Document available at:

Before the
National Telecommunications and Information Administration
Washington, DC 20230

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

Comments of Gordon Irlam.


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

                                                  18 August 1997
Gordon Irlam respectfully submits comments in this proceeding. Gordon Irlam is employed as a software developer. He is a domain name registrant and operator of a DNS server. He has been using the Internet for 12 years, participates in the IETF, is a member of the W3C Advisory Council, the Association for Computing Machinery, and the American Economic Association.

Table of Contents

A. Appropriate Principles

B. General/Organizational Framework Issues C. Creation of New gTLDs D. Policies for Registries
A. Appropriate Principles

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

The following additional principle is suggested:

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

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

Two implications of this principle are:

B. General/Organizational Framework Issues

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

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

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

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

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

2. How might current domain name systems be improved?

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

C. Creation of New gTLDs

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

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

11. Should additional gTLDs be created?

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

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

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

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

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

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

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

D. Policies for Registries

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

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

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

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

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

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

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

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

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

Exclusive domains should be phased out.

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

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