Name System can go live.

It is anticipated that it would take at least 18 months to complete all the above. Therefore, in the meantime, NSI should be given an extension to operate through the transition period.

9. Are there any other issues that should be addressed in this area?

There are other issues, but these will require subsequent submissions.

10. Are there technical, practical, and/or policy considerations that constrain the total number of different gTLDs that can be created

Although there are no real technical limitations to the number of gTLDs that could be created, there are practical and policy considerations that should constrain the total number of different gTLDs that are created. An increase in the number of gTLDs will place a greater burden on the administration and maangement of the TLD domain name space.

With every additional relevant domain name space, a large organisation or entity will feel compelled to obtain each and every relevant domain name space. Therefore if the rational for the new domain names was to facilitate the inclusion of new domain name holders, the reality is that large multinationals are far better placed to take advantage of the new domain name spaces and have more of a motivation in the short term so to do. Consequently, there will still be a fight for the finite number of domain name spaces available.

Each additional gTLD, in particular if no standard or guidelines are offered, acts as the basis for further litigation over domain name spaces. The motivation for the litigation may vary from the mere vague and ambiguous words in which the domain name space definition is couched, to the perceived value of the domain name space. For example, in relation to the .arts Domain Name Space proposed by the gTLD MOU, what is meant by the word "culture" may vary between the opinions of Senator Helms and the late Mr Robert Maplethorpe!.

In short, it appears that unless the new gTLDs are accompanied by clear and coherent definitions and standards/guidelines as to their interpretation, that these additions become more of a problem than a solution.

11. Should additional g-TLDs be created

No. Instead the existing .COM Domain Name System should be expanded as suggested earlier and the .US domain space should be utilised by non-multinational organisations whose primary markets are in the United States. Outside the US, entities should be encouraged to seek country code domain names in accordance with the guidelines of their local registry. The exception to this rule should be those multinationals etc who seek a .COM domain name space.

Please note that a global international business should be deemed as one which derives or intends to derive at least 10% of its business revenue outside of the country in which it is registered as its principle place of doing business.

Due to the unique nature of the gTLDs a higher fee could be charged for initial setup. The extra fee would justify the added expenses due to its global nature and perceived prestige in the market place.

12. Are there technical, business, and/or policy issues about guaranteeing the scalability of the name space associated with increasing the number of gTLDs?

To the extent that the DNS is finite, it leads some to fear they will be left out. However, the theoretical availability of the .COM domain name space is 37²² which suggests that it represents only a remote likelihood. Practical limits on the system may suggest it is a limit of approximately 1 billion records - enough for now - but changes will be needed later. In short, there is manageable scalability sufficient to meet current demand, and that is the message that needs to be made. More to the point is that with the introduction of each new gTLD the likelihood of litigation due to the uncertainty or increased demand goes up considerably. Obviously a new gTLD means more work for the DNS operations administrator, but if the manager of the American Numbering plan can cope, I am sure the DNS manager should also be able to handle the load.

13. Are gTLD management issues separable from questions about ISO country code domains?

gTLDs, in their administration, have taken on a global dimension and as such require a global perspective. Whereas, country code domains require only consideration within a territorial, legal, technical and administrative frame of reference.

TLD management is further complicated by the fact that it falls into two areas that can either be combined (as presently in the case of NSI) or separated due to demand:

DNS Operations service provision, i.e., the management of the actual central database and the configuration file; the configuration, maintenance, modification, and (when necessary) upgrading of the BIND software.

Registry services i.e. the provision of registration policy and administration, including the review, delegation and when appropriate, confiscation of domain names. The focus is on: the investigation into the availability of a domain name; the ensuring of its appropriate delegation, and where necessary, the resolution of disputes of domain names via arbitration.

On the other hand, due to lower levels of demand than TLDs, country code management can afford to combine the services and still provide adequate coverage for its territory.

Finally, it should be observed that the TLD registry is often distracted by increased levels of litigation activity surrounding the application and use of the TLD domain name space. Alternatively, country code management faces no where near the same level of litigious activity.

14. Are there any other issues that should be addressed in this area?

None, at this time.

