From: "Eric S. Raymond" <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 12:39pm
Subject: Comments on the NTIA's DNS administration proposal

Comments in a MIME attachment as requested.
<a href="">Eric S. Raymond</a>

What, then is law [government]? It is the collective organization of
the individual right to lawful defense."
-- Frederic Bastiat, "The Law"

I have been an Internet user and developer for twenty years. I am a
technical expert in several areas relating to Internet technology;
but I have a slightly different perspective which is perhaps more
relevant to the policy questions NTIA is now considering. I invite
you to examine the following references:

The Jargon File

How To Become A Hacker

The Cathedral And The Bazaar

These should suffice to establish that I have been a careful and
analytical observer of the Internet's culture and its evolution over
that time. I can produce evidence that these distillations of the
Internet culture are widely considered authoritative on the Internet's
history, culture and folkways by the very hackers who built the net and
maintain it today.

Thus, I write not only as a technical expert but as an accepted tribal
historian and keeper of lore for the culture most central to the
Internet's development. I venture here to speak not only for the
hacker culture's head, but for its heart as well.

In examining the draft proposal at


I have several concerns about it.


First, I consider the list of governing principles dangerously
incomplete. Two more belong on the list:

5. Freedom of Expression

DNS and the DNS-related infrastructure must not be permitted to become
a means by which any party can impose viewpoint discrimination either directly
on the DNS namespace or indirectly on any other portion of the Internet's

Assuring that this remains the case is not just a matter of idealism
but of practical politics as well. It has been correctly noted that
the Internet interprets censorship as damage and routes around it.
Internet hackers will not, by and large, accept a DNS regime which
permits registry authorities to refuse domain name requests on content
grounds. Thus, some equivalent of this principle will be necessary to
preserve the authority of the new administration.

6. Protection of Small and Individual Users

In order to counter the tendency of institutionalized power and money
to gain undue influence, DNS administrative policy should be explicitly
biased towards protecting the interests of individual and small users.

Otherwise, fees will tend to drift upwards and policies will tend to
become gradually more restrictive at the behest of commercial (and
governmental) entities who find the antics of the small users messy and
inconvenient, and who have a specialist interest in capturing the
regulatory process.

I do not make this argument from hostility to commercial users; like
most hackers today, I am a friend to the free market and am delighted
to see the Internet sustained by a robust private-sector for-profit
infrastructure rather than governnment largesse. But I know many
hackers fear that short-termism and arrogance on the part of large
commercial users might smother the very attitudes and traditions that
give the Internet long-term value to all users.

The only way to prevent this kind of institutional drift is to build
in a conscious, constitutional bias against it. Accordingly, the
composition of the at-large block of board members, as proposed, is
wrong. In order to preserve the Internet's vital character, a clear
*majority* of at-large directors should represent "non-commercial and
not-for-profit" use.


The proposal assumes a dichotomy between (a) market-driven registries
with no portability, and (b) non-profit registries with portability.
This dichotomy is false. If both profit and non-profit registries be
permitted, and if at least one initial registry of either class offers
portability, the market *itself* will determine what mix of options
best meets user needs.

Given this, there is no reason to arbitrarily limit the initial creation of new
gTLDs. Instead of setting a number such as 5, why not simply require the
posting of a substantial performance bond? The amount of the bond should be
set to be within the resources of a medium-sized nonprofit, revocable on a
finding by the B.O.D. of your hypothetical oversight organization that the
principles of operation had been violated.

This approach would probably effectively limit the initial new gTLDS. But
it would have the great advantage of not institionalizing the theory that
gTLD space is a limited resource dispensable at someone's whim.


Most hackers readily concede the value of protecting trademarks; but
we are also conserned about abuses like Zilog's onetime attempt to
trademark the letter Z, or PepsiCo's more recent attempt to trademark
a shade of blue. Such abuses, if permitted to reach into the DNS
namespace, could pollute vast portions of that namespace.

This also issue intersects with freedom of expression. The principle that
Fubarco's trademark deserves protection must not be interpreted as a
license to deny issue of (say) a fubarco-sucks domain name to somebody else.
This is an issue about which the hacker culture feels very strongly.

The requirement that domain name applicants certify no trademark
conflict is problematic. How is an applicant to *know*, short of
being an IPR lawyer? If the act of so certifying creates legal
exposure, it will create a disproportionate chilling effect on small
parties and is therefore bad policy. If it does not create exposure,
why have it at all?

According to my proposed principle 6, the burden of proof in a trademark
dispute should be on the larger party. This should be *explicit* in the
dispute-resolution guidelines.

The proposal for a time limit on objections and dispute resolution is good.
In order to create a self-correcting market in enforcement terms, choice
of law should be at registry level.

The concept of a priori trademark "clearing" should be rejected
outright; it would be an invitation to abuse and inevitably lead to
favoritism towards large parties. This problem whould be better
handled a posteriori by better search tools which allow trademark
holders to automatically, frequently and rapidly search for


There is widespread feeling among core Internet users that NSI has
committed monopoly abuse of a privilege it should not have gained in the
first place, pricing access to the "classic" TLDs unreasonably far
above costs and offering very poor service. NSI's initial position in
the new market should not reward it for past misbehavior.

Accordingly, the "agreed-upon formula" for prices should be cost-plus
and in perpetuity -- and NSI should be required to make periodic
independent audits of its costs and its pricing formulas available on
the Web.


I would add to the Registry requirements the following *very important*
constraint: the databases, software, and source code used by registries
to fulfill their registry functions *must* be available for public
inspection over the net at all times. No exceptions, ever.

Only in this way (by exposing the registry process to full and
constant peer review) can the security, transparency, and fairness of
the registry process be maintained.


From: Andrew Parker <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 12:38pm
Subject: Registries vs Registrar Competition

The government is concerned about the technical management of
Internet names and addresses. And well it should be. Most fast-growing
companies have a plan for managing growth. How many companies or
organizations have experienced the kind of growth that the Internet has?
None. And yet itÆs growth is managed by a small handful of academics.

This is how I think Internet names and addresses should be managed in the

One organization manages the registry because this requires coordination,
and allow any independent, private or public company in any market or
region (as geographic location is not an issue for on-line registries)
compete for the business of registering domain names with the registry.
Modern information technology is sufficiently advanced to eliminate any
potential coordination problems between a single registry and a number of
competing registrars. InternicÆs single-handed management of both the
registry (database) function and the registration (sign-up) function, is
a testament to this. However, InternicÆs monopoly is unfair and most
people would agree that there should be competition at some level,
especially for such a large commercial operation. A single registry would
allow an individual or company to register its name with all TLDÆs at

Introducing competition at the TLD registry level would be wasteful
because it gives each registry ownership of that TLD. The Green Paper
argues that market forces would decide which new TLDÆs would survive and
which would not, but this would happen regarless of whether the new TLDÆs
are on a shared database or on independent databases. In fact, the public
wouldnÆt even know or care whether there is more than one registry.

And keep the registry as a non-profit entity. Who is the government
responsible to? The shareholders of NSI or the hundreds of million
Internet users world wide?

Introduce competition where it would serve the public the most: where the
technology meets the consumer - the registrar. The registrars are the
ones building relationships with individuals and companies wishing to
register names. Registrars are the interface between the people and the
database, and this is the area where competition would result in the
greatest increases and innovation in service and the lowest prices. I see
a day when domain names are bought and sold like shares in a company. It
is like the stock market - you donÆt want competition for the information
and backup system (the registry) you want competiton between stock
brokers (registrars).


Andrew Parker

US Citizen
The opinions expressed above are my own and not those of my employer.


From: Sharon Eisner Gillett <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 12:42pm
Subject: DNS comments attached, Word 7.0 format

CC: Sharon Gillett <>

Is it the intent of the Green Paper to vest responsibility for the management of each TLD database in a single, for-profit entity? This is neither technically required nor customer-friendly. I urge you to clarify and/or amend the Green Paper to allow multiple entities to share the management of each TLD's database, using technology that is readily available for such shared access.

