From: "Martens,Rena - LGA" <>
To: "'NTIA'" <>
Date: 3/23/98 4:18pm
Subject: Comments of AT&T Corp. (Microsoft Word for Windows '97)

CC: "Kleinman,Ava - LGA" <>

Before the


National Telecommunications and Information Administration

Washington, D.C. 20230


Improvement of Technical Management ) Docket No. 980212036-8036-01

of Internet Names and Addresses ) RIN 0660-AA11









Mark C. Rosenblum

Ava B. Kleinman

Its Attorneys

Room 3252J1

295 North Maple Avenue

Basking Ridge, New Jersey 07920

(908) 221-8312



March 23, 1998







A. Plan and Schedules 12

B. Comments on Appendix 1 15

C. Selection of Registries 18





Before the


National Telecommunications and Information Administration

Washington, D.C. 20230


Improvement of Technical Management ) Docket No. 980212036-8036-01

of Internet Names and Addresses ) RIN 0660-AA11



AT&T welcomes the opportunity to respond to A Proposal to Improve Technical Management of Internet Names and Addresses, Discussion Draft 01/30/98, published as a proposed rule of the Department of Commerce ("DOC") and the National Telecommunications and Information Administration ("NTIA") on February 20, 1998 (generally referred to as "the Green Paper").

As is discussed in more detail in Section II below, AT&T agrees with the overall plan for governance as proposed in the Green Paper and believes it provides a good framework for moving forward. However, it is critical that detailed and workable procedures be in place to assure both a successful transition to a new system for managing Internet names and addresses, and the viability of that ongoing system and structure of governance (addressed in the Green Paper in the context of a "New Corporation"). Accordingly, several areas need further elaboration and clarification beyond what is discussed in the Green Paper. Of greatest importance, the new system of governance needs to be based upon procedures, policies and processes -- instead of individuals -- to ensure continuity, sustainability and security. The charter of the New

Corporation must be carefully created and enforced to assure an effective and fair board structure, a clear set of responsibilities, and responsible overall management. In order to ensure that these important procedures and safeguards are put into place, the DOC should convene a Task Force, whose output would be available for public comment, comprised of a broad array of representatives from the Internet stakeholders (e.g., providers, commercial and consumer users, and technical communities). The Task Force would propose detailed recommendations to the U. S. Government for the governance of Internet Domain Names under the New Corporation, and for a transition to that new system of governance. One important consideration of the Task Force and the government is that global representation and participation in the board of the New Corporation is critical to ensuring that the Internet successfully evolves into a meaningful and viable medium for the 21st century and to ensure acceptance of the proposed transition plan by other governments.

It is also crucial to ensure that trademark rights not be compromised under any new transition plan for Domain Names and Internet Governance. AT&T appreciates the NTIA's recognition of the real and growing conflict between Domain Name registrations and prior trademark rights, and the serious negative consequences for consumers who are increasingly relying on the Internet as a commercial medium. In order to protect those rights, especially during the critical transition stage when new processes are being tested, all Domain Name Registrars must be required to follow a uniform Dispute Resolution Policy that incorporates widespread accepted doctrines of determining whether bona fide conflicts between Domain Name registrations and prior trademark rights exist and that employs strict and swift resolution of trademark disputes. AT&T further supports a long term work effort to achieve consistency in rules relating to trademark and Domain Name usage and to increase the use of technological solutions, such as directories, to lessen conflicts between trademarks and Domain Names. These issues are discussed in Section III below.

The Domain Name System ("DNS") and the ability to register DNS names are key to the operation of the Internet. As explained in Section IV below, AT&T supports NTIA's intention to move quickly towards a more formal structure to support these functions. However, any change offers opportunities for mistake and abuse, and it is vital to the Internet as a whole that the transition from the current registration system to the new environment be carefully planned and executed. Accordingly, the plans for the transition should be public, they must be tested before the actual transition takes place, and they must be implemented as soon as possible. The transition plans should address contingencies that might arise during the transition and ensure that small problems do not grow into major disasters. Government and industry should proceed quickly, but carefully, mindful that some extra planning and testing time up front frequently saves a great deal of time fixing things after a failure. Of great concern -- and warranting special attention -- given the nature of the Internet today, security must be an important consideration both during the transition and in the new environment.


AT&T agrees with the concept of coordinated governance, embracing regions of the globe as well as types of stakeholders (e.g., providers, commercial and consumer users, and technical communities). The Internet stakeholder community is evolving, just as the Internet is evolving into a commercial medium. As this community moves from an Internet that serves millions of users to an Internet that serves magnitudes more, the governance structure must be flexible enough to address the changing nature of the stakeholders and the increasingly global distribution of stakeholders. To meet this goal, AT&T concurs with the basic four areas of authority for the New Corporation.

It is critical, however, that the DOC/NTIA go beyond the general principles set forth in the Green Paper and establish concrete processes that will govern the overall responsibilities of the New Corporation and clarify the fiduciary responsibilities of its Board and Staff. To this end, AT&T strongly urges the DOC/NTIA to convene a Task Force, with broad representation (e.g., providers, commercial and consumer users, and technical communities), to draft a comprehensive set of rules under which the New Corporation will operate. In particular, the Task Force should address such crucial issues as developing the charter for the New Corporation; specific responsibilities of the Board; representation on the Board; removal of Board members; frequency of Board meetings; general staffing requirements; budget; funding; and transition planning. The New Corporation should be based and operated from documented processes and procedures that are developed from present operations of the IANA during the transition period, and provide the basis for ongoing operations. Clearly, a business model is a sound premise on which to base the structure of Internet governance (and the transition to a new governance plan).

The proposed timeframe for handing off full control from the U.S. Government ("USG") to the New Corporation seems reasonable; however the term "stable" needs to be defined other than a state achieved on September 30, 2000. AT&T suggests that more detailed criteria be defined and included in some form of a comment process, such as via the proposed security study, in order to ensure that the New Corporation is capable of supporting the full set of expectations before functions are fully transitioned to them. The criteria for determining when there exists a "stable" form of governance should include:

1. required levels of security have been met and thoroughly tested (see below);

2. personnel with defined expertise and experience are in place and fully trained;

3. contingency plans for replacement of staff and equipment are established and documented;

4. facilities are identified, existing and operational;

5. there exists an operating budget for 12 months along with a realistic plan for obtaining funding/budget for the next two years (including funding to cover potential litigation, replacement costs of employees and equipment, insurance for board members). How funding will be provided is a critical element. Ideally, initial funding should come from the Infrastructure Fund; however given the unclear legal status of that fund, back-up or replacement funding needs to be identified; furthermore, ongoing funding needs to be developed; and

6. policies have been shown to be workable and viable for the industry. One suggested method of ensuring stability is to authorize dual operations between IANA and the New Corporation systems for two months.

Contingency plans are needed in case these criteria (or the final criteria) are not met.

The NTIA must adopt firm rules in the area of security of the DNS. To guard against cache contamination incidents, AT&T urges the rapid deployment of the DNS Security Extensions (RFC 2065). AT&T notes, in this regard, that the cryptographic mechanisms involved are authentication-only, and hence pose no export or key recovery issues. The root key is of critical import to this entire scheme. Both technical mechanisms, such as split-key secret-sharing schemes and administrative mechanisms, should be employed to safeguard this key. No one country should hold a controlling number of key shares. Finally, all major DNS servers – the root servers and the servers for all gTLDS -- must be operated securely. This includes physical security, cryptographically protected access mechanisms, and personnel screening.

The Green Paper is clear that global representation is essential; further discussion is needed on how that representation is identified and recruited, and this is an appropriate subject for the Task Force to pursue. In order for the transition to private-sector governance to succeed, it must be truly and inherently global. As AT&T understands the Green Paper's intent, although the New Corporation would be based in the U.S., participation by entities located in other regions, and in other nations, is essential to ensure that the decision-making process is global.

AT&T agrees that having a stable legal authority is essential to ensure the success of the New Corporation, and locating it in the U.S. initially, where the USG has legitimate authority to undertake such transitions, is a reasonable first step. Once the transition is made successfully, the New Corporation's board may choose, if authorized under its charter, to have the flexibility and freedom to move some functions to other regions of the world. AT&T also recognizes that the continued role of the USG, even in transition, is controversial in some sectors of the Internet community. AT&T supports the goal of the USG to ensure global acceptance of, and global representation in, the New Corporation. Ultimately, stability and predictability are critical success factors to help ensure that the Internet achieves its full promise of a commercial medium accessible to all around the world.


AT&T agrees that additional gTLDs should be carefully added during the transition to private sector coordination of the Domain Name system and afterward, but only so long as valuable trademark rights of trademark owners are protected and consumers are not confused. Consumers rely on trademarks as source indicators that indicate quality and consistency. Consumer confidence and trust in the Internet will be enhanced if consumers can be assured that Domain Names which contain well-known trademarks do indeed link the consumer to the owner of that well-known mark.

The registration of bogus confusingly similar and dilutive marks is a significant and growing problem for all trademark owners and consumers. There are numerous accounts of consumer frustration and confusion when they type an address that contains a well-known trademark and access a different entity's site that is attempting to trade on the goodwill of the well-known trademark.

AT&T has experienced an exponential growth in Domain Name disputes with third parties during the past year. In recent months, these new cases have numbered on an average 15 to 20 per month. Virtually all of these cases have involved bad faith attempts to usurp AT&T's trademark rights or to sell such rights back to AT&T or others. Similar to other trademark infringements, alleged Domain Name infringements must be carefully reviewed and investigated, communications initiated with the alleged infringer, and often negotiations conducted with the third party, through attorneys.

Overall, AT&T finds both the current NSI and the proposed CORE procedures inadequate in protecting and addressing trademarks interests and concerns. For example, the current Network Solutions Dispute Policy only covers disputes that involve a trademark that is identical to the registered Domain Name. This is problematic in cases where a cybersquatter registers a Domain Name that slightly differs from a well-known mark but is still likely to cause confusion. In these cases, a trademark owner's only recourse is the courts. Considering the number of Domain Name disputes, this is a major concern to trademark owners who may easily find themselves with a court docket of 100 cases per year. Furthermore, both trademarks and Domain Names serve as source indicators to consumers. Accordingly, this dispute policy also has a major negative impact on consumers who are seeking a particular company's product or service and are linked to an entity that has absolutely no connection to the trademark contained in the Domain Name. The proposed POC/CORE approach seems to assume that new Registries might develop their own dispute resolution policies specific to their gTLDs, but that WIPO and other dispute resolution entities could act as the arbitrator in disputes. Based on AT&T’s review of the revised third substantive guidelines, AT&T has significant concerns about the protection afforded to trademarks under this approach, which have not been adequately addressed by the CORE Registries/Registrars. For example, AT&T understands that several CORE Registrars have pre-registered thousands of Domain Names. AT&T is concerned that many of these pre-registrations will be dilutive of and/or confusingly similar to prior valid, subsisting trademarks.