15. Should a gTLD registrar have exclusive control over a particular gTLD? Are there any technical limitations on using shared registries for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist?

Exclusive control over a TLD.

The server model of the DNS requires that there be one primary or master server that replicates down. Thus, the system does promote the delegation of a specific TLD to a specific registrar with regard to operational management. However, this does not mean that there can not be competition for the award of the registrations of TLDs. This competition could be further fostered through shared access to the TLD database. Therefore, the operation managers competitive edge rests in how effectively he administers the system. On the other hand users will only register with an efficient and cost effective registration administrator.

Technical limitations on using shared registries

As long as the software being used is interoperable and configured to facilitate shared access to the registries, there is no real technical reason why it cannot be achieved.

Co-existence of exclusive and non-exclusive g-TLDs

Exclusivity is a matter of function not technology. The latter will allow for the co-existence of exclusive and non-exclusive gTLDs. But, such a co-existence may foster confusion in the minds of the consumer and a fear of fragmentation. However, the system will break down rapidly if the links and pointers in the software are not maintained.

It is likely that non-exclusive gTLDs, due to competition, will cost less than exclusive gTLDs and therefore the market may force out exclusive gTLD registries.

16. Should there be threshold requirements for domain name registrars? Who will determine these and how?

The requirements should ensure quality of service and competent execution. Operations administrators must have experience (a minimum of three years) and a defined level of skill, e.g. CNE, MCSE etc. Registration administrators must have both experience/technical qualifications with the Internet and legal/examiner competency/qualifications, e.g. three years experience at their respective PTO. With time, a standardised exam should be introduced and there should be periodic inspections of sites to ensure competency.

The registrar must submit a specification of resources including:

The IP address of the two principle servers (that must be operationally independent of each other) and the identity of all other relevant hardware, software and personnel. The submission must include reference to resources to maintain and update all records of servers, nodes, domain name holders, back-up facilities, software to detect and remedy viruses, to provide adequate security, the provision of power supply back-up and a mirror site in order to ensure minimal downtime of the servers and loss of records. The site must demonstrate robustness, scaleability and interoperability. It must run on the server (at least BIND version 4.9 with a plan to move to BIND version 8). There must be an undertaking to ensure compliance with RFC2010. The site must demonstrate its adherence with RFC's 974, 1034, 1035, 2065, 2136, 1996 and 1995. Also, a plan must be in place to ensure the compliance within one year of starting operations with RFC's 1886 and 1884.

The prospective registry must submit a three-year operations and financial plan. With regard to the latter, the entity shall have to demonstrate the existence of a $1 million comprehensive liability insurance and $500 000 working capital.

In the case of the Alpha registrar, in addition to the specification of requirements as stated above, it must demonstrate a plan to ensure that the server will never go down for more than five hours at a time and during such time there is an adequate back-up in place. It must also provide an undertaking that all records received for one day will be replicated down through the DNS tree by no later than the day and a half after being so received.

Requirements as to the domain name registries/registrars will need to be established through the policy making body. Similarly they will need to be supervised by an inspectorate trained and capable of supervising domain name registry operations.

17. Are there technical limitations on the possible number of domain name registrars?

Yes, the number of ISO country code domains and TLD's will define the number of registrars. There should be a registrar for each country, and one for .COM, .NET, .ORG, .EDU, .MIL and .GOV.

18. Are there technical, business and/or policy issues about the name space raised by increasing the number of domain name registrars?

The increase in the number of domain name registrars will in the short term require greater technical skills to ensure the correct configuration of the files. Such skills are in very short supply. Also, it may lead to a need to separate the TLD and root name servers in order to optimise the administration of the DNS. As a practical matter, there are obviously increased administrative headaches with regard to keeping all of the domain name registrars in the loop and connected to all the other respective domain name registrars.

19. Should there be a limit on the number of different gTLD's a given registrar can administer? Does this depend on whether the registrar has an exclusive or non-exclusive rights to the gTLD?

As a matter of principle a registrar should be allowed to operate only one registry at a time. By so doing, the registrar will be able to concentrate his skills on one TLD, and thereby, be able to specialise and maintain a high standard and quality of service.