A major problem with current domain-name management is lock in: a customer cannot switch to an alternative domain name manager if s/he is unhappy with the service provided by the current database manager. Switching requires the customer to change domain names, which renders obsolete all hypertext links based on the current name. As the customer has no control over the links pointing to him/her, switching names is impractical, resulting in effective monopoly power for the database manager. Not only is such a situation undesirable, it is also unnecessary.

Portable domain names are technically feasible and should be a requirement of a re-designed DNS, just as portable telephone numbers are a pre-condition of a competitive telephone system. I urge you to re-design this aspect of your proposed system.

Sharon Eisner Gillett

Principal, Victory Research

51 Everett Ave.

Watertown, MA 02172


From: Mohammad Masoumeh Akeem <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 12:46pm
Subject: Domain Name Draft

After reading the Green Paper and public comments on the Paper, there is
still one issue that isnÆt being addressed: The issue of international

The Proposal establishes competition between registries and between
registrars, but not necessarily international competition, in that there
is no guarantee that other countries will play a role in this

The issue of the registries is important, because there will always be
some level of monopoly and therefore potential for excessive prices, poor
service, and lack of innovation at the registry level. The question then
becomes, who will have control over the power of this/ these monopolies?
This power has formerly been granted to Internic. If there is one
non-exclusive registry for all gTLDÆs, as proposed by CORE, this would
eliminate any coordination problems from the simultaneous registration of
similar domain names. However, this monopoly should be a not-for-proft,
INTERNATIONAL organization. If there is a separate registry for each new
gTLD, each would have its own small exclusive monopoly. (A single
registry is the better option simply because it is non-exclusive).

Even though the Internet is international and Internet companies such as
registrars could virtually operate from any country, it is still better
to have a dispursed network of registrars for local representation. There
is a certain level of service and customer knowledge that can only be
achieved when customers are serviced locally. For instance, if a customer
wishes to telephone his registrar over a complication with the
registration of a particular domain name, he will want to make a local
call to an actual person speaking his own language.

There should also be a single dispute resolution mechanism with an
international component. There should be at least one representative from
all of the major countries in order to eliminate bias. Rulings should be
swift and final. It is also important for TLD name dispute resolution
that there be a single registry so that a company doesnÆt have to go
through half a dozen registries to check if another party has registered
its name in any of the new TLDÆs.

I believe the gTLD-MOU, while not totally perfect in my opinion, comes far
closer to these goals than your draft as it stands.

The international community cannot be ignored. In the future, Internet
usage by people living outside the US will far surpass usage by


Mohammad Masoumeh Akeem


From: "Eduardus Isworo" <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 11:50am
Subject: Comment on the green paper

Hereby I submitted a comment on NTIA's green paper
This file is a Microsoft Word 97 file

Get your FREE, private e-mail
account at













The Internet as a global telecommunication media grows rapidly in the past several years. The need to introduce new regulations may seem obvious, since the current system still needs improvement. The proposed system introduced by the NTIA however, should be evaluated to ensure its effectiveness. This paper will comment the NTIA’s proposed regulation from the private non-profit corporation view.

A new rule, set for the Internet domain name systems, in my opinion, should be based on several considerations, which are:

  1. The Internet as an international media
  2. Although the Internet emerged from the U.S. government investments in packet-switching technology and communication networks, the Internet is better considered as a worldwide communication network that eliminates national boundaries. The fact that most of Internet users, service providers and servers reside in the United States seems like giving the United States government the unofficial authority to control the whole Internet, but the evolution of the Internet also involves many related users, companies and even government authorities from many areas of the world. Therefore, the U.S. authority over the Internet is not considered a legal authority, since the worldwide use of the Internet. It is impossible for a country to govern the Internet for it has no legal authority to Internet users in other countries.

    All efforts to govern or restructure the Internet should be based on an international organization, such as the International Telecommunication Union (ITU), for it is impossible to perform them on national basis. The presence of IANA (The Internet Assigned Number Authorities) shows that the Internet world still has a close relation with the U.S government. This is the primary premise that should be underlined that the first move to reorganize the Internet domain name system is to make the Internet independent from any government authorities. If a new corporation should be formed, then it has to be formed by an international agency, not by any government.

  3. The need to restructure the Internet Domain Name System
  4. The main question that may arise from the proposed paper is do we really need to restructure the Internet domain name system. This leads us to reconsider the current situation. The fact that current system works well does not mean that it need not to be restructured. New regulation may be needed to guarantee a better system and a better environment where the Internet can grow even faster. One important matter that must be considered in restructuring the DNS is that there will not be any discrimination between individual users and companies in the proposed regulation. The Internet should be maintained as a global media, regardless the economic or other motifs. A new corporation formed to regulate the Internet domain name system should, therefore, covers many kinds of user interest, individuals and companies.

  5. Legal considerations
  6. Setting the appropriate rules to the Internet means considering the appropriate legal and economical aspects. Since the Internet eliminates the national boundaries, those considerations are hard to set. Even if all Internet users accept those issues, there are other issues that can disrupt the implementation of the Internet rules. For an example, we have to set a basis where a condition can be considered an illegal action of private property or not. The Second Level Domain (SLD) is where most conflict happened. The new system must ensure that there will be no conflicts. Therefore the system must set appropriate rules to assure legal rights of trademark owners. On the other hand, the systems must recognize individual rights. A new corporation formed to regulate the Internet must consider the way to resolve trademark disputes legally, which will lead us to the issue what kind of jurisdiction it should use.

  7. The effectiveness of the new regulations

One possibility of the new system to regulate the Internet domain name system, as proposed by NTIA’s paper, is to form a private non-profit corporation. It has to be assured, however, that the formation of that company will create an effective basis to regulate the Internet. To be effective, the non-profit agency should present users and providers and agencies from all areas. It means, the agency should absorb as many as it can Internet players from all over the world although it means that it will be difficult to make an agreement in the newly formed agency.

Another disadvantage of forming a new non-profit agency is that its effectiveness will be reduced because of national boundaries. It will be difficult to force a rule to companies outside the U.S. And even if the newly formed non-profit agency is an international agency, it will be very difficult to manage its authority. One solution for these problems in the proposed paper is to obey the rules of the country where the "A" server resides. It will give the Internet a basic legal consideration but will leave customers non-transparent systems, in which user has to know the law before a specific domain name can be assigned to the user. The Internet should conceal its complexity and for me it is necessary for the Internet not only to conceal its technical complexity but also its legal and political complexity.

The ideal private non-profit corporation to regulate the Internet domain name system

The creation of a new non-profit organization to regulate the Internet domain name system is a good way to regulate the Internet domain name system. However, since the Internet is a worldwide media, it is questioned whether the U.S. government is the one who has the authority to form such organization. It will be better if the newly created organization will be an international organization that acts independently from any government. The organization comprises all related organizations and authorities that play major role in shaping the Internet. For an example, this organization can be formed by the International Telecommunication Union (ITU) and consists of IANA (The Internet Assigned Number Authorities) and other regional IP registries (ARIN, RIPE and APNIC) and many other major players in the Internet world.

The main requirements for the proposed system to work well is to make sure that the rules create a fair deal for Internet users, the both companies and organizations and individual Internet users. The proposed system concerns most on companies, especially the trademark owners. It seems unlikely that this system will treat every user fair if it is based on trade mark owner’s needs. New approach should be considered so that everyone can be treated fair by these rules. The system to overcome the trademark problem, in which the registrants should agree that, in the event of a trademark dispute involving the name registered, jurisdiction would lie where the registries is domiciled or where the "A" server is maintained, is hard to accept. This system will require registrants from other countries to obey the law of one country and therefore make this system difficult to perform well.

The other thing that the U.S. government must consider is whether the U.S. government has the authority to govern all TLDs (Top Level Domain)? The creation of a new private non-profit organization with major participation of the U.S. government will make the U.S. government gets too involved with the organization. One can doubt that such organization will be free from government’s interest.The paper proposed by NTIA implicitly states that the U.S. government has the authority to rule the Internet. In my opinion, the U.S government authority should be limited in its national boundary and should not be brought to the Internet world; therefore it will limit the U.S. government authority only on the .us domain.