Another problem encountered in Domain Name disputes is that many infringers provide false identities and contact information in the InterNIC records. In order to address these real and growing concerns, AT&T supports requiring each Domain Name applicant to certify under oath that the contact information it provides in the Domain Name registration is true and that it does not know of any individual or entity with superior rights in the Domain Name it seeks to register. In addition, AT&T strongly endorses the recommendation that the registration search tools be updated to meet trademark owners' needs.

All too often, however, bona fide measures such as those described above, intended to promote compliance with trademark laws, prove to be illusory, because determined cybersquatters will use false information, swear false affidavits, and otherwise provide information that is intended to deceive and mislead.

Because of the easy circumvention of these rules and the associated costs, AT&T agrees that an efficient and necessary alternative to protracted Domain Name litigation is on-line dispute resolution. AT&T supports the view that all Registrars be required to resolve disputes within 90 days after an opposition is filed against a specific Domain Name. Registrar policies must be consistent and incorporate identical dispute resolution procedures that address this problem and provide redress to the trademark owner. The standard which should be applied in all dispute resolution proceedings to determine whether Domain Name registrations conflict with prior trademark rights is whether the Domain Name registration is "identical or confusingly similar to or dilutive of any demonstrable trademark right." This standard is well developed and adheres to principles already established in many countries throughout the world. Absent resolution in this time frame, however, the trademark owner is free to pursue its recognized legal remedies in its own country or in any country in which jurisdiction lies.

AT&T also strongly supports the right of the losing party to file a de novo action locally. In addition, AT&T recommends that a reasonably long limitations period (e.g., five years) be established by rule in which any party that believes it was being damaged could petition to cancel a Domain Name. This five year time period is consistent with established law throughout the world.

AT&T further recommends that strong measures to discourage cypersquatters and individuals incorporating famous marks in their Domain Names to attract attention to their sites be included in all policies relating to the Domain Name application process. And both Domain Name registrants and Registrars should face specific penalties, including the loss of Domain Name registrations, if false information or false or misleading certifications are made.

AT&T is also in favor of a 30 day waiting period for registering a Domain Name. The proposed Domain Name should be publicly published for opposition purposes during this period in an easily searchable database maintained by the Registry. This would provide trademark owners with a cost-effective opportunity to halt the use of infringing Domain Names before irreparable harm is done to their name and reputation. This process is also beneficial to Domain Name applicants, because it protects them against investing time and marketing dollars in a Domain Name that they may ultimately be precluded from using. Although it may be argued that this waiting period is inconsistent with the speedy operations of the Internet, such a limited waiting period will ultimately benefit consumers who will be able to find the product and service providers they are seeking on the Internet without the confusion that may otherwise occur, and the harm to legitimate trademark holders if potential infringements are not disclosed and resolved at the outset. Moreover, this is in fact consistent with virtually all of the trademark registration procedures in practice throughout the world.

Obtaining jurisdiction over the Domain Name applicant is problematic for trademark owners. To reduce the burdens of enforcing trademark rights, Domain Name applicants should be required to agree, as part of the Domain Name application, to submit to jurisdiction in the United States and perhaps a specified list of other countries that have well-developed judicial systems and trademark and unfair competition laws.

AT&T also supports a mechanism whereby only the owner of famous marks could register or block any third party Domain Name registrations for such famous marks or any confusingly similar or dilutive marks in all the gTLDs of all Registries. Famous marks would be evidenced by a variety of factors including widespread registration, consumer surveys, judicial recognition and other factors that are legally recognized to establish strong trademark rights. A universal list of such marks would be created and registrars would be required to preclude any attempt at registration of any of the marks on the list or any confusingly similar marks.

Appendix 2, which establishes minimum dispute resolution and other procedures related to trademarks, as drafted, is a good starting point in establishing minimum requirements to which all registries must adhere. However, AT&T recommends some changes. In particular, the trademark dispute resolution process should be identical for all registries, and trademark owner/industry working group input should be required in the formation of this process.

Finally, AT&T supports the undertaking of studies to evaluate the effects of adding new generic Top Level Domain Names and related dispute resolution procedures on trademark owners. Such studies should address a variety of issues, including but not limited to the following:

1. Determine how many and when new gTLDs should be introduced;

2. Decide whether the establishment of a trademark clearance process is necessary;

3. Determine whether a searchable database is needed, and how it might be funded and supported;

4. Identify what information from Domain Name databases should be accessible for "clearing" of trademarks;

5. Address known effective dispute resolution mechanisms and whether additional protections should be or could be established by the registries; and

6. Conduct an in-depth focus on choice of law and jurisdictional issues and trademark infringement liability for registrars, registries, and technical management bodies.

The study should be undertaken with participation of (1) individuals and organizations responsible for trademark protection; and (2) individuals and organizations concerned with the commercial growth and functioning of the Internet. AT&T welcomes the opportunity to participate in such studies.


A. Plan and Schedules

It is critical that public transition plans be developed to ensure a timely and orderly transition from the current system (NSI) to the new system (competitive registrars and multiple registries). Under the auspices of the DOC and NTIA, and in line with the Green Paper, a major component of the Internet’s infrastructure -- Domain Name registration and governance -- is undergoing substantial change. It is thus crucial that industry and government move forward in a way that helps reduce concerns about the stability of DNS and the registration system. One way to do this is to put a well-documented plan in place and to publicly document the progress of the plan.

Consistent with this goal, AT&T recommends the following specific action items to implement the transition:

1. It is important to quickly nail down key dates. One of the most important dates to establish is the date that registrars other than NSI will be allowed to register names in .COM. This key transition date should be as soon as possible, and preferably before the end of the NSI contract extension.

a. If the transition cannot be completed before the end of the extension, there should be a contractual agreement between NSI and the New Corporation covering transition details. Such a transitional contract should be of short duration (certainly less than 18 months), should be non-profit, should explicitly recognize the New Corporation’s oversight authority and that the New Corporation has title to the Registry data, and should state that the New Corporation handles appeals and can conduct audits. These criteria should apply to all Registries.

b. Other key dates include: (1) naming of the Task Force that writes the New Corporation charter, completion date for the formal charter, naming of the initial board of the New Corporation, and selection of the New Corporation CEO; (2) the date the New Corporation is formed; (3) selection of the first Registries other than NSI; (4) selection of the first Registrars other than NSI; (5) start of registration for the first new gTLD(s); and (6) the date the New Corporation takes over each of its four major functions (number block policy, oversee root servers, oversee gTLD policy and criteria, coordinate protocol parameters).

2. It is also important to understand what will happen if the key dates are not met. Solid, public plans, schedules, and progress reports are essential to stability and should be required from the New Corporation and from all potential Registries. The plans should include intermediate milestones so that progress can be tracked during the transition, and any contracts that affect key transition dates should include penalties for delays. Such penalties could be financial ($1,000/day late charge) or business oriented (if a Registry is not ready at the promised time, its gTLD(s) are given to another Registry).

3. Some specific contingencies that should be addressed include:

a. What happens if NSI is not ready to start shared operations for .COM, .ORG, and .NET on the agreed upon date? Financial penalties are most appropriate in this case.

b. What happens if a Registry cannot handle the load it receives once it goes into service? Registries should be encouraged to support each other in situations like this.

c. What happens if a Registry refuses to accept registrations from some Registrars? Since a Registry controls all changes to the gTLD(s) it controls, the Registry’s contract with the New Corporation should require it to accept registrations from any Registrar that meets minimal requirements set by the New Corporation. Financial penalties should apply if it does not, with the possibility of reassigning the Registry’s gTLDs to another Registry in specified cases.

d. How are disputes between registrars and registries handled? One way to cut down on conflicts might be to make Registry logs generally available so timestamps can be examined in case of disputes about who registered a name first. This would make it easy for Registrars to discover when the earlier request was submitted and could cut down on calls the registry needed to handle. It would not eliminate disputes where the Registrar did not trust the Registry’s logs, however, so the agreements between the New Corporation and each Registry and between Registries and their Registrars must include dispute resolution procedures.

e. Where may applicants for domain names send complaints about Registries and Registrars? The New Corporation will need to set up procedures to handle such complaints.

4. The Green Paper states that "the first five entities" to meet the criteria in the Appendix will be allowed to become Registries. When does the clock start to measure "the first five" (some groups will feel they already meet the criteria)? Who will decide who "the first five" are and determine that they do meet the criteria? How soon could the first new gTLD go into service? Without answers to these questions the current confusion over gTLDs can only increase.

Finally, the plan specified in the Green Paper does not set out specific benefits to the participants who cooperate or disadvantages to participants who do not cooperate. Why should NSI agree to this plan? What happens if they don’t? Why should CORE (which has a reasonable amount of support internationally and in the Internet community) cooperate rather than setting up a competing system?

B. Comments on Appendix 1

Appendix 1, which specifies recommended registry and registrar requirements, makes a good start at laying out requirements for Registries and Registrars. AT&T recognizes, however, several inconsistencies in the proposed rule, and recommends specific clarifications that should be written into the rules.

First, there is an inconsistency between the overview of Registrar requirements ("Registries ... may remove domain names from the registries if at a later time the registrar which registered them no longer meets the requirements for registrars.") and Registrar Requirement 2a, which indicates that a registrar must have a plan to ensure that its obligations are fulfilled even if the registrar goes out of business. If the intent (as in 2a) is to keep registrants whole in the event of a Registrar failure, Registries should not be permitted to remove names. Instead, there should be an orderly transition to new Registrars for customers of the failed Registrar where customers of the failed Registrar are notified by the Registry and given access to a list of other Registrars for the gTLD in which they are registered and be asked to renew with one of the other Registrars when their current registration expires. Because the failed Registrar’s customers will typically have different renewal dates, their current registrations will expire at different times and should not cause a "log jam" at the other Registries.

Further, in line with the Registry Requirements’ intent that Registries be robust, AT&T recommends that each Registry be required to have a disaster recovery plan and disaster recovery tests/drills at random intervals. There should be a period of test operation of any new Registry before it actually accepts registrations. Both the hardware/software systems and the operations plans should be tested during this period. The primary DNS server for the SLDs supported by the Registry should be considered part of the Registry and should be subject to the tests noted above.

There should also be a requirement on the length of time that records must be kept. This should be as long as the maximum period when a registration can be challenged.