As stated earlier, there is a need to ensure that there is only one registrar controlling the operations administrative aspect of each of the TLDs. It is suggested that the operations function be carried out under an exclusive licence. However, the registration function does not necessarily need to be exclusive, but can be offered by a number of entities within a commercially competitive environment via a shared database. Thereby ensuring that the end user obtains a professional and cost-efficient registering service for its domain name.

20. Are there any other issues that should be addressed in this area?

There are issues but it is not appropriate to raise it at this time.

21. What trademark rights (e.g. registered trademarks, common law trademarks, geographic indications, etc.), if any, should be protected on the Internet vis-a-vis domain names?

With the establishment of domain names as electronic brand names and their subjection to trademark law, there seems no alternative but to protect all trademark rights that may exist in domain names. However, because in so many ways currently a domain name is inconsistent with trademark law concepts and analysis, it must be deemed a sub-set made distinct by its anomalies. For example, domain names preclude the co-existence of businesses on the Internet that offer different goods or services, because domain names are unique.

It should be ensured that any domain name application that is purely motivated by a desire for piracy should be denied as a matter of policy. A simple test of legitimacy should be applied at the time of the application, like requiring the production of, for example, a company registration number, copies of the audited company accounts etc.

There should be no limitation on the number of corresponding domain names to trademarks that a corporation can attain. This is merely good housekeeping on the part of the corporation. Further, a corporation should be allowed to register a domain name consistent with its company name.

There needs to be consideration as to how to protect not only famous trademarks, but also secondary or inherently weaker marks that are operating as domain names, which are no less important to the corporate trademark registrant.

Finally, there needs to be a serious review of the implications of the Hasbro case and its interpretations of the 1995 Federal Trademark Dilution Act. This is because the decision is so pervasive that it may consequently usher in a wave of vexatious litigation, the likes of which has not yet been seen in this area of the law.

22. Should some process of preliminary review of an application for registration of a domain name be required, before allocation, to determine if it conflicts with a trademark, a trade name, a geographic indication, etc.? If so, what standards should be used? Who should conduct the preliminary review? If a conflict is found, what should be done, e.g. domain name applicant and/or trademark owner notified of the conflict? Automatic referral to dispute settlement?

There should be a process of preliminary review of an application for a domain name prior to its registration and issuance. I would refer you back to my response to question 3.

Components of a preliminary review should include: searches for the availability of the prospective domain name and the existence of any relevant trademarks. Also, there should be a period of publication/opposition, in which a trademark registrant is given an opportunity to challenge a prospective domain name applicant. Finally, there should be a declaration by the prospective domain name holder of an intent to use the domain name within three months of it being issued.

The standards that should be applied to the preliminary review should be that of a reasonable trademark/domain name examiner.

The preliminary review should be conducted by an individual who is competent to carry out such an examination. Such an examiner should have at least three years experience at the Patent and Trademark Office or the equivalent.

Where there is a conflict between a prospective domain name holder's application and a cited trademark registration, the domain name holder will have 30 days to overcome the conflict to the examiner's objections to the domain name application. At the end of the 30 day period the examiner may: decline to register the domain name due to the cited reference or allow it to be registered with the proviso that its use shall be exclusively limited to the domain name space or allow for unlimited use of the domain name. In other words, where the examiner directs that there shall only be a limited right of use of the domain name, it shall act as merely an addressing mechanism.

The determination of the examiner is final. Any appeal should be on the basis of an error in fact or law and a petition should be made to the appropriate Trademark Appellate Court.

1. Aside from a preliminary process, how should trademark rights be protected on the Internet vis-a-vis domain names? What entity(iies), if any, should resolve disputes? Are national courts the only appropriate forum for such disputes? Specifically, is there a role for national/international governmental / nongovernmental organizations?

Trademark rights should be asserted by trademark holders and relief sought through the courts. Jurisdiction and venue for the seeking of such relief shall be in the territory in which the domain name holder resides. As a matter of principle the registry shall promote the use of arbitration or informal negotiations between the parties. In the spirit of RFC1591 the role of the registry shall be limited to merely informing the parties of a dispute and facilitating their exchange of contact information. It shall in no way involve itself in the dispute between the parties, in particular, it shall not place any domain name on hold, but bind itself to the decision of the court or arbitrator as the case may be.