Comments and suggestions on the proposed system

  1. Jurisdictional provision
  2. One of the way to overcome the trademark problem, in which the registrants should agree that, in the event of a trademark dispute involving the name registered, jurisdiction would lie where the registries is domiciled or where the "A" server is maintained, is hard to accept. This system will require registrants from other countries to obey the law of one country and therefore make this system difficult to perform well. If the new private corporation is an international company, this problem will be easier to be settled down. The corporation, that consist of many organizations and agencies from all over the world, can propose a petition that in the case of trademark or other disputes they must obey to the law where the registries domiciled or where server "A" is maintained. One this petition is agreed by its member, it can have legal power to force its implementation. Voters are not restricted from companies or certain users. Based on the consideration that all users have to be treated fairly, all Internet users have rights to vote.

  3. Resolve mechanism of the trademark disputes
  4. The registries should be required to resolve disputes within a specific time period after an opposition is filled within a specified number of days. However, the proposed 30 days seems not enough to resolve the disputes. A 60 days time period is more appropriate time to settle any dispute between a trademark holder and a registrant.

  5. The decision making process
  6. The new governing body in its decision-making process must publicize the decisions before making the decision. The governing body should inform the public all of their activities and their options and give the public adequate time to response to the proposed decision. This will allow everyone a chance to say what he wants on a given issue.

  7. The creation of new gTLDs

As of there is a possibility that the gTLDs will rapidly change overtime, the new governing body should conduct a public vote as to what new gTLDs should be added to the already existing list. The governing body should also explain why each of it should be added to the system. It will avoid constant modification of gTLDs.



From: David Price <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 12:54pm
Subject: Domain Name Draft

ItÆs apparent from the Green Paper that the government would have
authority over resolving trademark disputes which could be a problem,
especially where there is a dispute between an American company and a
non-American company. This will create situations where the American
company is favored, or appears to be favored in the dispute, that is to
say that the judge would be biased. This is an important issue because
there is the potential for conflict - international conflict.
International law is not as infallible as other areas of law. Potential
for political tension is too high and this is a risk we cannot afford to
take. There has been a lot said and written about the great potential of
the Internet for the future of technology, communications, commerce and
education, but very little about the potential for political conflict
posed by the Internet. The Internet is unprecedented in its ability to
cross international boundaries, cultural and language barriers, and trade

The governmentÆs authority over trade mark disputes will make smaller
nations nervous and protective at a time when we are just beginning to
understand the benefits and realize the fruits of free and open trade.

Trademark disputes may become the spark that ignites defensive
governmental actions as each nation tris to defend its interests and the
interests of its businesses, fearing that the United States will secure
undue control of the Internet and the lionÆs share of its commercial

Embargoes, war... It may never come to that, but consider the fact that
in the new global economy, information is more valuable than natural
resources and whomever controls the Internet and its laws and its growth,
controls a major conduit of information.

Intellectual dispute resolution procedures are already set up within the
world intellecual proporty organization and they should have unbiased
jurisdiction over any domain name registration disputes.

Please, let us not take the first steps down the steep slippery slope of
American Government intervention. Keep the Internet open.

Best Regards,

David Price

Articling Student
Stanton & Jokes
The opinions above are solely my own, and are not meant to represent either
my employer nor my Internet Service Provider


From: L J Davies <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 1:09pm
Subject: Response to Green Paper

[An attachment was originally included here]


I have great pleasure in sending you my comments on the Green Paper. The
attached file, response.doc, is in MSWord95 format.


L J Davies

Research Fellow,
Information Technology Law Unit
Centre for Commercial Law Studies,
University of London

Response to the US Government Paper on the Improvement of Technical Management of Internet Names and Addresses


The proposals contained in the Green Paper were a welcome addition to the debate on the need to regulate and manage the Domain Name System. However there are several points which I feel need to be addressed or at the very least clarified.

A complaint that the proposals appear to constitute an effort by the US government to control the Internet and its regulation seem to be borne more from an underlying fear of the government's intentions rather than from an analysis of the Green Paper. The point, though made several times within the Paper, that the US government wishes to relinquish control of the management of both the Domain Name System and the allocation of numerical addresses needs to be stated with more force and clarity than at present.

The requirement for the new not-for-profit corporation to be constituted under US law is sound and sensible given its location. There is, however, a justifiable fear that this could entail US law, both state and federal, as being the principal law for all contracts that derive from the corporation. The issue comes down to that of the forum for dispute resolution. Though the US could be the correct forum for disputes between the corporation and registries, there may be a valid concern that the registries or registrars would be forced to contract under US law with domain holders. This is perfectly acceptable with registries, registrars, and domain holders based in the US but it would create difficulties for registries, registrars, or domain holders based in other jurisdictions. Any requirement to contract under US law could be viewed, whether it is justifiable or not is not the issue here, as an indirect attempt by the US government to control the operations of the registries and registrars. The issue of contractual independence needs to be clarified. The issue is of some concern and, at the sake of repeating the theme of the Green Paper, it might help to present the proposals as simply a set of proposals that address a specific issue. It should be stressed that the proposals do not, and could not, be seen to attempt to regulate Internet activity as a whole.

The difficulty of determining who would be eligible to stand for the board, in particular who would be classed as a Internet User has been discussed in other submissions and needs no further treatment.

One issue which should be kept in mind is the structure and activity of the corporation and the effect which this will have on anti-trust laws. Though the corporation will be incorporated under US law its activities will be subject to other jurisdictions and efforts should be made to ensure that its activities do not fall foul of the relevant anti-trust provisions. In order to ensure that this be so it might prove better to create two not-for-profit corporations, one to deal with the management of the Domain Name System and the other to deal with the allocation of numerical addresses. Given that these two functions are not mutually dependent this could be carried out with relative ease.

Another concern is the proposal to create several new gTLDs and to have each gTLD handled exclusively by a registry. This proposal will create several problems which will detract from the need to introduce competition into the registration and management of the Domain Name System.

The creation of new gTLDs will most probably lead to an increase, however slight, in alleged trade mark conflicts. It will, however, have a far greater effect on the perception that this will be a great concern and perhaps that the control and management will eventually fall under the sway of big business. The first concern cannot be helped if the gTLDs are to be created. Their creation appears to be an accepted fact and will most probably take place. The second concern will most probably be alleviated by sensible dispute resolution mechanisms provided by the registries or registrars, and indeed by a board that is not biased towards or against on or other of the players in the management of the Domain Name System.

The most worrying aspect of the proposal is that of granting a registry exclusive dominion over a gTLD. This would create several anomalies for the market. It would effectively mean that the registry with the domain that is perceived to be the most attractive would immediately attract registrations of SLDs, irrespective of their performance in the market. This would not only present them with an immediate advantage over other registries as to their long term viability but would also remove from the competition equation, the need to compete on their own merits. Other registries would be unable to effectively compete simply because their gTLDs are perceived to be less attractive and so attract far less registrations.

The second problem is that this proposal would effectively lock in domain holders once registered. They will effectively be forced to remain with a registry through its exclusive hold over the TLD. A far preferable solution, and one which should be technically feasible, though it will be technically complex, is to enable registries who qualify to jointly administer a TLD. Thus successful registries will be those who succeed due to their own merits and not simply due to their luck in the draw for the gTLDs, whilst domain holders will be able to move their registration to the registry of their choice and not be locked into a registry simply by way of the need or desire to hold an SLD under a particular gTLD. The fact that domain holders would be able to move registries would do more to encourage competition between registries than simply a hope that market mechanisms might discourage such behaviour ever could. Registrars should also be required to compete in a similar manner.

Thank you once again for the proposals which proved to make interesting reading and add greatly to the on-going debate.


Lars J Davies


From: Hsiaosu Hsiung <>
To: <>
Date: 3/23/98 1:18pm
Subject: Mitretek comments on January 30th discussion paper

Attached in Word version 6.0 for MS-DOS are Mitretek Systems' comments
regarding the January 30, 1998 discussion paper entitled "Proposal to Improve
Techical Management of Internet Names and Addresses".