In addition, the rules proposed in Appendix 1 do not address the significant security concerns that exist regarding the registries.


1. As noted in Section I above, all major DNS servers—the root servers, and the servers for all gTLDs—must be operated securely. This includes physical and operational security, cryptographically protected access mechanisms, and personnel screening.

2. Given the inherent redundancy in the DNS, it isn’t particularly important that most root DNS servers be highly available; if one should fail, the others would carry the load with no user-visible effects. However, changes to any zone can only be made easily to the primary server (i.e., for that zone; accordingly, the primary servers should all be high-availability machines, with appropriate hot spares. Furthermore, regular disaster recovery drills should be practiced in the event that one such server will be down for an extended interval.

3. Registry machines are equally vital, and should be held to the standards of primary zone servers.

For Registrars, the requirements of Appendix 1 may be too strict. Since the Registry actually holds all the key data for a registrant, the requirements on Registrars should be reduced. AT&T doubts that any current or proposed gTLD registrar (including NSI and the members of CORE) meets all the requirements of Appendix 1. In order to support a speedy transition, the qualifications for Registrars established by the gTLD-MOU group could be taken as a starting point in place of the Registrar requirements in Appendix 1 and the existing Registrars authorized by that process recognized by the transitional one, conditioned only upon immediate re-opening of the application and acceptance process. Once established, the New Corporation may supplement the gTLD-MOU criteria or establish different criteria for new gTLDs as needed.

Ideally, customer-to-registrar communications should be digitally signed. Among other things, this will prevent the equivalent of "slamming" by registrars. Customer public keys should be known to the registries, and used to validate any requests purportedly coming from that customer.

Although AT&T agrees that registrars should be operated securely, this should not be imposed as a requirement. Anyone can switch registrars at any time, so the free market should ensure proper behavior.

AT&T supports the idea of checking whether Registries and Registrars meet certain minimal criteria; however, it is not clear how the certification process will be funded. This is particularly true in the case of Registries, where the checking should include site visits to the primary and backup sites. AT&T recommends that Registry and Registrar certification be funded by non-refundable application fees.

C. Selection of Registries

There are two types of TLDs in use today. Country Code TLDs ("ccTLDs") are based upon ISO-3166 two-letter codes. Generic TLDs ("gTLDs") are not tied to any specific locale. There are clear differences in responsibilities such as registry designation between generic top level domains and country-code top level domains since it is generally recognized that countries should have control over use of their ccTLD.

AT&T proposes that all registrars approved by the New Corporation can make registrations in all gTLDs. The New Corporation, not the registries, has the final authority on determining registrars' qualifications. National governments should have control over the Registrars and Registry for their respective ccTLDs.

AT&T recommends addressing these two types of TLDs in the following manner:

1. ccTLDs: let the national government designate the Registry, Registrars, and the policies. The national government may choose to ask the New Corporation to manage the ccTLD, in which case the New Corporation can treat the Registry function as generic TLD functions and can accept national-based policies if they do not incur significant extra costs or duties on the New Corporation-designated Registry. For historical reasons, the "MIL" and "GOV" gTLDs should be considered ccTLDs for the United States;

2. gTLDs: the New Corporation will designate the Registry responsible for each gTLD, and should allow for the possibility that one Registry will be responsible for more than one gTLD. The designation process should be competitive with registries structured as not-for-profit organizations and distinct from any other structure in their organization/entity, with no equity interest in the databases based upon their operations and support. Contracts between the New Corporation and these Registries should include an appeals process, periodic review of the registry's OA&M and auditing procedures, testing of transition to another Registry, acceptance of registrations from any certified Registrar, and the ability to terminate the contract for non-compliance or any breach of contract or public trust. The Green Paper proposes that the New Corporation select Registries and that Registries then select the gTLD they want. This mechanism is too far removed from the Registrars who will actually have to sell names in the new gTLDs. AT&T suggests that consideration be given to the proposals of the POC in this area to see if their mechanism of letting the Registrars select the Registry could be applied. Regardless of the selection process, all designated Registries would be required to meet the New Corporation criteria before they are allowed to take registrations.


AT&T appreciates the opportunity to provide its views on these important issues, and respectfully requests that the DOC and NTIA take full account of these comments and suggestions as it moves forward in its efforts to establish a stable and secure environment for Internet governance.

Respectfully submitted,


By     /s/   Ava B. Kleinman

Mark C. Rosenblum

Ava B. Kleinman

Its Attorneys

Room 3252J1

295 North Maple Avenue

Basking Ridge, New Jersey 07920

(908) 221-8312



March 23, 1998

certificate of service

I, Rena Martens, do hereby certify that on this 23rd day of March, 1998, a copy of the foregoing "Comments of AT&T Corp." was served by electronic transmission at and hand delivery, on the parties listed below:

Karen Rose

Office of International Affairs

National Telecommunications and Information Administration

U.S. Department of Commerce, Room 4701

14th and Constitution Avenue, N.W.

Washington, D.C. 20230


    /s/   Rena Martens

Rena Martens


From: Jeff Kuester <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 4:28pm
Subject: Comments

Please find attached comments to the Improvement of Technical Management of
Internet Names and Addresses from the American Bar Association Section of
Intellectual Property. The comments are in Microsoft Word for Windows 95
Version 7.0a format. Please acknowledge receipt.

Response to the U.S. Government Paper on the Improvement of Technical Management of Internet Names and Addresses



The American Bar Association Section of Intellectual Property Law applauds the work of the U.S. Department of Commerce for taking the time to propose and request comments on the "Improvement of Technical Management of Internet Names and Addresses" [Docket No. 980212036-8036-01] as published in the February 20, 1998 issue of the Federal Register ("Proposal"). The Proposal constitutes a proposed rule of the Department of Commerce and solicits public comments for the purpose of issuing a final rule.

The following comments represent a draft report by the Federal Trademark Legislation Committee ("Committee") of the Intellectual Property Law Section of the American Bar Association. However, the following comments have not been submitted to nor approved by the Board of Governors or House of Delegates of the American Bar Association and, therefore, should not be construed as representing Association policy.

In light of the over 430 public comments, the Committee believes that the Proposal correctly recognizes in a section entitled "The Trademark Dilemma" that cyberspace will only function as an effective commercial market if businesses have confidence that their trademarks can be protected. It is also recognized by the Proposal that management of the Internet must respond to the needs of the Internet community as a whole, and not trademark owners exclusively. Thus, in suggesting that trademark holders should have the same rights on the Internet that they have in the physical world, the Proposal accurately notes that a fair balance should be struck between the interests of the Internet community as a whole and those of trademark owners.

a. Common Law and State Law Trademark Rights Should Not Be Ignored on the Internet

Although the Proposal explicitly requests comments on a few trademark-related questions, all of which are addressed below, the Proposal unfortunately does not address one of the most widely criticized problems with the current domain name dispute system ­ the problem known as reverse domain name hijacking. A domain name is the subject of a reverse hijacking when it is automatically suspended without legal justification. The current domain name dispute system allows for such unwarranted suspensions by both ignoring the common law and state trademark rights of many trademark owners while also impermissibly exaggerating the scope of protection provided by a federal trademark registration.

The current domain name dispute policy maintained by Network Solutions, Inc. (NSI) includes considering only one factor, that of a federal trademark registration, before suspending a domain name for an indefinite period of time. Unfortunately, such a policy often results in what effectively amounts to automatic permanent injunctions against domain name owners in situations where the facts would clearly not support such injunctions from a court of law in the physical world. As a result, the current domain name dispute system and that suggested by the Proposal do not reach the goal of ensuring that trademark holders have the same rights on the Internet that they have in the physical world. Instead, the current policy ignores common law and state trademark rights by allowing only a domain name owner having a federal trademark registration to prevent a challenged domain name from being automatically suspended. Furthermore, the owner of a common law trademark does not have the same ability under the policy and, according to the policy, that domain name owner must file a lawsuit against NSI or the complaining party to prevent the domain name from being placed on hold.

By determining whether particular domain names should be placed on hold and effectively enjoining further use of those domain names, NSI is usurping the role of the courts and reaching ill-founded conclusions based on the single factor of federal trademark registration. As a result, legitimate domain name owners with non-infringing uses have been forced to bring suit against NSI and others simply to maintain the status quo. According to the policy, filing such a law suit within 30 days of receiving notice from NSI is the only way for a legitimate domain owner to prevent the domain name from being suspended after a trademark owner with a federal registration in any class of goods or services complains to NSI. For example, the court in Interstellar Starship Servs., Ltd. v. Epix, Inc., Civil No. 97-107-FR, 1997 U.S. Dist. LEXIS 18873 (D. Or. Nov. 20, 1997), entered summary judgment in a declaratory judgment action brought by a domain name owner making a clearly non-infringing use in a very different area of goods and services. The NSI policy forced this matter into court even though the dispute might well have been resolved without the need for court intervention.

While the cybersquatter problem is certainly real, the current NSI suspension policy unfortunately only works most "effectively" against legitimate users depending on the continued availability of their domain names. Cybersquatters, on the other hand, typically have only trivial uses of the ransomed domain names at best, such as Dennis Toeppen's aerial views of Pana, Illinois at the domain name "" See Panavision Int'l, L.P. v. Toeppen, 945 F. Supp. 1296, 1300 (C.D. Cal. 1996). Thus, it is often a meaningless threat against cybersquatters to have NSI place such a domain name on hold, leaving legitimate domain name owners as the class of domain name owners most adversely affected by the NSI policy.

As indicated above, the current policy has been widely criticized. For example, the Internet Subcommittee of the International Trademark Association (INTA) produced a Proposed Domain Name Registry Policy (available at that criticized the current policy of NSI (referred to as InterNIC) as follows:

While this and predecessor dispute policies appear to be well-intentioned attempts to address the legitimate concerns of trademark owners, they are rife with problems, including the following:

A trademark registration does not confer trademark rights in the U.S. and certain other countries, but rather merely constitutes rebuttable evidence of those rights. In the U.S., substantive trademark rights arise from actual use as a trademark. Accordingly, the current policy will sometimes accord unjust advantage to the prior registrant over the prior user who may have superior rights. This is unfair.

Under trademark law, most trademark owners do not own the right to use the mark in question to the exclusion of all others -- many similar and identical marks coexist happily and peaceably in commerce because rights are very often limited by industry and by strength of the mark in that industry. Thus, the current policy may favor the rights of one trademark owner over the legitimate coexisting rights of another, solely because one registered and the other did not. This is unfair.

