1 (5)
Number: 405
|
|
|
1 (5) |
| Contact, telephone | Date | Reference | S-kl |
|
|
Patrice Washington Office of Public Affairs National Telecommunications and Information Administration (NTIA) Room 4898 14th St. and Constitution Ave. NW, Washington, DC 20230 |
Request for Comments on the Registration and Administration of Internet Domain Names
A. Appropriate Principles
It is our view that the presented principles in general are adequate for evaluation of proposals for the
registration and administration of Internet domain names. It is important to realize some basic
concepts that have governed Internet at large and should specifically govern the Internet domain
name system:
B. General/Organizational Framework Issues
The current domain name system was installed at the beginning of the explosive growth of the
Internet and was adequate at that time i.e. in the beginning of the ´90:s. Today with the enormous
expansion on a global scale a different domain name system, taking into consideration competition
and decentralization as well as international policy oversight, would be more appropriate.
The kind of authorities that would have both a dedicated Internet focus and being global does not
formally exist today. However, looking to other telecommunication sectors there are organization
established to take care of such issues on an international scale, e.g. ITU. It is our belief that in the
long run organizations like ITU will play a central role in the creation of fair and competitive
frameworks for the Internet.
The current telephone number administration may give at least to some extent an indication in what
way the administration of Internet domain names may evolve. Although there does not exist
international telephone prefixes comparable with the gTLD's, the national prefixes are being
administered by an international treaty organization and a similar system should, in the long run, be
adopted for the Internet domain name system. This is valid on an international scale. On a national
scale only relevant national authorities can have a say and should not be focus of international
agreements as long as there are not a conflict between the schemes adopted.
It would probably by economically a disaster to try to retire the current gTLD's (.com, .org and .net)
as there are significant resources invested in the registration and use of these domains. Taking the
global nature of the Internet into account there is also a need for international, generic, domains.
Many corporation and organizations are international and will benefit from not having to register
their name in each and every country, where they are active, under a specific country code if not
necessary for other reasons.
With respect to the technical frameworks with multiple registrars and multiple gTLD's it will be
necessary to provide databases with correct information independent of where from the information
is requested. There is thus a need for a global decentralized database system with corresponding
robust and secure real time transactions to keep the system synchronously updated. This may also
effect the update processes of the domain name root servers as well as interactions between
customers requesting database information (comparable to today's "whois"-service).
As comparable with the Ipv6 framework it is not advisable with a flag day, i.e. a single instance in
time where a change from current system to a new system is done. On the contrary, it will be
necessary to run both systems in parallel for some time, phasing in the new and phasing out the old
system. It is our belief a suitable point in time for opening up competition within the current gTLD's
(.com, .org and .net) would be at the termination of the NSF contract with Network Solution Inc, 1
of April 1998. Before that the administration of the new gTLD's must have proven to work and
shown its capability of acting as administration of the new gTLD's.
C. Creation of New gTLD's
The number of gTLD's should be kept small. The reason for this is that many international companies
will register their name in all possible gTLD's to protect their names. If thousands of new gTLD's
will be created this will cause sever costs for such companies (but on the other hand be a reasonable
business case for many new companies having name protection as there business idea). Another
reason is the today rather simple way of identifying a company on the Internet. In most cases
www.<company-name>.com will do. With thousands of new gTLD's this would not be that simple
any more. From a technical point of view, however, there are no such limitations on the amount of
gTLD's that could be included in the domain name system.
In essence, the administration and management of gTLD's should be kept apart from corresponding
administration and management of the ISO country code domains. On an international scale a system
should be adopted which not necessarily will be similar to the system used within each and any
country as long as these systems inter-operate on a technical level. While decentralization and
competition may be feasible on a global scale, a more centralized approach might be more adequate
in a small country like Sweden.
In conclusion, the number of gTLD's should be kept low while the number of organizations
providing registration services under these domains should be kept sufficiently high that a
competition is effectively achieved.
D. Policies for Registries
To make it easy for customers to change from one organization to another providing registration
within a certain gTLD there should not be any exclusive control over gTLD's. The amount of
investments lost if an organization registering domain names with exclusive control over a gTLD
goes out of business or for some other reason stops this service might be significant. If, on the other
hand, multiple registries are registering domain names for the same gTLD using one common
database, it will be much simpler to change provider and still keep the investments done in registered
domain names. However, in principle both exclusive and non-exclusive gTLD's is capable of coexist.
To prevent from registries too easily going out of business and, by that, stopping their registration
services there should be at least some minimal requirements on such a registry company or
organization making it plausible it will be able of keeping the business going in a longer perspective,
i.e. that they have the financial power of implementing and running such a business. From a technical
perspective there is a need for understanding of the Internet domain name system and that the
registration service is provided over the Internet.
Depending on the technical solution for a distributed shared domain name database there are no
serious limitations on the number of registrars. There are probably a financial limitation in the sense
that there is a limited market for such a service. What may happen is that a rather large amount of
domain name registrars will initially start up but in the long run fewer of these will be profitable,
grow and continue their business, grow maybe via buying smaller not so profitable registrars. In
essence the same scenario that has seen evolved with the commercialization of the Internet will
probably also be seen in a deregulated Internet domain name service provision.
E. Trademark Issues
So far, domain name registration have been on a first come first served basis which have made it
possible for organizations to register a lot of brand- and company names and later reselling them
"back" to the company with that brand or name (a.k.a. hoarding of domain names). To prevent such
business it should be necessary with some kind of domain name acceptance period where the name to
be registered is announced and where a possible owner of the name will have a chance of proving its
ownership.
It is our belief that current praxis used for international trademarks and names also should be used
within the Internet domain name registration system. Example of international entities that
accordingly should be involved are World Intellectual Property right Organization (WIPO) and
International Name and Trademark Association (INTA) as they have a well established system for
name and trademark conflict resolution.
To facilitate for name conflict resolution domain name databases must be easily accessible by
registrars and customers making it simple to check if a name have been previously registered within
the Internet domain name system. This functionality will be simplified with a limited number of
gTLD's.
Thomas Svensson
Vice President Telia Network Services
Head of Routernetworks/Internet