I refer you to my answer to questions 3, 22 and 23 respectively.

In addition, I would state that conflicts over trademarks would be reduced through the use of the international classification scheme within the .COM domain name space. Thereby, allowing parties with different goods and services to co-exist over the Internet rather than pushing them towards unnecessary litigation in an attempt to overcome the unnatural limitation of the DNS system and its inevitable result that the winner takes all.

There is no doubt that a global database of all trademarks/servicemarks/ domain names and company names needs to be created. Such a database should be placed under the control of the self-governing body running the Internet. The secretariat of the self-governing body would be the appropriate resource to create and maintain the database relying in part on the electronic data provided by all the country PTOs. Once created the database would facilitate more efficient registration and protection of domain name rights once they have accrued. In terms of the assertion of intellectual property rights arising out of the domain names with the incorporation of the international trade classification system and the emphasis on the utilisation of the country code-based registries, wherever it is possible, would ensure more effective territorial enforcement of rights.

Finally, as a matter of principle, the database should be made available to the public in order that every registrant should be able to check their own details.

Every domain name applicant should provide a basis for requesting a domain name. The basis should be consistent with the TLD requested. If such information is not forthcoming then the application should be denied. At a minimum the applicant should give his name, business or home address (whatever is appropriate in the circumstances), telephone and fax number and email address. Where the application pertains to a commercial organisation then evidence should be furnished in the form of a corporate registration and a set of accounts for the previous financial year. Such information should be evaluated by the examiner of the Domain Name Application. The criteria for the evaluation, will be done on the basis of the examiner's experience and where necessary will resort to the applying of principles analogous to those in trademark law.

The overarching principle for the evaluation of a domain name application will be that of ensuring a bona fide application does prevail, but that in no circumstances should an individual/entity intent on piracy of domain names be allowed to achieve their objective.

26. How would the number of different gTLD's and the number of registrars affect the number and cost of resolving trademark disputes?

Evidently the more registrars and gTLDs there are the larger the burden of management upon the DNS administration system. There is therefore a greater likelihood of litigation arising out of the consequent mistakes and disputes over domain names. With the increase in the number of disputes, the time for adjudication will be longer.

1. Where there are valid, but conflicting trademark rights for a single domain name, are there any technological solutions?

As mentioned earlier in this response, courts have ordered that disputing parties share their domain name space via the use of either icons mapping to their respective sites or a drop down menu that facilitates hyperlinking to the said site. In short, where there is only one domain name space, it is still not difficult for the website to facilitate the ability to link to other sites and thereby facilitate co-existence albeit rather uneasily.

28. Are there any other issues that should be addressed in this area?

None at this time.

This document has been written in response to the questions asked above. There are a number of substantial issues that have not been addressed in this document that can form part of a further clarification to this response as required.

John Wood

Prince plc

011-44-181-748-7448

Email: Johnw@prince.com

www.prince.com/prince

###

Number: 365

From: Marc Hust <mhurst@skyscape.net>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 4:47pm

Subject: NOI: A Canadian View

There should be grave concern regarding the current ISOC/ITU gTLD MoU plan that is being implemented.

Jay Fenello, of Iperdome Inc, is absolutely correct in his observation that this debate is not about the operability of Top Level Domain Registrations... it is about power. It is about the control of binding international law and policy and how these standards will be applied to this new universal communications tool known as the electronic information superhighway, the Internet.

For many years the Internet was run among a group of collegians as a vehicle to communicate among one another, to trade research data, or to monitor remote laboratories. Commercial use of the Internet did not occur on any scale for almost 20 years. Within this time this basketweave of intercity telephone lines and mainframes was managed, governed, by a small handful of people who, for no other reason, had the job by default as they were the only people who understood how the network *ran**could be run*. Law, policy and procedure are not as serious an issue when a network is a few hundred thousand hosts deep and does not overlap onto other continents.