Putting an active domain name on hold is effectively a form of injunctive relief, granted by InterNIC after the registrant has invested in the challenged domain name. Thus, in some circumstances, the dispute policy will allow certain trademark owners to trump the legitimate rights of other trademark owners and obtain the equivalent of injunctive relief, without meeting the stringent standards for such relief required by law. This is unfair.

InterNIC and other NICs are fundamentally unsuited to play a judicial role, particularly since domain name / trademark disputes are highly contextual in nature and require a close examination of the facts and equities of each case in order to reach a fair result. No amount of tweaking the current dispute policy will remedy this.

InterNIC risks being pulled into court whenever they put an active domain name on hold -- this is a resource drain on the NIC, whose resources are better spent on NIC administrative activities, rather than funding litigation to defend the dispute policy.

In short, the current dispute policy does not serve the interests of all trademark owners, but only some. The policy can result in injustice and it can get the NIC dragged into court (often to prevent just such an injustice) because the NIC is playing a judicial role that it is neither authorized nor competent to play.

The current domain name dispute policy is apparently a result of NSI's own concerns of liability for contributory infringement which might arguably result from otherwise failing to suspend an allegedly infringing domain name after receiving notice of such alleged infringement. As the Proposal correctly states, it is important that infringers, rather than registrars, registries and technical management bodies, should be liable for infringement. Unfortunately, while the Proposal also correctly notes that U.S. trademark law imposes no general duty on a registrar to investigate the propriety of any given registration and that the law provides no basis for holding that a registrar's mere registration of a domain name, at the behest of an applicant with which it has an arm's-length relationship, should expose it to liability, the Proposal shortsightedly suggests that registries should simply expect to incur legal expenses in connection with trademark disputes as a cost of doing business.

Instead of simply giving registries over to increasing legal expenses and further clogging court systems across the country by perpetuating the current reverse domain name hijacking problem, the option of legislative activity should be thoroughly considered. For example, the solution to this problem may well be a statutory limitation on liability against trademark infringement and dilution for domain name registrars, registries and technical management bodies. Such limitations on liability could easily be modeled after those provided to innocent printers and publishers, 15 U.S.C. § 1114(2)(A), (B) (1994), limiting liability to injunctive relief. While the creation of such limitations on liability might alone be sufficient to result in the adoption of domain name dispute policies that are more balanced and equitable, it may also be necessary to condition the granting of such immunity on the adoption of such policies.

In any event, the Proposal specifies that the domain name dispute policies of registries must be "independently reviewed." It should be clarified that this independent review should include a test for compliance with the principle of giving trademark owners no greater rights in cyberspace than they have in physical space, and for elimination of the present reverse domain name hijacking problem.

b. Temporary Initial Suspensions

The Proposal seeks comment on a variety of trademark related questions. According to the Proposal, the first area of questioning involves a temporary initial suspension and includes the following:

Mechanisms that allow for on-line dispute resolution could provide an inexpensive and efficient alternative to litigation for resolving disputes between trademark owners and domain name registrants. A swift dispute resolution process could provide for the temporary suspension of a domain name registration if an adversely affected trademark holder objects within a short time, e.g. 30 days, of the initial registration. We seek comment on whether registries should be required to resolve disputes within a specified period of time after an opposition is filed, and if so, how long that period should be.

Clearly, there is great risk associated with automatically suspending a domain name based merely on the statements of any trademark holding party alleging an adverse effect. Unlike trademark registration oppositions, which rarely delay uses by registration applicants, the suspension of a domain name will unavoidably prevent use of that domain name. Thus, regardless of whether there would be any actual adverse effects on the complaining trademark holder, there is clearly great likelihood that the delay in domain name availability will adversely affect the domain name registrant. For new small online businesses waiting to open their cyberdoors, this delay could be devastating.

As discussed above, the granting of an automatic suspension is effectively a form of injunctive relief. Thus, an automatic suspension could effectively deny due process if a domain name registrant is not given an opportunity to respond to the allegations of the trademark holder before the domain name is suspended. If a temporary initial suspension is necessary, the Department of Commerce should consider a policy giving the domain name registrant such an opportunity to respond before the domain name is suspended.

Regarding whether registries should be required to resolve disputes within a specified period of time after an opposition is filed, the establishment of such a specified period of time is certainly warranted. Regarding the length of the period, the dispute resolution process should be exactly what the Proposal suggests ­ swift. Specifically speaking, a period of two weeks would seem to be adequate to render an administrative decision on this extremely time sensitive issue.

c. Jurisdiction, Clearing, and Minimum Procedures

The Proposal also seeks comments and suggestions on implementing a system whereby, at the time of registration, registrants could agree that, in the event of a trademark dispute involving the name registered, jurisdiction would lie where the registry is domiciled, where the registry database in maintained, or where the "A" root server is maintained. This system sounds promising since, as the Proposal suggests, trademark holders are often concerned that domain name registrants in faraway places may be able to infringe their rights with no convenient jurisdiction available in which the trademark owner could file suit to protect those rights. Regarding suggestions for implementing such a system, it would seem that all new domain name registrants should simply execute a domain name registration contract including contractual provisions which unequivocally create such an obligation.

Comment and suggestions are also sought for implementing mechanisms for "clearing" trademarks, especially famous marks, across a range of gTLDs. Such mechanisms would certainly provide great assistance to domain name owners and trademark owners. In addition, robust searching interfaces should be provided to accommodate slight variations in words and phrases.

The Proposal does not propose to establish a monolithic trademark dispute resolution process, suggesting that it is not clear what system would work best. While there will always be doubts as to what system might work best, it is clear that allowing each registry to even partially establish its own dispute resolution mechanism would be much worse. Unnecessary confusion would certainly result from such disparate treatment of the relevant issues.

d. Joint Submission by INTA, AIPLA & ABA IPL Section

The following suggestions were proposed as a joint submission by the International Trademark Association (INTA) and the American Intellectual Property Law Association (AIPLA) to Commerce. The Committee believes that they are well-taken and therefore endorses them as well.


A. Generic Top Level Domain Names (gTLDs)

1. There should be a "go-slow" approach with respect to the addition of new gTLDs, since they pose a policing problem for trademark owners.

B. Application and Renewal

1. The application must include:

* detailed identifying information, including applicant's name; state of incorporation or partnership (if applicable); address; telephone/facsimile numbers; and e-mail address.

* designation of an agent for service of process. The application should include the agent's name, address, telephone and facsimile numbers and e-mail address.

* statement of bona fide intent to use the domain name within a time certain.

* statement of bona fide reason for using the domain name, e.g. the domain name is the applicant's name or trademark or a variation thereof.

* statement of intended use of the domain name, e.g. as e-mail, website, BBS, or other use, and the goods/services in connection with which the applicant intends to use the domain name.

* statement that the applicant, to the best of its knowledge, believes that the domain name is available and that the intended use will not infringe the trademark or other rights of any other party.

2. Renewal

* Renewal of a domain name should be required annually.

* Renewal applications must confirm and update information contained in the original application.

C. Publication

1. Applications, including all the information included therein, should be posted within a week of receipt by the registrar on a publicly-available, searchable website.

2. All registrars should use the same application and should post the application on one website shared by all registrars. This site should, in essence, function as a centralized listing of directories from individual registries.

3. The website should be easily searchable across all registries. This website also should be updated as often as possible, preferably at least every 30 days by each registrar.

4. The registration database should be monitored for "deadwood", i.e., domain names which are not actually in use, and an appropriate removal policy implemented.

D. Dispute Policy

1. Neither the registrars nor the registries should be responsible for developing dispute resolution policies. There should be a single dispute policy for all registries and registrars. The proposed non-profit corporation should seek input from the trademark community in developing the uniform dispute resolution system.

2. Alternative dispute resolution mechanisms, such as mediation and arbitration, should be available. Where possible, these procedures should be conducted on-line and should be expedited.

3. There should be an administrative dispute policy ("Registry Dispute Policy") that, at least during the experimental period, is limited to cybersquatting, false application/renewal statements and other clear instances of abuse.

4. The trademark community must be included in developing an appropriate registry dispute policy.

5. Registries and registrars should be located in countries which are signatories to the Paris Convention (an international agreement providing that member nations shall give foreign nationals the same treatment as the member nations provide their own citizens) or TRIPs compliant by the year 2000.


It is vitally important that the trademark community have adequate representation on the Board. Organizations that are representative of the global trademark community, should sit on the Board, along with other voices from the business world.

If the Internet is to serve the needs of international commerce, including business' interests in maintaining brand equity and investment and consumers' interests in relying on such brand equity to identify the product, service or company of the consumers' choice, the business community must have a significant voice in Internet domain name policy.

As the Board is currently envisioned, it is weighted particularly heavily toward the Internet technical community, which may not fully understand or appreciate the relevance of trademark concerns to business and consumers. Because trademarks are the lifeblood of international commerce and the backbone of the worldwide web, the business and trademark community should comprise at least one-half of the representation on any governing body, in balance with the technical community.


From: Jane_Keeler <>
To: "''" <>
Date: 3/23/98 4:49pm

1) the internet is not only or primarily a place of commerce. To define it
as such would be to gut the possibilities out of it. It is a mode of
communication, which may include every sort of human endeavour, including
research, social work, social change, and counselling for example, some of
the areas I work in.

It has an enormous educational possibility as well as the possibility of
linking groups with common interests, whether they are environmental,
spiritual, or commercial. It is a mode of romance for some individuals,
and is corrupted by pornography by others.

It is essential that you NOT put into place a naming operation that focuses
on only commerce and business. This would be a grave error.

2) The internet should not be administered by representation. It is an open
forum, and must be kept open to individuals, not to associations only..In
the proposal at hand, there is the description of a board which would be
dominated by commercial/business users. This would produce a biased form of
governance that has little to do with the best sides of the internet.

I urge you to make changes that will not damage the current openness, and
non-commercial viability of the internet.

Jane Keeler
Human Rights Educator


From: Dan Steinberg <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 4:50pm
Subject: GP Comments

Before the
National Telecommunications and Information Administration
Washington, DC 20230

In the Matter of )

Notice of Proposed Rulemaking

Comments of Dan Steinberg

Dan Steinberg
Box 532, RR1
Chelsea, Quebec

23 Mar 1998

-------------------------------------- break



I. The Proposal
A. The Coordinated Functions
B. The Competitive Functions
C. The Creation of New gTLDs
D. The Trademark Dilemma (ref. App. 2)
E. The Intellectual Infrastructure Fund

II. The Transition
A. The NSI Agreement
B. Competitive Registries, Registrars, and
the Addition of New gTLDs (ref. App 1)
C. The Root Server System
D. The .us Domain
E. The Process


The GP is an excellent start towards privatization of the internet.
These comments are made with the intention of fine-tuning an already
adequate document, and should not be construed in any was a criticism of
the paper or the process. Most of the suggestions are made with respect
to the constitution and conduct of the new non-profit, and should not
raise any roadblocks to rapid implementation of the transition rules.