Comments regarding our submittal may be directed to me at: or at 703-610-2931.

Hsiaosu Hsiung
Sr. Principal Engineer
Telcommunications and Networking Division

CC: H. Gilbert Miller <>

Mitretek Systems welcomes the opportunity to comment on NTIA's Proposal to Improve Technical Management of Internet Names and Addresses. In our following comments, we address the issue of early establishment of the not-for-profit corporation mentioned in your discussion paper. We also provide comments to the paper that you issued for public comment.

Mitretek agrees that overall policy guidance and designation of new TLDs and control of the Internet root server system should be vested in a single organization, one that is representative of Internet providers and users. The importance of representation of all of the Internet community, providers and users from around the world, cannot be overemphasized. As the discussion paper makes clear, the functions discussed are best performed in the private sector. While government funding may be appropriate in the transition and initial periods, the cost efficiencies of the discussed functions will improve when the community is paying for these functions. Buy-in and bottom-up consensus by all the stakeholders are essential to the initial and on-going success of this initiative.

In order to be operational by the target 30 September 1998 date, a number of issues and activities need to be addressed. Examples of these activities include:

1. Creation of the not-for-profit corporation
2. Establishment of a board for the not-for-profit corporation that would be acceptable to the Internet community
3. Transfer of existing IANA and NSI functions to the not-for-profit corporation
4. Transfer of the root system and the appropriate databases to the not-for-profit corporation

Accomplishment of these activities by 30 September is a daunting task. Existing government procedures may not be flexible enough or have the authority to accomplish these activities in the time available. Furthermore, as discussed above, the issues and activities necessary to implement this initiative are best performed by the private sector.

As a result, we suggest consideration of the following two questions:

1. In order to achieve the goal of getting private sector and Internet community buy-in, what is the minimum set of decisions that the government believes that it must make prior to delegation of this initiative to the private sector? That is, what is the minimum set of principles that must be agreed to prior to the private sector and Internet community taking ownership of this initiative?

2. Instead of starting a new corporation, can an existing not-for-profit corporation, one with the administrative and corporate resources already in place, and one already performing roles similar to the proposed coordinated functions, assist in the immediate implementation of this initiative?

Due to the urgency of getting the not-for-profit corporation in place, we believe that use of an existing not-for-profit corporation offers the ability to quickly implement this initiative. Mitretek Systems would be willing to establish an independent division inside of our existing organization to undertake the roles of the suggested not-for-profit corporation. Mitretek is a not-for-profit, private sector, public interest, technology and scientific corporation with expertise in telecommunications and network analysis. Our corporate characteristics(1) are similar to those suggested for the not-for-profit corporation (i.e., objective, independent, and free from any real or apparent conflict of interest).

Mitretek could establish an Advisory Board (i.e., similar to the proposed governing board) comprised of a minimum number of participants. After being initially established, the Advisory Board would become self-perpetuating (i.e., growing to ensure the appropriate and self-determined level of participation and size) and self-managing (i.e., identifying and addressing issues setting priorities, meeting dates, etc.). The Advisory Board would work to define specific roles and responsibilities for itself and the suggested not-for-profit corporation, based on a minimum set of principles accepted a priori. Further, the Advisory Board would provide definition of, input to, and review of the policy, administrative, coordinated, and other appropriate functions to be performed by Mitretek, acting as the suggested not-for-profit corporation. Also, for example, the cost required to implement and perform the functions of the Advisory Board and the not-for-profit corporation, could be proposed to, reviewed and accepted by the Advisory Board on an annual basis. These costs would then be used to establish an appropriate fees schedule and appropriately passed on to the Internet community. The ultimate success of the Advisory Board, and the not-for-profit corporation, rests solely on the ability to get buy-in and consensus from the Internet providers and users. The authority of the Advisory Board would derive solely from the continued participation (i.e., intellectual, operational, and financial) of the Internet community.

Mitretek is willing to further define and discuss this approach.

Responses to Comments

1. Comment sought on whether registries should be required to resolve disputes within a specified period of time after an opposition is filed.

Our view is that a domain name registration should be suspended if a dispute is lodged against the name. The dispute should be filed within 30 calendar days after the domain name is registered and made public. The private, not-for-profit corporation proposed in the discussion paper should have the authority to mediate the dispute and to resolve it within 60 calendar days after the dispute is lodged. We do not believe the authority to resolve the dispute should lie with the registries since they are not necessarily objective and may have bias towards one of the disputing parties during the resolution process. The not-for-profit corporation should provide a fair, unbiased forum for dispute resolutions.

2. Comment sought on the proposal for "clearing" trademarks across a range of gTLDs.

We agree with the proposal to "clear" trademarks across a range of gTLDs. We believe such a "clearing" process is the only reasonable method to avoid trademark disputes. In order to accomplish the "clearing" function, registrars need a search capability in all applicable registries to review proposed names, assuming such searching capability is available. The means to interconnect registries for the purpose of conducting the searches should be managed under the auspices of the not-for-profit corporation, which would act as an independent third party when designing and implementing the interconnections.

3. Comment sought on the jurisdictions for resolving trademark disputes.

Notwithstanding the points made in the discussion paper, we believe that the registrars (not the registries) should take primary responsibility for avoiding trademark disputes. Our reasoning is that the registrar should review proposed names in the appropriate registries before submitting the name to the registries. If the registrar fails to adequately perform this review and identify potential trademark infringements, the registrar should be held responsible for the trademark dispute. Therefore, we recommend that the location where the registrar is domiciled be chosen as the jurisdiction for adjudicating trademark disputes.

4. Comment sought on the pool of applicants for new gTLDs being limited.

During the transition of the address functions to the private sector, we recommend that no new registry providers be allowed. Instead, adopt the "pioneer preference" approach and limit the pool of applicants supporting any new gTLD registries to those companies already supplying registry services. After the not-for-profit corporation is in place, the
not-for-profit corporation will coordinate with the Internet stakeholders to allow additional companies to compete for business with the new registries that were created, just as those and other companies will be allowed to compete for business for all registries.

During the transition, NSI should continue to act as registry for the ".com", ".net", and ".org". However, as soon as possible, preferably before 1 October 1998, the not-for-profit corporation should establish a competitive environment(2) to perform some or all of NSI registry functions. NSI should be allowed to compete for these functions like any other bidder. If the competitive environment cannot be established until after 1 October 1998, then arrangements should be made to continue NSI's contract until the contracts for all three registries can be awarded.

NSI's registrar functions should be similarly competed during the transition phase. NSI should be allowed to compete for these registrar functions like any other bidder. The awards should be based on merit and competency and need not be limited to "pioneer preference" contractors.

5. Comment sought on new gTLDs being made immediately available.

Regarding the addition of five new registries, each of which would be initially permitted to operate one new gTLD, our view is that some new registries and gTLDs should be created, although five may not be the appropriate number. For example, we suggest that one of the new registries, in some way, address the depletion of unique addresses in the ".com" space. We suggest that a new registry with the ".tmk" gTLD be created to reflect that an address such as "companyname.tmk" indicates the "companyname" is the trademark of the company. Under this arrangement, some holders of "" addresses would have to migrate to the ".tmk" registry.

We recognize the difficulty in dealing with the trademark infringement issue. Since no easy solution is now available, the not-for-profit corporation will need to coordinate solutions with the Internet community of users, as well as those applicable international organizations and jurisdictions dealing with the protection of trademarks and intellectual property. However, we offer the concept of the ".tmk" gTLD as a possible solution to minimize the possibility of trademark infringement.

We believe that merit exists in the combination of gTLDs and ccTLDs, of the form "", to further distinguish users who wish to use identical company names as part of their address. The combination of ".tmk" and ".cc" provides additional differentiation of address holders and should help avoid trademark conflicts. As has been suggested in other comments to the discussion paper, those address holders that do not wish to be associated with a particular country code could use the gTLD ".int" to reflect their international status.