Today, the Intenet is governed by very much the same people as *always* *it was before* (the IANA and the ISOC *can be* *are* considered as part of this group). They are definitely more experienced than before at managing the network and setting procedures but they have no where near enough experience to deal with the myriad of issues that have sprung up in relation to the communications industry and its effect on global trade.

The International Telecommunications Union has had little involvement in the Internet and the autonomous, community oriented, policy creating procedures (such as RFC's) that have evolved. This acquired level of cooperation has allowed "rough consensus" to be the de facto standard in how new policy and procedure is implemented on a network wide level. In fact many people have grown to accept that decisions made by this process, although they might not agree with the outcome, are the right decision for the moment due entirely to the level of input and consideration that is weighed before an idea gets imnplemented. In fact many have *characterised the Internet as being analogous to the democracy of ancient Rome* *used the analogy that the Internet is the second coming of ancient Rome's democracy.

The gTLD MoU

After a few years of the accepted RFC procedure attempting to tackle the task of adding new Top Level Domains to the existing roots Don Heath, president of the ISOC, created the IAHC in attempt to determine "if" new TLD's should be added to the root. The outcome was the unmandated gTLD MoU which, as many know, among the productive ideas it is to implement there are several very serious issues that effect international law, national sovereignty and non-scaling monetary issues.

Some of the more serious issues that need to be addressed are;

- Being a Canadian Citizen, I solicited the IAHC on several occasions to identify the Canadain government representative who had been consulted and who had granted permission to the IAHC/ISOC/ITU/WIPO allowing them to set rules and regulations regarding the operation of sovereign Canadian businesses that must be accountable to the Canadian Government and Canadian Law. This person was represented as being "an official with no name". I later verified that the Canadian government was never consulted.

In fact, the Canadian government has no official position on these issues as it feels the industry, the stakeholders, academics and business people, should decide what is best for the Internet.

- Much of the operating procedures are dictated by having to operate in the language of English only. This will complicate issues once the Internet becomes less English centric and more regional based especially in emerging nations of the southern and eastern hemispheres. This also is a violation of constitutional rights in many countries English is not an official language.

- The involvement of WIPO as the potential "final say" on trademark disputes will cause many groups to lose their right for legal recourse in a sovereign jurisdiction. While WIPO claims it has the support of its member nations and that its' member nations have no say in these affairs having signed treaties governing trademark and intellectual property issues, many governments were never informed that these liberties were being denied to them and their citizens in regards to these telecommunications issues.

- The International Telecommunications Union wishes to be the group that controls all of the domain name and IP space for the Internet. Unfortunately this group does not have sufficient Internet experience to allow for the nurturing that has been required by the community that would allow the Internet to naturally involve. The ITU has predominently been a standards oriented body. Until recently the Internet has been overseen, to a certain extent by the Internet Engineering Task Force. A community based effort that allows input from anyone who has a concern about how the Internet is being run. The IAHC circumvented the IETF when Don Heath sidelined "Draft Postel" in October of last year and handed the Domain Name debate to the IAHC. This group, that although were involved in the listserv process, held many of their discussions behind closed doors and refused to release the minutes of such sessions to the community as a whole.

It has also a fact that much of these closed door efforts come at the fear of large corporate entities such as Microsoft and Disney. These groups, who make the bulk of their revenues off of licencing of intellectual properties, fear a larger name space will make the policing of infringements on their properties that much harder to enforce. This is a contradiction of intellectual property law where many trademarks are industry and application specific. There is a distinct difference between the usage of properties such as "coke.food" and "coke.metals" that many corporate trademark lawyers refuse to accept as realistic. However this is a U.S. centric issue as the liability of loss of potential income, unlike in most other nations, is a just reason to seek retribution for damages or infringements.

The IAHC plan, should not be scrapped, but should be allowed to exit among other private, industry and regional government sponsored plans.

The Alternative

Groups that feel that this initiative should be left up to private industry have been waging their own wars for years. One camp in particular is the "Alternative DNS" camp.

This group feels that many different models should exist that would allow more choices and freedoms in how Internet Domain Names should be deployed, how they should be paid for, how they should be policed and whether or not they should be allowed to fall into private ownership.