All of the issues raised in the Green Paper should be interpreted with
the concept of BALANCE firmly in hand. The interests of any group
should not be allowed to override the global interests of the internet
community as a whole.

It is fairly easy to recognize the need for action and why a certain
process was selected.
This does not make the process outlined for the transition period
optimal. It is strongly suggested that the rate of insertion is more
important than any actual number of new TLDs.

Trademark issues, while important, should not be the only overriding
concern when establishing policy.

It is not clear that there is a need to mix numbers and names in
governance. Unless there is an overriding policy or technical reason
for oversight of these functions to be placed in a single body, it would
appear optimal to keep them separate. This greatly simplifies the
constitution of the board that will run the new non-profit corporation.

-------------------------------------- break

Before the
National Telecommunications and Information Administration
Washington, DC 20230

In the Matter of )

Notice of Proposed Rulemaking

Comments of Dan Steinberg

1. Dan Steinberg respectfully submits comments in this proceeding
published at 63 F.R. 8825 (No. 34, Feb. 20, 1998).

Dan Steinberg has been involved with the internet from both legal and
technical standpoints for several years. His practice involves issues of
convergence between technology and law, and the current internet
situation is one of many instances where it has proven valuable to have
dual expertise. No claim is made that involvement with these issues
gives anyone any special rights. Rather, it is felt that all comments
should be judged on their individual merits. These comments are made as
an individual and a Canadian citizen, and not as a member of any
particular organization or group, or even ther representative of any

-------------------------------------- break

I. The Proposal

A. The Coordinated Functions

There is no denying the need to create a separate organization or
organizations to perform the existing IANA functions. However, it is
possible that the new corporation may be given an overbroad mandate.

The proposal calls for the establishment of a non-profit corporation to
have oversight on:
* IP address policy
* root server system
* domain names
* technical protocols

It is felt that this is a lot of disparate authority for one
organization. Allocation of number blocks, oversight on the operation
of the root server system, oversight of policy for determining the
insertion of new TLDS and coordination of the development of technical
protocol parameters appears to be a large mandate. Of the four,
oversight on the operation of the root server system and oversight on
policy regarding the insertion of new TLDs appear well-suited for a
single organization. The remaining two oversight functions could be
left to the internet to govern itself, as it has traditionally done.

The proposal states:

The new corporation will be funded by domain name
registries and
regional IP registries. Initially, current IANA staff will
move to this
new organization to provide continuity and expertise
throughout the
period of time it takes to establish the new corporation. The
corporation should hire a chief executive officer with a
background in
the corporate sector to bring a more rigorous management to
organization than was possible or necessary when the Internet
primarily a research medium. As these functions are now
performed in
the United States, the new corporation will be headquartered
in the
United States, and incorporated under U.S. law as a
corporation. It will, however, have and report to a board of
from around the world.

Under US law, there are many different forms of not-for-profit
corporations. Considerable uncertainty would be removed if the exact
nature of this not-for-profit corporation were to be established.

Board of Directors

If oversight of the IP number function is removed from the new
corporation, the representation can be made more balanced. No specific
changes to the constitution of the board are proposed at this time,
other than the removal of the representation of the IP number

The creation of this board is apparently itself dependent on the
creation of several membership organizations. Details on how such
membership organizations will be created are necessary in order to
ensure that all stakeholders are properly represented. This is meant as
a comment rather than a criticism. It is recognized that many details
had to be left out of the early implementations of the Green Paper.

One stakeholder area that does not appear to be adequately represented
is software vendors. Browser vendors such as Netscape and Microsoft
currently have certain features built-in that represent the overwhelming
preference of registrants for .com domains. This has both legal and
practical considerations. If the new TLDs are to thrive and the
registry for .com to avoid ongoing anti-trust suits, it will be
necessary to make modifications to browser software. It would be unfair
to force software vendors to make changes without giving them a voice or
some means of representation at the board. The browser example is only
one of many possible software changes that may become necessary to
implement ongoing policies changes of the new corporation. Thus, there
should be representation of software vendors and other stakeholders in
the new corporation.

B. The Competitive Functions

Discussion of competitive functions has tended to be rather emotional in
the past. It is important to recognize that these issues are emotional
and that no one is likely to be happy with any compromise

Although necessary for the transition process, it should be recognized
that the separation of the registrar/registry function is not a
technical requirement. NSI currently acts as both a registrar and
registry. Other organizations also act as registrars, and their
continued existence points to a market need for additional registrar
functions. Although the registry/registrar duality appears to have been
created for anti-trust considerations, I see a continuing need for
registrars even if anti-trust considerations become moot. There will
always be those that have preferences to "buy retail" or "buy direct",
and there are considerable value-added services that registrars can

One of the complaints about competitive registries is that they lead to
lock-in. "Lock-in" is perhaps not as important as some might think.
The internet survived many years of lock-in to NSI. It is possible to
have a very competitive market that still exhibits lock-in. The
computer industry prior to the open systems initiatives is a good
example. DEC, Data General, IBM and many others all competed for the
right to lock in customers to their proprietary systems. When this
became unsatisfactory, no government intervention was necessary to
effect a change. The marketplace demanded it.

It is also possible to have a strong competitive industry even if
consumers are locked-in after their initial choice. Any system with
components such as ski racks, stereo systems with advanced features,
even the old proprietary hardware systems from DEC, IBM, etc. requires
commitment to one product line. Complaints about these industries (if
any) arose when there was limited choice of vendors to lock into. No
one went after Microsoft until they had overwhelming presence in the
market. Thus, this is not an issue that requires government
intervention. The market is sufficient to handle this aspect of the
industry, even if some claim that lock-in will harm the consumer.

-------------------------------------- break

C. The Creation of New gTLDs

The actual number of players that are ready to implement new TLDs today
is not a large number. While there is concern that a large number of
TLDs could 'break' the internet (or seriously destabilize it), no one
has suggested that "large" is a number smaller than 1,000.

It is suggested that the number of transitional TLDs be increased to at
least 20, preferably 30. This will not satisfy everyone, but allow all
current stakeholders to have something for their efforts. In addition,
it will serve to diminish the possibility of subsequent lawsuits from
unsatisfied players.

Long-term, the absolute number of TLDs is less important than the rate
of insertion and the mechanism to verify that the net has not become
destabilized. This issue is best addressed by the new corporation, once
they have the results of the required research.

-------------------------------------- break

D. The Trademark Dilemma (ref. App. 2)

It is important that registries not be given the obligation of acting as
a court system. Only those who have been adequately trained in this
function should have the responsibility of such decisions. The numerous
disputes where NSI has been directly named as a defendant highlights the
fact that such organizations are ill-prepared to act as courts.

It is not clear what benefit accrues (if any) from the proposal to
suspend domain names if an objection is lodged within 30 days of the
name's registration.

This proposal harms both sides:

* domain registrants are once again subject to uncertainty under which
they have no control.

* trademark holders suddenly get a new obligation to check domain names
or possibly lose the chance to complain about an infringing mark.
Although the Green Paper does not automatically make the domain holder
free and clear after the 30 days (congressional intervention would be
required to substantially modify trademark law), it might create a
presumption that the mark holder was not vigorously defending their mark
in the first place.

A registry should not be forced to suspend domain names if an objection
(baseless or not) is lodged within 30 days of that name's registration.
TO quote the Domain Name Rights Coalition:

"Doing so would prohibit timely personal, political and commercial
speech, and is essentially a 30 day waiting period that would be
particularly injurious to small business interests without righting any
serious wrong.

Further, the automatic suspension is contrary to trademark law because
it effectively grants an automatic injunction against the domain name
owner where the trademark owner has not proven any likelihood of
confusion or infringement."

This suspension could only be partially justifiable in the case of
famous marks, but famous mark protection is far from universally
accepted worldwide. What this provision serves to do is create new
uncertainty and new law of questionable extraterritorial impact.

From the Green Paper:
Trademark holders have expressed concern that domain name
registrants in faraway places may be able to infringe their
rights with
no convenient jurisdiction available in which the trademark
owner could
file suit to protect those rights. At the time of
registrants could agree that, in the event of a trademark
involving the name registered, jurisdiction would lie where
registry is domiciled, where the registry database in
maintained, or
where the ``A'' root server is maintained. We seek comment on
proposal, as well as suggestions for how such jurisdictional
could be implemented.

It is important to recognize (as the Green Paper has done) that the mere
existence of a domain name identical to a trademark (TM) does not
necessarily constitute infringement. In the past year, this has gone
from academic theory to several court decisions. It is also true that
TM rights could be infringed by domain holders in faraway
jurisdictions. The rights of TM holders have always been subject to
infringement in faraway jurisdictions without effective recourse. The
defect is more in a lack of uniform international trademark law than in
the domain name system. The domain name system simply enhances to
possibility of infringement by lowering the cost of doing business. It
is unfair to expect the domain name system to solve a problem that TM
practitioners have been attempting to solve for many years. The current
method solution has always been to seize infringing goods at the
border. This is admittedly not a perfect solution, but the best one
available before the advent of the internet and domain names.

Trademark holders have also called for the creation of
mechanism for ``clearing'' trademarks, especially famous
marks, across
a range of gTLDs. Such mechanisms could reduce trademark
associated with the addition of new gTLDs. Again, we seek
comment on
this proposal, and suggested mechanisms for trademark

The problem with this is that the notion of famous marks is far from
universal. Some jurisdictions have created common law notions of famous
marks via court cases. Others have flatly rejected extending the common
law (or civil law equivalent) to include the notion of famous marks.
There is just no practical way to implement trademark clearance and
maintain any semblance of an international internet. It bears repeating
that trademark law is far from universal. Some jurisdictions have laws
that have no equivalent in the US. For example, the UK concept of
trademark bullying has no equivalent in the US.

We stop short of proposals that could significantly limit
flexibility of the Internet, such as waiting periods or not
any new top-level domains.

This is a good idea. Waiting periods serve no useful purpose in that
they would create a new obligation on mark holders without fixing the
problem. If a waiting period is to be of use, it would have to be
implemented across all the various country code TLDs as well as the
generic TLDs.

Unless an acceptable minimum dispute policy could be
"first-come-first-served", it is unlikely that forcing registries to
implement policies is a good idea. Experience with the many iterations
of NSI's dispute policy have shown that the registry is not necessarily
the best place to resolve disputes or even to develop policy.