Hsiaosu Hsiung
Sr. Principal Engineer
Telecommunications and Networking Division

Dr. H. Gilbert Miller
Vice President
Center for Telecommunications and Advanced Technology

1. Mitretek is a not-for-profit, private sector, 501(c)(3) tax-exempt corporation. Mitretek is a public interest, technology and scientific company with expertise in telecommunications and network analysis. There are no ownership rights and no stock in Mitretek, which is held in trust for the public by a self-perpetuating Board of Trustees. We have no affiliations or associations with telecommunications services, including Internet Service Providers (ISP) or product manufacturers. We work exclusively for users of telecommunications services. We do not manufacture hardware or software products, and we refrain from competing with any telecommunications providers. Mitretek imposes on itself and staff stringent conflict of interest policies. As a tax-exempt organization, we are prohibited from political lobbying or influencing. As a not-for-profit organization, our corporate and individual performance is not evaluated against a 'profit bottom-line'. Rather, our corporate and individual performance is measured against public interest and technical metrics of quality, utility, and responsiveness. We are experienced in, and have the facilities for, the handling of national security classified information, as well as highly sensitive, proprietary information from telecommunications providers. These corporate characteristics ensure that our work is objective, independent, and free from any real or apparent, financial or technological, conflict of interest.

2. One example of a competitive environment, and there are clearly other examples, is the not-for-profit corporation establishes rules to conduct competitive procurements to award initial contracts to the registries. In this example, each registry is associated with only one TLD. The not-for-profit corporation would conduct competitive procurements to award the contracts for each individual registry to a single contractor, with each contract having a limited term. The not-for-profit corporation would compete the contracts for the registries at the end of each contract's term. A contractor could be allowed to compete for more than one registry. As part of the procurement rules, the not-for-profit corporation and its advisors would decide how many registries one contractor could simultaneously operate.


From: "J.Kammerer" <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 4:34pm
Subject: Filing

Seems to be working well. Why not leave well enough alone?

John Kammer


From: "Karl Denninger" <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 12:47pm
Subject: Submission from MCSNet / Karl Denninger, DNS Position Paper

Please find enclosed our submission for the Department of Commerce Inquiry
being performed at the present time, with comments due to close today, March
23rd, 1998.

Please acknowledge receipt of this transmission.


Karl S. Denninger


March 24, 1998

Department of Commerce

Docket Number 980212036-8036-01

RIN 0660-AA11

Proposed Rule and Request for Public Comment

ATTN: Karen Rose

Please find enclosed MCSNet’s response to the NTIA document referenced above in regards to the Domain Name and IP Number Assignment System for entry into the public comments.

It is our hope that this submission, along with the many others which you have received, will assist the US Government in formulating public policy which is in the best interests of all stakeholders and users of the Internet.

If there are any questions regarding this submission, or if you or any member of the Department’s staff would like to discuss the points which we raise in this memorandum further, we remain at your disposal in the public interest.





Karl S. Denninger



Biographical Information On The Author

Mr. Denninger is the majority owner of Macro Computer Solutions, Inc. The firm was founded in 1987 to serve the Chicagoland regional marketplace with computer integration and networking solutions of various kinds.

During the 1980s, MCSNet was a major hub for Unix systems communication using the UUCP protocol after AT&T shut down its regional gateway, "ihnp4". MCSNet, in concert with a few other organizations, was responsible for preventing the collapse of inter-system communication over electronic mail and Usenet following that occurrence, with this pre-dating the widespread commercial availability of true online Internet connectivity.

Mr. Denninger’s networking and computer industry experience stretches back nearly fifteen years. He is a current member in good standing of the Advisory Council (AC) to ARIN, the American Registry of Internet Numbers, has served on the Board of Directors of the local chapter of Uniforum (a national Unix-related user’s group) and has been active in the domain name system controversy as both a participant and an affected party since the beginning in September of 1995.

Background Information

In September 1995, the entire world of Domain Name Registration was turned on its head. It was on that date that Network Solutions, Inc., obligor of a cooperative agreement with the National Science Foundation, was freed to and began charging for domain name registration and renewal.

At the time the Domain Name System held approximately 150,000 names in the COM, ORG and NET zones. Within the next two years that was to increase by ten-fold, to over 1.5 Million domain name delegations.

NSI’s actions were particularly onerous for several reasons:

    1. They were imposed on prior registrants, including our company. MCSNet has held the domain name "" since 1988, and has been using it in interstate commerce since that time. The original domain registration agreement which we entered into was not even filed with NSI – it was, instead, filed with General Atomics, an organization which no longer is involved in the domain name system.
    2. MCSNet has never consented to any modification of our domain name agreement with the US Government or its agents. This includes, but is not limited to, the cost of that registration, the terms of that registration, and the procedure under which disputes would be mediated and ultimately settled. All of these things were changed by NSI since our original domain agreement was filed with General Atomics in 1987. It is the position of MCSNet that these unilateral changes amount to an improper modification of an agreement entered into with the US Government via its agents. It is further our contention that such modification has materially harmed our firm and tens of thousands of others like ours throughout the United States.
    3. NSI is, for US companies, the only reasonable and available domain name registration system with commercial-grade operations. The .US top-level domain has been mismanaged by the IANA in the extreme, including the grant of geographic monopolies to organizations with no business presence in the areas involved, lack of verification of appropriate infrastructure, and near-complete absence of oversight. Our customers have been materially harmed by the decisions of the IANA in this area, including the City of Aurora as recently as 1997. As this remains the only other choice for a US based organization, NSI has an effective monopoly on the process at this time, and has had that monopoly since September of 1995.

In September of 1995, MCSNet, along with others in the Internet industry, petitioned the IANA to open the root domain (".") to competition and allow us, along with others, to participate as registrars and add additional top-level domains.

To date the IANA has refused this request.

In 1997, the IANA, under increasing pressure including two filed and active lawsuits (one of which has since been dismissed by the plaintiffs) tossed the "hot potato" to a group that was hand-picked by Mr. Postel known as the IAHC.

This group begat what is now called "CORE", and the gTLD/MOU. We believe that the process by which this was undertaken was and is corrupt, in that no attempt was made to adhere to the principles of representation for the various stakeholders and users of the Internet namespace. To be specific, we believe the IANA and IAHC violated reasonable process for a public proceeding by:

    1. Hand-picking the representatives, with the primary person doing the selection being Mr. Postel. Mr. Postel was in turn not accountable to the public interest. We also believe he has a severe conflict of interest himself due to his grants of monopoly registration status to multiple firms throughout the United States for the .US TLD.
    2. The IAHC, while claiming to take public commentary, in fact had several members on the committee who instead engaged in a public campaign of harassment and name-calling against any individual or organization who opposed their specific agenda. This included but was not limited to allegations of "piracy" of the root of the DNS system.
    3. The IAHC, contrary to the public interest, drew up a process that would give the IAHC and their progeny exclusive control over the root domain for all TLDs not owned by a sovereign. Their model further dictates a business model (that of shared control over all non-country TLDs), grants exclusive oversight to themselves, including the power to levy fees amounting to a tax, and provides no public oversight of this process. Further, the IAHC deliberately moved the governance of this group to Switzerland with the effect of evading United States anti-trust law. In short, the IAHC and CORE created an exclusive monopoly system for control over all DNS registry operations. This was cloaked in the guise of "competition" by allowing anyone who wanted to conform precisely to their rules and regulations to "buy in" to the monopoly. However, this does not change the fact that CORE has explicitly rejected being one of many forms of operation for TLDs – they have insisted on being the only form of management permitted.
    4. CORE’s business model, in fact, appears to be viable only if they can maintain complete control over all non-country TLDs and their registrations. Should such control fail, and CORE TLDs be a minority of those available to customers, the free market will determine pricing and it is highly likely that their business models will fail. Should CORE be able to obtain exclusive control over TLDs, they may determine pricing in a cartel-like fashion. The trend is already clear – we have yet to see pricing from a CORE registrar which is under $50.00 per year, and some are publishing fees at the present time as high as $95.00 annually.
    5. CORE was the first, and to date, only registrar who has deliberately embarked on a course of action intended to fracture the root of the DNS system. By claiming two TLDs that the IAHC and CORE knew at the time were in prior use (.ARTS and .WEB), CORE has put itself in the position of forcing people to choose between the existing TLDs and their version. This course of action is not in the public interest, and further, it opens the door to serious questions of legal liability. It is important to note that no other DNS group has done this.