One movement that should be supported is the model of many cooperating regionally oriented domain name services. Countries like Australia, Canada and India are in the process of forming National Root Server Confederations. These Confederations operate under the laws of their perspective sovereign governments and recognize many different combinations of domains in addition to the NSI legacy root servers. In countries such as Australia, where international transit is charged to companies by the kilobit, it has made economical sence to localize the traffic for DNS. It is estimated that a DNS server in Australia being secondaried in the continental United Sates could cost as much as $24,000 US per year to operate.

This issue is compounded when taken into consideration the various exchange rates that could be in effect. Fortunately Australia has a strong dollar.

Dispute resolution should be a mute point. However we must take into account the United States legal system and the trend of large cash settlements for potential lost business and damages. In other countries, international court cases are just to expensive and time inefficient to be an option for resolution. Businesses in these predicaments are compelled to negotiate settlements that are amicable to all parties. Under the eDNS program all disputes are to be handled in the courts of an appropriate jurisdiction before the root will act on rendering a decision. This removes the liability from the registry which is the only appropriate solution for a United States based dispute.

The Free Solution

By allowing competing "free market" models the impetus will be to promote negotiation, compromise and cooperation. Having several root server confederations recognize one another is the only way you will foster free trade, growth and true competition.

It is not in anyones best interest to have a fragmented name space, that is competing ".com's" or ".web's", that have parrallel addresses with different content. By opening up the name space to interested parties industry groups will be driven to offer higher levels of service and more innovative products and properties.

An open market would also allow for franchising and recipricating marketing agreements. Larger registries could allow smaller companies in emerging nations to compete on a global level for business that might not have otherwise had access to certain markets. Until infrastructure in all nations reaches a universal standard costs will always give geographically challenges companies distinct disadvantages within their market segments.

Last Thoughts...

The US government should be acknowledged for fostering the Interenet to where it is today.

The gTLD MoU is flawed if it is to be "THE" authority for DNS from now on. It should, however, be allowed to compete among other models to test its' viability.

Regional Root Server Confederations should be fostered thus allowing soveriegn concerns to be addressed in the subject jurisdictions.

Free market private models should be forced to negotiate there way into the root servers by cooperating with the regional/national Root Server Confederations.

Delay the formation of any Global Internet Government until after the Domain debate is settled.

These are my observations from over two years of being involved in the enhanced DNS debate.

There are no right or wrong answers and no one model should be forced upon the industry or the world community.

LET THE MARKET DECIDE.....

Sincerely

Marc Hurst

SKYSCAPE Communications Inc

Toronto, Canada

mhurst@skyscape.net

416.484.4739

###

Number: 366

From: <Jay@Iperdome.com>

To: NTIADC40.NTIAHQ40(dns)

Date: 8/18/97 5:21pm

Subject: Iperdome's Response to the NOI

Iperdome's Response to the NOI

This response has been prepared by Jay Fenello, President

of Iperdome, Inc. Jay has a computer engineering degree

from the University of Florida, and experience as a design

engineer who has worked on communications projects at IBM.

Jay also has an MBA in Entreprenuership from the University

of Arizona, through a program that consistently rates as

one of the best in the country. Jay has spent the last

several years providing entreprenuerial consulting to

assorted small businesses.

Iperdome is a startup in the new registry industry. It

was formed exclusively to offer personal domain names under

the .per TLD.

Background

Contrary to appearances, the Domain Name Crisis is *not* about

domain names. It is about control. It's about how the

Internet will look 20 years from now, and who will make those

decisions.

The immediate challenge facing all Internet stakeholders is

how to deal with the IAHC proposal. It is the result of a

process initiated by the IANA, and orchestrated by the ISOC,

ITU and WIPO. While we don't necessarily oppose these four

groups' involvement in Global Internet Governance, we do

oppose their unilaterally deciding to take over this

governance, especially when it was done behind closed doors,

without legitimate authority, and counter to Internet

traditions.

A, E, I, O, U, Y	= 0
B, F, P, V	= 1
C, G, J, K, Q, S, X, Z	= 2
D, T	= 3
L	= 4
M, N	= 5
R	= 6