Registrars should not be forced to have a domain name dispute policy.
Domain name disputes, as with all other disputes involving trademark law
need to be settled by court proceedings. Trademark owners should have
no superior rights in cyberspace than they have in any other medium of


There is no problem with doing a study of the effects of new TLDs on
trademark issues, but it is unlikely that such studies could provide
useful data until many years in the future. Court cases (including
appeals) take several years. Until consistent judgements are handed
down in several jurisdictions, the actual effects will be difficult to

-------------------------------------- break


E. The Intellectual Infrastructure Fund

No comments on this section.

-------------------------------------- break

II. The Transition

A. The NSI Agreement

1. NSI will effectively separate and maintain a clear
between its current registry business and its current
business. NSI will continue to operate .com, .net and .org but
on a
fully shared-registry basis; it will shift operation of .edu
to a not-
for-profit entity. The registry will treat all registrars on a
nondiscriminatory basis and will price registry services
according to
an agreed upon formula for a period of time.

No comment submitted.

2. As part of the transition to a fully shared-registry
system, NSI
will develop (or license) and implement the technical
capability to
share the registration of its top-level domains with any
registrar so
that any registrar can register domain names there in as soon
possible, by a date certain to be agreed upon.

No comment submitted.

3. NSI will give the U.S. government a copy and
documentation of
all the data, software, and appropriate licenses to other
property generated under the cooperative agreement, for use by
the new
corporation for the benefit of the Internet.

This is appropriate under the terms of the cooperative agreement. This
data will be necessary for the development of global whois service.

4. NSI will turn over control of the ``A'' root server and
management of the root server system when instructed to do so
by the
U.S. government.

In order for control to be turned over, it is necessary to have tested
infrastructure in place. Rather than rush to implement such changes, it
is preferable to 'do it right'. It is suggested that this action be
explicitly deferred at least until the new corporation is established
and has a chance to work on developing policies and infrastructure.

5. NSI will agree to meet the requirements for registries
registrars set out in Appendix 1.]

No comment submitted.
-------------------------------------- break

B. Competitive Registries, Registrars, and
the Addition of New gTLDs (ref. App 1)

The concept of introducing new TLDs in advance of the establishment of
the new corporation is a good one. It will serve to meet some of the
pent-up market demand and reduce some of the current uncertainty. It is
recognized that the government is bound in part by limitations on what
can be done without legislation. In this spirit, it is recognized that
a limited number of TLDs that meet certain criteria is the best
short-term way to introduce new TLDs.

Irrespective of the results of the process to admit the first 5 entries,
there must be clear recognition of earlier work done in order to defuse
the potential for lawsuits. Even if some of the pioneer registries
don't have a perfect claim for their respective TLDs, it would take
court action (or at least legislation) to prove this fact. Both court
action and congressional intervention take time and there would be
considerable uncertainty in the interim.

However, this short-term solution should not be confused with the
long-term implementation. Long-term, the rate of insertion is more
important than absolute numbers (5, 10 100, etc). Development of
policies with regard to the rate of insertion should be left for the new

-------------------------------------- break

C. The Root Server System
IANA and the U.S. government, in cooperation with NSI, the
IAB, and
other relevant organizations will undertake a review of the
root server
system to recommend means to increase the security and
management of the system. The recommendations of the study
should be
implemented as part of the transition process to the new

It is always a good idea to study a problem before developing a
solution. However, the timing of the study might be sub-optimal. It is
important to recognize that the terms of reference are about to change.
For example, several new TLDs are about to gain entry into the root. It
is not known at this time what impact these new TLDs might have on
security or management issues. The changes might be sudden or gradual.
It will be extremely difficult to measure such a rapidly-moving target
and it is possible that the very act of measurement could impact the
results. This is analogous to Heisenberg's Uncertainty, a known
phenomenon in physics.

It is suggested that this study be deferred (with the possible exception
of the collection of baseline data) until after the new TLDs have been
operational for a while. This would suggest that the bulk of the work
be performed under contract to the new corporation. This will allow the
new corporation to best describe and implement any required changes to
either the security or management of the root server system.

-------------------------------------- break

D. The .us Domain
No comments on this paragraph. This matter is of concern for US
citizens only.

-------------------------------------- break

E. The Process

It is laudable to seek strong consensus before implementing any
policies. However, it should be recognized that consensus may be
difficult to recognize. In a meeting of 100 people, 5 people shouting
will drown out the other 95, even if the other 95 have reached

Thus, the continued oversight by the US government is welcome. It is
expected that some course corrections may be required.

As a Canadian, I might normally be concerned with the establishment of
the new corporation as a US entity. But in the circumstance, it is
better for me to have a US-constituted corporation than live with
ongoing uncertainty. Dealing with the large political and economic
presence of the US is a business reality. Pierre Elliot Trudeau, a
former Prime Minister of Canada made the following remark in 1969:

"Living next to you is in some ways like sleeping with an elephant: No
matter how friendly and even-tempered the beast, one is affected by
every twitch and grunt."

-- Pierre Elliot Trudeau speaking to the Press Club in Washington,
D.C. (March 25, 1969)

My position is that it is better to be in bed with an elephant than
sleeping outside in the cold. Any Canadian who wishes to be seen on the
world stage must recognize that the US represents a significant portion
of the world economy. To be subject to US jurisdictions in some areas
is nothing new. Canadians with .com, .net or .org domains are already
subject to the actions of a US corporation (NSI) and US courts.

It is understood that several non-US stakeholders have objected to the
constitution of the new corporation as a US non-profit. Should the US
government decide that these objections are valid, Canada, with its
strong traditions of both common law and civil law, would be an ideal
compromise location.

-------------------------------------- break

Annex 1

Open Root Server Confederation
Karl Auerbach
Canadian Association of Internet Providers (CAIP)


From: Nominalia <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 3:52pm
Subject: DNS proposed rule -Comments form FCR/Nominalia

To whom it may concern,

Attached you'll find the comments submitted on the matter of the proposed rule
on Technical management of the Internaet nmaes and addresses by the Fundacio'
Catalana per a la Recerca.

The comments are in Word format, version 6.0.01 form Macisntosh. A second mail
with the same comnets in ascii follows

Fundacio' Catalana per a la Recerca

Before the

National Telecommunications and Information Administration
Washington, DC 20230



In the Matter of: )

Improvement of Technical )

Management of Internet Names ) Docket No. 980212036-8036-01

and Addresses; )

Proposed Rule )


Comments of the Fundació Catalana per a la Recerca

(Catalan Research Foundation; FCR)

CORE Registrar #01 through its domain-name

registration department, Nominalia


Josep Antoni Plana
Fundació Catalana per a la Recerca
Passeig de Lluís Companys, 23
E - 08010 Barcelona (Catalonia; EU)

National Telecommunications and Information Administration
Washington, DC 20230

	In the Matter of			        )
	Improvement of Technical Management   ) Docket No.980212036-8036-01
    	of Internet Names and Addresses               )

 Comments of Fundació Catalana per a la Recerca ("FCR")

I. Introduction

The Fundació Catalana per a la Recerca (FCR; Catalan Research Foundation) is a non-profit institution whose aims consist in promoting scientific research with a wide view of science in the service of society. Among the many research centers, functions and infrastructures , created, managed, funded and/or supported by FCR there are the Catalan Supercomputation Center (CESCA) or the so-called "Scientific Ring" (Anella Científica) the high-speed telecommunications network rallying all Catalan Universities and Research Centers.

FCR's funding comes from both the public sector (the Catalan Government) and private-sector institutions and companies, from the financial, industrial and telecommunications sectors.

Consistently with its pioneering role in other Internet activities and services in Catalonia, the FCR has actively participated in the so-called "gTLD-MoU process", being one of the original signatories of the Generic Top Level Domains Memorandum of Understanding, and an active player within the Policy Advisory Body from its inception. FCR was also one of the first organizations to apply, be approved and become member of the Council of Registrars (CORE), though a newly created department, Nominalia.

In this capacity we would like expressly supporting the comments (and related suggestions) already presented by the three distinct organizations/bodies which constitute the structure of the gTLD-MoU process: the Policy Advisory Body (PAB); the Policy Oversight Committee (POC), and the Council of Registrars (CORE). In fact we are satisfied to agree one way or another with most of the comments already submitted, as it appears to be a widespread consensus on the perceived shortcomings of the GP. We feel obliged, nevertheless, to stress once again some basic issues and submit our own concerns in relation with the Proposed Rules for the Improvement of Technical Management of Internet Names and Addresses, best known as the Green Paper (GP).

II. Summary of Concerns

The GP puts forward as its declared goals to bring competition to the domain name registration business, while trying to disengage the USGov from its (funding) commitments with the Internet management functions. In doing so the GP identifies four basic principles that should guide the process: stability; competition; private, bottom-up coordination and representation. We can only applaud this program, and adhere to all of its elements. Unfortunately we don't believe that the detailed provisions set out in the GP, and even its basic assumptions and choices, are in line with none of these goals.

Our major concerns are the following:

A) USGov regulation or International private-sector bottom-up coordination?

While claiming Internet self-governance as a goal, the GP in fact provides for an unprecedented USGov micromanagement of the Internet. This is at odds with both USGov repeated previous statements and the widespread preferences of Internet users.

We don't want to discuss now and here whether the US Government has the legal authority, even under its own legal system, to regulate Internet administration issues. This would certainly be discussed before the relevant courts, if needed. Nor do we pretend to discuss if the issuing of the GP is in itself a violation of the US-EU joint statement on Electronic Commerce, as this will be undoubtedly be directly discussed by both Administrations.

The point we want to make here is much more modest: just how weak is the GP justification on this regard, and how dangerous are the consequences of such approach.

Stating that the Internet has grown out of US Government investments is a too partial point of view that not only overlooks the important contributions from other parts of the World, most specially Europe (both form its private and public sector) but also minimizes the substantial contributions from US-based private corporations. We are all indeed aware and grateful to the USGov for its generous funding of the Internet, but this should not lead us to unilateral micromanagement of the Net basic administrative functions by the USGov.

This attempt of what is perceived as excessive, unwarranted, unneeded regulation is in fact producing a visible countereffect: for the first time there is serious discussions in various Internet fora about the convenience to "export" Internet's most crucial management functions outside the US, and locate them in a more neutral country. We believe that this move is not at all necessary, at least now, but the discussions in reaction to the GP are highly symptomatic.

The GP carries a double hazard: excessive (unneeded and unenforceable) top-down governmental regulation where the only role that Governments could and should play is ensure that the self-regulatory private-sector driven processes takes place in a competitive market, and unilateral action by a single government in a field that is global by its very nature. We firmly believe that we shall let the Net take its own decisions through its processes right now, accelerating as much as possible the restructuring of IANA as a more stable institution and letting that institution, along with IAB; IETF; IESG and ISOC take the pertinent decisions.