At approximately the same time, ARIN, the American Registry for Internet Numbers, was created out of NSI’s IP number-assignment capacity. The flaws in the procedure, and the myriad ways in which this organization can cause restraint-of-trade and other interference with legitimate Internet commerce will be explored in the last section of this submission.

The Green Paper

MCSNet applauds the Department of Commerce in its intervention in this matter. We are particularly pleased that CORE has been prevented from completing what we believe was an active taking of the entire top-level domain infrastructure. We also believe that CORE intended to and, if allowed to proceed, will destroy essential civil and legal rights in Cyberspace for American Citizens.

However, we remain concerned that much of what the Green Paper has laid forth is based on "bad science", and as such a number of conclusions not supported by the facts have made their way into this proceeding. To the extent possible, this response will attempt to address these points one at a time. Unlike many of those who have espoused their views on these topics, we will not attempt to present a proof by assertion. Rather, we invite the Department of Commerce, along with any other interested parties, to investigate the facts and foundations upon which these arguments will be made on their own.

Misguided Beliefs

The following points, while distasteful to some people within the Internet community, are nonetheless factual statements. Many individuals and even some organizations have attempted to argue that the axioms below are simply impossible.

I urge the Department of Commerce to very closely examine the claims that have been made by many which are counter to the axioms in this section. The Department should look specifically for the net benefit to the organizations who would advance positions contrary to the below, and carefully verify the truth of any such assertions. Asking "experts" is insufficient, as many of the so-called "experts" on the Internet have vested interests in the answers they may proffer, and absent being under oath are free to lie outright to the Department’s representatives. For this reason it is crucial that the Department seek proof of any assertion raised in these matters – including those which I present below.

    1. There is and can be only one "root" nameserver set, and one "root" zone. False. MCSNet, in concert with many other organizations, has been operating on a completely separate root server set for over a year – with no operational problems. In fact, MCSNet has staked its continued operation as an ISP on the stability of the eDNS Alternative Root System. We’re not the only ones dong this – there are dozens, even hundreds, of organizations pointing to these alternative roots, including firms in Australia, the eDNS system, and others. The only harm which can come to the Internet as a whole is through an organization attempting to "claim" a TLD which another firm or group has already put into service, thereby causing a query to resolve ambiguously. It is only in the case of intersecting sets of TLDs, where the intersections are ambiguously resolved, that problems will arise.
    2. (Insert name) is trying to steal control of the root and/or a TLD. False. In fact, the only conflicts recorded in the top-level domain system (TLDs) were created by the organization that has made this charge! CORE has claimed as theirs two top-level domains (ARTS and WEB) which were in prior use by another organization for a significant period of time, with actual, live commercial registrations, prior to the IAHC and CORE being constituted. It is interesting to note that of all the organizations that are actually operating in a live environment today, only CORE has attempted to undertake this act. Those in the Alternative Registry camp have steadfastly observed the rights of others to develop TLDs without interference.
    3. If we open the root to all, the domain system will collapse. False. We know this to be false, because at present, the majority of the root servers are configured with the entire COM, NET and ORG zones on them. This constitutes more than 1.5 million domains! No reasonable individual has postulated that under any scenario there would be even 1/100th of that number of TLDs in use. Even at 1/100th of that load, this is 15,000 TLDs! If we were to force NSI to pay for their own domain servers for COM, NET and ORG and add 15,000 TLDs tomorrow, we would decrease the load on the root server system by ninety-nine percent. To suggest that the Internet DNS system would collapse under such load is intentionally misleading and dishonest. In fact, if we were to do that tomorrow, performance from the root servers would rise to levels never before seen on the Internet as 99% of the load would instantly evaporate. We must also take into account the changes in caching behavior at the root level from an addition of a large number of TLDs. This effect, however, is minor in comparison to the effects of removing more than 1.5 million entries from those same machines.

It has been suggested that the current load on the root servers is approximately 500 queries per second. If we added 15,000 TLDs and removed COM, NET and ORG, forcing NSI to pay for their own infrastructure, the query load would drop to 5 queries per second, assuming equal caching behavior. That assumption, however, is fallacious – so let’s assume that we degrade cache coherency by a factor of 10 (one full order of magnitude). We are still at 50 queries per second in this scenario, a rate of processing that is easily handled by even an inexpensive Pentium machine!

It is unfortunate that from misguided beliefs and starting points one develops policies and procedures that, upon examination in light of the facts, make no sense. This is where the Green Paper fails in its essential purpose.

It is incumbent upon the Department of Commerce to verify the above facts. Further, it is incumbent upon the Department of Commerce not to issue final policy based upon "junk science" and doomsday scenarios that just simply are not true.

Guiding Principles As We Move Forward

We believe that the final policy promulgated by the Department of Commerce with regards to DNS issues must take into account the following facts:

    1. The US Federal Government provides funding for six of the existing "IANA" root nameservers, either directly (ie: those operated by the US Department of Defense) or indirectly (those operated by either public Universities or under contract through the NSF). As such, it has the unquestioned right and ability to control the policy for additions to the root zone (".") within those machines.
    2. The US Government does not have the right to compel any organization, whether it be a private concern, ISP, or other network operator not funded by the US Government to use the IANA-sponsored root servers. Any attempt to impose control over the root of the DNS system which violates the reasonable sensibilities (or the rights) of those individuals, businesses, and other organizations risks their defection to other, competing root DNS systems.
    3. Such defections, if and when they occur, are already proven not to "break the Internet". Claims to the contrary are false. In fact, a good part of the Australian Internet infrastructure today is running on alternative roots – as are several sizeable ISPs in the United States. Hard proof exists in the form of these organizations that no threat to stability is posed by these events, or by organizations forming coalitions of root server operators to best serve their constituencies. In fact, there is significant theoretical evidence to suggest that it is in the best interest of closely-connected sites to do precisely this in order to improve service to their customers by avoiding highly congested links.
    4. It is in the public interest for the US Government promote maximal diversity in the TLDs which are available. Areas of diversity and potential trouble spots which should be investigated and taken into consideration by the Department of Commerce rulings include:

A Roadmap

MCSNet strongly urges the Department of Commerce to issue regulations and guidelines controlling the entry of new TLDs into the government-controlled Root Nameservers under the following conditions and guiding principles:

    1. Any Organizations May Qualify as a registry, and operate one or more TLDs for their benefit either individually or as a group with others in a shared model. The barriers to entry in this marketplace must be strictly limited to those that impact the operational integrity and stability, as well as accountability to the public, of the system as a whole. It is specifically in the public interest that low-cost TLD providers enter the marketplace – including user groups, trade associations, "Freenets", and others with an interest in public service. We believe that the following criteria are appropriate:
    1. The US Government, in those root servers it controls, must aggressively protect the public interest by investigating any significant concentration of market power in the Registry Industry. This includes, but is not limited to, preventing an organization from registering all pr a significant number of TLDs in a given category (ie: "Legal"; .LAW/.LAWYER/.LEGAL/…..). Existing anti-trust law appears to be more than sufficient to deal with any potential or actual problems in this area. It is, however, crucial that such registrations receive the highest level of scrutiny where it appears that a registry is attempting to control a given category of product or service.
    2. To obtain additional TLDs, the Registry must show market acceptance of the existing TLDs or dispose of them. We suggest that to obtain an additional TLD that a registry must show through a certified public audit (perform by a firm such as Arthur Andersen) that it has a significant number of actual, active and in-use registrations. We believe that this number should be set somewhere between 5,000 and 25,000 active SLD registrants. Note that a firm which registers 100 domains in a TLD would count under this system as one registrant; this restriction is necessary to prevent "stuffing the ballot box". This procedure should be repeated for each additional TLD that a firm or consortium wishes to register.
    3. Existing registries operating in the alternative roots must be recognized. These organizations have put forth significant investments in this technology, and further, they are positioned right now, today, to provide diversity in the namespace. "First come, first served" is the guiding principle within the existing Copyright, Trademark, and Patent legal spaces in the United States today. We urge the Department of Commerce to follow existing legal precedent and insert into the root servers those operational registries which exist today, in the order which they are able to prove their operational status.