We want to encourage the USGov to take its full responsibilities in the areas where its authority is undisputed, namely the contract with NSI and the .US TLD, and to refrain from regulating those issues that, by its very nature, can only be solved through global private-sector consensus, but never through (unilateral) governmental intervention.

B) The GP fails to promote a competitive environment for the domain-name registration system.

The GP only allows (provides) for a rachitic competition with NSI, the current monopolist in the gTLD market with weak registries and hard-to-monitor separation among NSI's functions as registry and registrar.

A recurrent concern with the GP is repeated inconsistencies. There is nevertheless a line that crosses it from the beginning to the end and with a consistent approach: and apparent need to protect NSI. The USGov should not sponsor any solution that does not amount to robust competition with viable players. Overprotecting the incumbent monopolist as a principle is a policy that should not resist any serious antitrust analysis.

On the same ground, the GP fails to acknowledge that, from the customer point of view, names do not compete perfectly. Only registrars do. And the GP proposal, reinforcing the power of registries, reduces registrars'' ability to pass the benefits of competition to its customers. The GP overemphasize competition from the supply side perspective, completely overlooking competition from the demand-side perspective. It fails to rightly identify the relevant product market, as names do not (perfectly) compete because they are not perfect substitutes, and sometimes no substitutes at all.

Also from an antitrust point of view the GP fails to identify the relevant market in a second way, as it misses the difference between two separate markets: the market for SLD registrations and the market for renewals. In the former, a weak competition might occur among TLDs representing names that are at least partial substitutes (a certain, but never very robust, competition might occur among the registries operating .com and .firm. It will almost NEVER be possible among the registries operating .shop and .org. But it is important to insist that in the market for renewals. there will simply be NO competition for most users, especially those more involved in Internee businesses.

The GP would lead to a world of parallel monopoly registries, weak registrars locked-n to registries and users treated as hostages.

It is a fact that a registry is somehow a monopoly. What is then important is to reduce its power and ability to harm consumers to the strict minimum. If we are to live with monopolies, let's have benevolent ones. This is what the gTLD-MoU has tried to achieve though its non-for-profit, shared registry system managing TLDs that are treated as public resources.

C) Having Multiple Registries in the gTLD Namespace is a Huge Policy Mistake.

So-called competing registries (in fact parallel monopolies) in the gTLD namespace are likely to screw up the Net. This amounts to granting exclusive (monopoly) rights to for-profit entities over public resources (generic names). It leads to multiple and incompatible dispute resolution mechanisms, thus exasperating the trademark problem. To multiple registration technologies and registry interfaces, thus making registrars' life a nightmare. To multiple and incompatible policies, making registrants lives more difficult. And for this system to work properly we should increase, not decrease, the level of government oversight/involvement, quite the contrary of what is the express goal of the GP. This governmental oversight and intervention is not only at odds with the proclaimed goals of the USGov, but is also unenforceable in practice. We have a perfect proof in the lack of intervention form the USGov when NSI, though its wholly-owned subsidiary WorldNIC has started to charge more than allowed by the so-called Cooperative Agreement to registrants.

In front of that solution, CORE solves most of the above problems, while providing with self-policing mechanisms in case it would lead to undesirable situations: it is a membership association open to all those wishing to operate the registrar business, not a separate for-profit entity with its own selfish (even if legitimate goals). It has an independent oversight committee (POC). As a whole, maximizes competition without increasing the risks to the Net stability.

We believe that most of the current confusion arises from a misrepresentation of the different functions involved in the domain name registration business. This confusion comes form the fact that NSI has been historically playing all the functions.

We all agree that it is at the registrar level where real competition happens, and where users need having a fair choice. Competition at the registrar level should therefore be encouraged and anyone fulfilling a minimum set of objective criteria should be able to participate.

We should all agree that there could be many (even if not unlimited) database operators. (this is what NSI was primarily intended to do; this is what Emergent does now for CORE). Multiple db operators increases the costs and perhaps the risks of technical consistence, but on the other hand may reduce the operational risks. Competition at the db operators level makes little sense, but the existence of multiple db operators serving different TLDs under a given Registry might be desirable under some circumstances.

But basic feature of a Registry is policy setting. Here is where for-profit companies should absolutely be ruled out. Here is where governmental intervention is impossible at the gTLD level. Here is where IANA interventions is and not be sufficient. And here is where competition makes no sense.

Finally, a Registry should fix an appeal channel, a mechanism for dispute resolution; an oversight function. And this function, as the registrar one, should be separated form the registry (that should be limited to direct policy setting and director or indirect db operations). In this regard we could discuss whether CORE/POC adequately handle this process. We could further discuss POC composition (being currently revised, increased and diversified, as a matter of fact). We could discuss whether the new IANA has a role to play in this function, and to what extent.

But we cannot discuss the fact that we should separate some of these functions, and that the gTLD-MoU is the proposal that better deals with these problems.

Here should lay the focus of our discussions. Pretending that the simple addition of multiple registries will solve any of the current problems is too naive. Accepting that all registries should be run as non-profit registries only goes half the way: if we want real competition registrars should be able to operate under all TLDs. If different TLDs are run by separate Registries with different technologies, interfaces, policies and criteria for admitting registrars (as the GP implies) competition will not occur. If in the other hand we end up with different registries managing different TLDs but all of them being non-for-profit; using a standardized interface, and technology, working with the same set of registrars (all of them), with the same dispute resolution mechanism for conflicts between domain names and intellectual property rights; the same policies.... Well, we'll end up with multiple clones of a single registry.

Taking as granted the need for multiple (unlimited) registrars, the real choice is between chaos and no increase of competition at the registry level on one hand and technical stability, policy coherence and, again, no increase of competition at the registry level. Because we all knew that Registries do not compete.....

If the USGov wishes to experiment with multiple registries (multiple parallel monopolies) with exclusive rights over generic names, we strongly encourage them to run such an experiment under the .US TLD, which is indeed of its only and undisputed competence and therefore will be able to exerce the level of control, oversight , intervention and mandatory changes that might deem necessary at any given moment. But please don't extend such futile experiments to the generic DNS, when we all know that they would bring very serious problems without solving any single current problem. And, what's worst, without the ability, or very little, for the USG or any other government to solve those problems afterwards.

D) Process and Consensus

Finally, and as to the process to be followed, the GP somehow proposes to re-do from scratch what we have been doing during the last year and a half. Pretending that the gTLD-MoU process has failed to reach consensus, it only provides with an alternative almost universally rejected.

Many commentators have already pointed out how unrealistic the GP is as to the selection process of the "five first registries". FCR wants to underline once more that there is a process that has gone through incredible, even if not perfect, public scrutiny in order to solve this issue.

Indeed, the gTLD-MoU process have been accused of failing to build consensus. And this is true in absolute terms: not everybody agrees with our proposal (which still would not prevent it form having consensus) and many people openly oppose it. But nobody will be able to point out any other alternative plan (or even any other Internet-related process so far) with such a wide and diversified support.

Please read again the list of signatories of the gTLD-MoU ( The crucial point is not that it has 200+ signatories, but the wide, and often contradictory range of interests they represents: ISPs, large, medium and small. CORE registrars but also trademark associations. ISOC chapters and user associations, but also telco and backbone operators. Non-for-profit Foundations, associations and institutes and law firms. Private sector and public sector. US-based, European-base, AP basic and truly int'l. orgs and corporations. Internet historical players and latecomers. People with an economic vested interest in this process, and lots of people with the sole interest of contributing to the best solution to the gTLDs issue.

Add to this the role played by those initiating and participating the IAHC/gTLD-MoU process: IANA, FNC; ISOC; IAB; INTA; ITU; WIPO...

In front of that everybody is well aware of the absolute incoherence of the interests represented by those opposing the gTLD-MoU, and therefore the radical lack of any possible "alternative consensus", as each and every group of opponents has only one interest to defend and is unable to present a whole design or solve the multiple issues at stake: those wanting hundreds of TLDs and that part of the TM community waning zero gTLDs; those claiming that the ACPs only serve big corps and those claiming that ACPs give too much weight to domain-name owners' rights; those claiming exclusive rights on generic words as TLDs and those opposing any "commercialization" or private-sector driven management of the DNS; those wanting competing registries and those trying to keep their current monopoly; those wanting more governmental representation within POC or IANA and those even opposing the presence of ITU and WIPO as non-voting observers; those claiming that any institution related to the Internet should be US-based and US-incorporated and those claiming that all Internet functions should be performed by int'l associations; ......

No, there is no alternative consensus. What's more the GP major contribution to the current debate will be (besides fixing the general consensus on the broad goals of increasing competition and stooping governmental intervention and funding) to show that the gTLD-MoU has in fact much more support that we all have realized. The simple reading of the comments received by NTIA cannot lead to any other conclusion.

III. Conclusion

FCR strongly encourages the US Government to take into account that most companies, organizations and individuals participating in this process have asked for the following points:

* Reassert IANA's authority over DNS issues

* Encourage and help reorganizing IANA as a more stable and representative entity.

* Refrain form regulating the gTLD namespace.

* Don't perpetuate NSI's monopoly over .COM, .ORG and .NET.

Respecfully submitted,

J.A. Plana




From: Deb Howard <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 5:01pm
Subject: ISP/C Comments on the Domain Name "Green Paper"

To the Department of Commerce:

The Internet Service Providers' Consortium (ISP/C) hereby submits formal
comments on 15 CFR Chapter XXIII, "Improvement of Technical Management of
Internet Names and Addresses; Proposed Rule", Docket No. 980212036-8036-01.

The attached commentary expands upon notes previously submitted by the ISP/C
based on our meeting with Mr. Ira Magaziner and Ms. Karen Rose of March 12,
1998. The attached commentary has had the benefit of input and discussion
among our members, and we look forward to presenting it to you. The
document is entitled "ispc-gp.doc" and is in Word 6.0 format.

A number of ISP/C member companies will be submitting commentary on their
own, as we have encouraged them to do based on their interests.

Thank you for your assistance.