Transitioning of the IANA function

In the last few weeks we have seen a highly disturbing process take root – the IANA appears to be attempting to hand-pick yet another "transition group" – this time to manage its own transition. We are particularly unhappy with this set of events for the following reasons:

    1. The existing IANA organization should not pick its own successors. This is a basic principle of public policy. If you are dissatisfied with the course of action that an organization has taken, you do not allow them to choose their own jury in deciding what the next course of action should be! A more colloquial manner of saying this is "don’t leave the fox guarding the henhouse". We believe that this applies here, and that the Department of Commerce should be the organization that manages and chooses the IANA successors.
    2. It is critical that the IANA be operated completely within the legal limits of law in the United States, and that no outside organization be allowed to skew the decisions of this body by offering to indemnify it. Specifically, CORE has stated, several times, that they are considering indemnifying the IANA in order to induce it to move forward with the insertion of their TLDs into the root. We believe this is clearly improper when a quasi-governmental function, being performed under the auspices of the US Government, is involved. It is critical that the individuals involved in the IANA function be personally liable for violations of Anti-trust law in the event that they are significant participants in decisions which later turn out to be either civilly or criminally actionable. Our legal system is the last and best check and balance on an individual’s actions in a public role such as this, and we must not allow that remedy to be short-circuited.
    3. The IANA is the organization which is directly responsible for the current problem, in that they have allowed an effective monopoly to continue in domain registration. Further, the IANA is responsible for the IAHC and CORE, along with their current positions. Any reorganization of the IANA must take into account the past actions of those persons constituting the IANA, in particular the grant of monopoly registrar status within the .US domain and the IAHC/CORE process. We believe that the current members of the IANA should be excluded from future involvement due to what we perceive as mismanagement and active interference with the domain name industry in the United States and elsewhere.

IP Number Allocation Issues

The area of IP number allocation is one in which people have also, for a long time, been making false and misleading claims. Before we set forth a policy which we believe is in the public interest and which the Department of Commerce should force ARIN to adhere to, we would like to rebut some of the falsehoods that have been circulated:

    1. We are running out of IP numbers. False. In fact, we are nowhere near running out of IP numbers. We have wastefully allocated a great deal of IP space, and the first place to start IF we were determined to actually be getting close to running out would be to reclaim that which is most-wasted today. Examples of this include:

This means that more than 30% of the available space hasn’t even been assigned. If you go back to organizations which have "Class A" assignments, such as MIT, you would find that in many cases 90% - or more – of those addresses are not in use. This is an incredible waste and we could likely recover more than half of all possible IP space were we to perform a cleanup on the existing allocations and force those not in use to be returned for appropriate-sized blocks.

    1. We must be "careful" with initial allocations to ISPs, even those which are multihomed. False again. Multi-homed ISPs number perhaps 5,000 worldwide. Let’s multiply that by ten in the next few years – an incredible growth rate for multihomed connections, and one which nobody believes will be achieved, but let’s go with it nonetheless. Further, let’s assume that any multihomed ISP can receive a /19 block immediately, with further allocations once proof that the /19 is 75% exhausted has been provided. What would this require? The answer is you can obtain 8 /19s from a /16, and 256 16s from an /8. This means that you would need 24 /8s – or approximately 1/3rd of the unallocated old "Class A" space – to give every conceivable multihomed ISP its own /19. Clearly we can do this without any risk of running out of space, now or in the future.
    2. It is not a big deal to require people to renumber. Absolutely false! The impact of calling your dedicated line customer base to inform them that they are required to renumber out of their existing address space while remaining your customer will be that many of those customers will cancel! Currently, only small providers are required to do this – MCI, Sprint, UUNET and other large providers have never, to the best of my knowledge, been told that they have to return a smaller allocation to receive a larger one. Yet this is commonplace when smaller ISPs go to the address registries to request allocation of new (or additional) space. This is a clearly anti-competitive requirement which has the effect of severely disadvantaging smaller providers in the marketplace.

To repair this damage, we urge the Department of Commerce to promulgate regulations which specify the following as minimum standards for behavior by any allocating authority with ties to the US Legal system, directly or indirectly:

    1. Any legitimately multi-homed provider must be able to receive an allocation of a /19 upon request from the allocating authority. It is clear that no technical damage will result from such a policy. Further, we urge the Department to enforce that once 75% of existing allocated space has been consumed, the allocating authorities be required to issue additional allocations without requiring that the original block be returned.
    2. Address allocation authorities must not be allowed to set fees that are in excess of documented direct costs for the service of performing an allocation. Allowing authorities to operate as a trade association, where purchased memberships grant voting rights, is acceptable. What is not acceptable is requiring membership to receive an allocation, or charging fees in excess of direct costs for performing an allocation.
    3. Existing allocation discrepancies which have disadvantaged providers should be investigated by the Department, with appropriate legal sanctions levied against those organizations and individuals who have refused or hampered legitimately multihomed providers from receiving needed provider-independent address space. We can’t fix everything in the past, but we certainly should attempt to set an example that anti-competitive behavior will not be tolerated.

We hope that the US Department of Commerce will see fit to evaluate the postulates, axioms, and paths set forth in this response with the due process and investigation that we believe this matter deserves.

Sincerely yours,





Karl S. Denninger

President, MCSNet


From: Ellen Rony <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 1:58pm
Subject: Comments of Ellen Rony and Peter Rony

Before the
National Telecommunications and Information Administration
Washington, DC 20230

In the Matter of )

Notice of Proposed Rulemaking

Comments of Ellen Rony and Peter Rony, authors of The Domain Name Handbook:
High Stakes and Strategies in Cyberspace

Ellen Rony Peter Rony
21 Juno Road 1501 Highland Circle
Tiburon, CA 94920 Blacksburg, VA 24060

23 March 1998

1. Ellen Rony and Peter Rony respectfully submit comments in this
proceeding published at 63 F.R. 8825 (No. 34, Feb. 20, 1998). We are
siblings and authors of a forthcoming book, The Domain Name Handbook: High
Stakes and Strategies in Cyberspace (R&D Books), which culminates two years
of research on DNS issues.

2. Peter Rony is a Professor of Chemical Engineering at Virginia
Polytechnic Institute and State University in Blacksburg, Virginia, where
he has taught since 1972. He is a pioneering author on the subject of
microcomputer interfacing and the author/originator of the popular series
of Bugbooks written in the 1970s and 1980s. Dr. Rony is a trustee of
Computer Aids for Chemical Engineering Education (CACHE). He is editor of
CAST Communications, a divisional newsletter of the American Institute of
Chemical Engineers. He has authored numerous scientific articles and
taught short courses worldwide. Dr. Rony earned a Bachelor of Science
degree from the California Institute of Technology and a Ph.D. in Chemical
Engineering from the University of California at Berkeley. His response to
the Department of Commerce Notice of Inquiry is comment number 137,
submitted on July 24, 1997.

3. Ellen Rony has worked in and written about the computer industry for 17
years. In "Discovering the Source," published in 1982 by PC Magazine, she
described the early marriage of technology, information and communication
in a predecessor to the Internet. She is editor of Cyberworld, the
technology section of FastForward magazine in Marin County, California, and
locally operates the Alexander Works, specializing in high tech marketing
and web site design and maintenance. She holds a Bachelor of Arts degree
from Pomona College and a Master of Arts in Communication from Stanford
University. Her response to the Department of Commerce Notice of Inquiry is
comment number 443, submitted on August 18, 1997,.


We thank you for the opportunity to comment publicly on this proceeding.
We believe it is appropriate for the US Government to put forth a proposal
for the Technical Administration of Domain Names and Addresses. Today's
global internetwork is the result of the pioneering ARPA network developed
in the late 1960s by the U.S. Department of Defense to create a
communications link for scientists and research contractors. We feel that
the U.S. government has a fiduciary responsibility to its taxpayers to
develop a means for an orderly and stable transition of the DNS to the
private sector .

The Competitive Functions

We concur that the functions of a registry should be distinguished and
separated from those of a registrar. However, we recommend that no company
should serve simultaneously in both capacities. We feel registries should
operate on a not-for-profit basis and do not trust "market mechanisms" to
prevent the concerns over switching costs and lock-in that are cited in
this proposal.

This proposal confers to Network Solutions, Inc. a continued exclusive
commercial monopoly advantage over the very lucrative .COM domain. It will
be able to provide registry services without any oversight or regulation.
We are disturbed that this proposal enhances NSI's dominant position
further by granting a triple bonanza: exclusive back-end control over .NET
and .ORG as well, while restricting all other registries to a single TLD.

With 1.5+ million existing .COM registrants, we expect that this TLD will
remain the virtual real estate of preference and that NSI will remain the
preferred legacy provider. The triple monopoly conferred upon NSI in this
proposal, plus the opportunity to serve both as registry and registrar for
the three most popular TLDs, seems misguided.

The Creation of New gTLDs

We question the wisdom of introducing new gTLDs into Internet name space.
We need to decide what is broken, and what needs fixing. In our opinion,
the existence of 'international' top-level organization-type domains
(iTLDs) is what is broken.

The creation of new gTLDs exposes this proceeding to international
reprimand. We feel that the only specialized non-geographic uses such as
.INTA deserve the hallowed gTLD designation and should operate under U.N.
oversight. The argument that businesses want an "international"
designation, thus seek a gTLD such as .COM, is specious. Every presence on
the Internet is indeed international. What makes one TLD more desirable
than another is the ease with which Internet users can find a specific
site. We feel the expansion of gTLD namespace will exacerbate the problems
of accessibility and accountability.

There is no consensus on what number of gTLDs is best for the Internet
marketplace, on what they should be or how they should be chosen. The
numbers 5, 7, 2,048 and up to 10,000 gTLDs have all been suggested.
Technical matters notwithstanding, we think you should anticipate a
scramble for names in ALL TLDs that are newly created, making the Internet
name space flatter and more cumbersome and the effort required to find a
specific site more elusive. We reiterate our desire to see domain names as
country-specific with better categorization of second level domains along
the approach followed by Australia, England and France, e.g., .COM.US,

The Trademark Dilemma

Domain names are NOT trademarks. They are mnemonic addresses to replace
the underlying dotted octet that humans find too difficult to remember.
Unlike trademarks, domain names are neither context sensitive nor
geographically specific. Thus, we disagree with the focus this proposal
places on trademarks.

We do not believe this proposal strikes a balance between the rights that
trademark owners hold in the physical world and those they seek in the new
virtual realm. This approach does nothing to resolve the issues of
trademark owners of identical marks, of those with superior common law
rights or those who have another legitimate nexus to a domain name. Whose
right prevails when they all want the same name?

The borderless realm of electronic commerce raises complex issues for the
tenets of trademark law and the protection of intellectual property rights.
Nations have been working on the harmonization of intellectual property
protection and registration since the Paris Convention of 1883. The
Internet is growing too rapidly to expect the development of some mechanism
for "clearing" trademarks in the short term. However, individual countries
have statutes and mechanisms to define and adjudicate true trademark
infringement. It is preferable to eliminate international trademark
conflicts from the DNS by focusing on two-letter country codes as the top
domains. A migration of all TLDs to appropriate country codes would move
make most trademark squabbles internal to a country and avoid the
involvement of an international tribune or committee.


If the U.S government insists that a company be given exclusive
administration over a TLD, then NSI, must not be allowed continued monopoly
control over THREE TLDs. This anti-competitive monopoly position has no
justification in the current DNS environment.

The .US Domain

The geographic specificity currently required by the .US domain has
rendered it ineffective for nationwide commercial use. There is much
opportunity for enhancing the US domain space without displacing the
current geopolitical structure. This proceeding should set forth ways to
begin the transition of all .COM, .NET and .ORG registrations to
appropriate country codes. We urge the U.S. government to explore the
assignment of SLD categories under .US and begin the process of migrating
from .COM to .COM.US. , .ORG to .ORG.US, and identifying other suitable
categories of SLDs.

Most sincerely,

Ellen Rony Peter Rony
Tiburon, California Blacksburg, Virginia

Vincit qui patitur

Available soon: E. Rony and P. Rony, The Domain Name Handbook: High Stakes
and Strategies in Cyberspace
Ellen Rony
Director, Alexander Works
21 Juno Road * Tiburon, CA * 94920

Phone: 415/435-5010
Fax: 415/435-5010


From: Dennis Nelson <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 2:26pm
Subject: Technical Management of Internet Names and Addresses

For the proposal to be properly reviewed and digested by individual
Internet users, it should be structured as follows:

-- First and foremost, it should list how the proposal would better
guarantee universal access and the maintenance and promotion of
individual rights.

-- Expected end-state of the Internet via the new domain structure and
process. What will it look like, and how is universal access and benefits

- What are the benefits to individuals, benefits to commerce, to
institutions, to the people's representatives?

- What are the risks in keeping the domain system as it is; risks in
changing the system as proposed?

-- Then, what are the technical aspects of the proposed changes. What
are the other alternatives for which a cost benefit analysis has been
done which clearly shows the proposal to be the better return on

-- What will be the costs and benefits of each proposal, tangible and

-- What would an individual have to do to amend the proposal before its

The suggestion aren't meant to be all inclusive. The intent is to ensure
the focus is on the individual versus business or other exclusive rights
and on worldwide sharing of Internet benefits through universal access
and openness rather than controlling interests or just financial gains.

Dennis R. Nelson
Strategic Planner


From: Brian Harvey <bh@CS.Berkeley.EDU>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 2:32pm
Subject: domain name proposal

1. The proposal is seriously misleading in its claim to support
democratic, even-handed representation of interested parties.
The even-handedness is to be achieved by offering representation to
these categories: (1) address space gurus, (2) technical experts,
(3) registries, (4) registrars, (5) users. These five groups have
roughly equal weight. And yet, categories (3) and (4) together have
exactly one member, a single corporation, which category (5) has
some tens of millions of members.

And even within category (5), five of the seven seats are effectively
reserved for the wealthiest one percent of the population, with only
two seats for the rest of us.

Categories (3) and (4) should not be represented in this proposed
organization at all. This would be like asking AT&T and MCI to
select the FCC commissioners. The whole purpose of the organization
is to supervise (3) and (4). Groups (3) and (4) exist to meet the
needs of the Internet, not the other way around.

2. Expanding the set of top-level domains is the wrong solution to
a legitimate problem. The legitimate problem is that a private,
for-profit corporation holds a monopoly on Internet namespace.
But making it harder to figure out a host name is not the answer!
Is anyone, including businesses but especially individuals, going
to be well-served if is one company and is
a different company?

The proposed solution is to add yet another level of indirection, so
that we just look up "ford" in some index and find out the domain
name. But if we can do that, why not just look up "ford" and find
the host address? You are creating a need for another (costly, no
doubt) intermediary -- just what the domain system was supposed
to eliminate.

This is a knee-jerk response: The solution to all problems is to
introduce competition. That *can* be beneficial, but each case
must be considered on its merits, and *this* case doesn't have
any merits.

3. Finally, the overall tone of the report definitely makes the
Internet revolve around commerce as its guiding purpose. Commerce
is okay; I've bought things over the Internet myself. But much
more important to me is the exchange of ideas via email and
newsgroups, with private individuals. That must remain the
overriding purpose of the Internet.