Kind regards,
Deborah Howard for the ISP/C

CC: Karen Rose

March 23, 1998

Ms. Karen Rose
Office of International Affairs, National Telecommunications and Information
Administration (NTIA), Room 4701
U.S. Department of Commerce
14th and Constitution Avenue, N.W.
Washington, D.C. 20230
Via electronic mail to

In regard to:
15 CFR Chapter XXIII: Improvement of Technical Management of Internet Names and Addresses;
Proposed Rule, [[Page 8826]]
DEPARTMENT OF COMMERCE, National Telecommunications and Information Administration
15 CFR Chapter XXIII
[Docket No. 980212036-8036-01]
RIN 0660-AA11

Improvement of Technical Management of Internet Names and Addresses
AGENCY: National Telecommunications and Information Administration (NTIA), Commerce.
ACTION: Proposed rule; request for public comment.

This response to the above docket number is made on behalf of the Internet Service Providers' Consortium (ISP/C). These comments include remarks made at a meeting held Thursday March 12, 1998, in which representatives from the Internet Service Providers' Consortium (ISP/C) met with Senior White House official Ira Magaziner from the Office of Domestic Policy and Planning and Karen Rose, Department of Commerce, International Affairs, in regard to the domain names proposal ("Green Paper") by the United States Government. ISP/C Participants at this meeting included Charles T. Smith, Jr., President, Internet Service Providers' Consortium and Deborah A. Howard, Chair and Executive Director, Internet Service Providers' Consortium.

About the Internet Service Providers' Consortium

Formed in June of 1996, the ISP/C is an international, not-for-profit trade association composed of individuals and organizations that functions to implement cooperative services to assist ISPs. As of March 1998, the ISP/C has 165 members representing 42 states and 10 international countries.

On behalf of its members, the ISP/C provides a unified voice on legislative issues, vendor relationships and other business issues impacting the operation of an ISP. The ISP/C provides a range of services, including hardware, software and support discounts, to ISPs. The group also offers a forum for the maintenance of an open global market for ISPs.

Information about the ISP/C is available at with a European mirror site at ISP/C representatives who participated at the "Green Paper" meeting can be contacted as follows:

Charles T. Smith, Jr.
(770) 934-6033, ext. 2902

Deborah Howard
Chair of the Board and Executive Director
(310) 827-8413 or (310) 448-1680

ISP/C's Position Regarding the Green Paper

The ISP/C wishes to express our general support for the Green Paper's initiatives, specifically, Section V. Principles for a New System, and the openness of its process as well as its expressed intent to turn the important asset of the Internet root servers and the domain naming process to the private sector. We believe an open and fair organizational process with a broad coalition selected from representative Internet stake holders is the best method to ensure stability and an appropriate transition process to the domain name system. In terms of specific suggestions for enhancing the Green Paper's proposals, we have several areas of specific concern.

Section V.
Principles for a New System
A. Internet Stability

For the stability of the Internet, the ISP/C believes it of the utmost importance that the process of establishing and maintaining the domain name organization should be conducted in an open manner. It is particularly important that individuals involved in selecting or serving on the domain names board be of sufficient stature in the Internet community so as to ensure the process is above reproach by responsible parties.

Furthermore in terms of stability, control and maintenance of the root servers is an essential part of the stability equation. It is vitally important that additional measures be undertaken to ensure that another split in the root servers will not occur to undermine the perceived stability of the Internet naming system.

Section V.
Principles for a New System
C. Private, Bottom Up Coordination
Bottom up coordination is important to this process. We believe there are sufficient models among existing Internet governance structures to support the reality of this process.

For example, in the IETF, individual members might decide, "someone needs to do something like this, let me see if other people feel the same way." If they can find other people who feel the same way, they find the appropriate tree of the IETF and go to the Area Director (AD) for that section of the IETF. If the AD agrees that the proposal is something that should be worked on, and that it is not overlapping with the work that others are doing, the process allows them to form a working group, appointing their own working group chair.

The working group then writes papers, and when the working group chair believes that the papers are finished, they hand them off to the AD, who makes sure the ideas contained within are not in conflict with other tracks, that they are correctly formatted, cover all the bases, etc. The AD then submits it to the IESG who comb through the working group proposal, decide whether or not it should be a standard (usually championed by the AD), and then if accepted, the proposal is assigned a number.

This "bottoms-up" process allows the foment of ideas to occur with checks and balances appropriate to the stability of the governance structure. We endorse such a model.

VI. The Proposal

A. The Coordinated Functions

Board of Directors Constituency: Lack of ISP Representation

One direct concern is that major stake holders, such as ISPs, are not explicitly represented in the current constituency of the non-profit corporation slated to be charged with domain name policy. Although it has been suggested that ISPs are represented through the IP allocation entities of APNIC, ARIN and RIPE, we do not concur with this opinion. In this regard, we suggest that the ISP/C be allowed to appoint a representative for small to mid-size ISPs, and that our colleagues at CIX have a voice as the representative for larger ISPs.

Board of Directors Constituency: Internet Users Representation

One concern we have is that no one "Internet membership" entity, such as the ISOC, would be appropriate to have the power to appoint the seven general Internet slots on the Board of Directors of the domain name corporation. In further regard to ISOC, it is not clear to us why there are proposed to be 2 IAB representatives; we would recommend only one such representative is necessary to be appointed by ISOC. It is appropriate that ISOC as an international organization with interest in domain name issues should indeed have an appointment slot for the domain name organization. We believe, however, that it is important to have a broad diversity of stake holders represented by the appointment slots.

In terms of "general Internet community", we suggest this definition is quite broad and subject to widely different interpretations. We did point up the fact, however, that representation from companies and corporate entities should not be dominated solely by trademark holders as there are many businesses which do not fall into that category who nonetheless have a stake in domain name policy.

Among general Internet membership societies, one in particular which we believe to be worthy of consideration to participate in this process is CAUCE (Coalition Against Unsolicited Commercial Email). With over 10,000 worldwide members, CAUCE is a grassroots organization with strong leadership which could be entrusted to appoint a responsible representative to the domain name non-profit corporation.

Board of Directors Constituency: International Representation

Internationalization of the representatives for this body is an important topic. As an international organization strongly committed to principles of internationalism, the ISP/C has several points to make in this regard.

One criticism leveled by international sources at the process outlined in the Green Paper is the fact that it recommends a U.S.-based non-profit corporation for its governance structure. This is not a criticism shared by the ISP/C. Due to the openness of process and the strict laws governing non-profit corporations in the United States, the ISP/C believes that having the domain name body be U.S.-based is preferable to having this entity's jurisdiction under Swiss law as suggested the ISOC/IAHC/CORE proposal. It is our belief that Swiss law tends to offer more protection for the corporation and less public review and openness than corresponding U.S. law, and it is for this reason that we strongly support this aspect of the proposal endorsing a U.S.-based, non-profit organization.

In terms of international representation among explicitly designated entities, the ISP/C wishes to express our support for the IP allocation entities, APNIC, ARIN and RIPE, to participate as outlined. We further suggest that if additional IP number registries are created in other areas of the world, they should have equal, rather than rotational, representation to prevent in-fighting among those entities.

The ISP/C would also like to suggest that a council of international delegates be appointed from the Internet bodies governing the country-level TLDs in the major Internet-using countries. These countries should include Australia, Canada, France, Germany, Great Britain, Japan, and the United States. Such an appointment process would broaden the perspective of the domain name body to be less U.S.-centric and more global in nature. Such an expansion befits a world-wide resource which nonetheless has its roots in U.S. government initiative and U.S. infrastructure investment.

In addition, appointment of international representatives with experience in country-level TLDs will help ensure that such representatives are both 1) well-informed of Internet matters and 2) broadly representative of their respective countries. One possible model for this process would be the United Nations Security Council.

VI. The Proposal

C. The Creation of New gTLDs

In regard to gTLDs, we support a "slow growth" expansion of the available gTLDs as we believe it is somewhat artificial to limit this number to five. We also support consideration of expanding the .US TLD in the manner outlined in the Green Paper. We do, however, endorse the notion that any possible expansion in new gTLDs be sought only after formation and consolidation of the domain name non-profit corporation, allowing that corporation to determine the consensus "best path" to follow.

VI. The Proposal

D. The Trademark Dilemma

This section as written is one sided. It contains paragraph after paragraph about protecting the rights of trademark holders, but not once does it provide protection for domain name holders. It has been our experience that duly registered domain name holders need some level of protection from trademark holders.

For Internet Service Providers in particular, the uninterrupted operation of our domain names is imperative. The suspension of a domain name does not just impact one segment of our business. Even a temporary suspension of a domain name can terminate, irrecoverably, our entire business.

Therefore, domain name holders should be protected from frivolous, unwarranted, and non-infringing attacks by trademark holders. "Just because" a trademark holder makes a claim of infringement to a registrar/registry, does not on its face mean that such a claim is valid. Registrars or registries must not be permitted to summarily seize a domain name based on nothing more than a claim of infringement by a trademark holder. This has previously been the case, and it should not be permitted to continue.

As current NSI policy stands, a trademark owner can simply mail a proof of trademark and state that they felt that there was an infringement of their mark. The trademark holder does not have to state how any alleged infringement is occurring nor do they have to provide any proof to substantiate such a claim. This is not a fair and balanced system.

Currently in the U.S., there is no need for a business to apply for a trademark or service mark in order to conduct business. Yet in effect, the current domain name system requires this additional process in order to protect a domain name registration. If a company wishes to protect itself from having its domain name taken away, it must process a trademark or hire an attorney to protect itself.

It is important to note that a trademark in the U.S. and in many other countries is offered to protect the rights of the trademark holder to conduct business within a given industry. A trademark does not grant an exclusive right to claim exclusive use of that trademark in any and all industries. As this is the case, any proposal to "clear" a name within a number of registries far oversteps the bounds of reasonableness in striving to protect the rights of a trademark owner. Doing so would create a situation whereby an owner of a trademark would enjoy exclusive rights to a domain name in any and all industries. This stance is inconsistent with protections offered under U.S. trademark law. Any attempt to overly broaden trademark protection is anti-consumer, as well as anti-small business.

The complexities of trademark disputes and ownership rights increase exponentially when we begin to examine these issues in their proper context which, given the constituency of the Internet, is global in scope.

The mere registration of a domain name that is the same or substantially similar to a registered trade mark is not in itself evidence of a trademark violation. With regard to registries in the U.S., it is not incumbent upon the registry to determine if a possible trademark violation exists, or to suspend or turn over ownership of the domain name as a result of that determination. The above assertions are borne out in U.S. case law.

Furthermore, Appendix 2 of the "Green Paper" includes the following provision:

"If an objection to registration is raised within 30 days after registration of the domain name, a brief period of suspension during the pendency of the dispute will be provided by the registries."

If such a "first 30 days" suspension provision is implemented, it MUST provide for sufficient time to register, and transition to, an alternate domain name to be used during the suspension period. Any cost associated with registering this temporary domain name SHOULD be borne by the challenger.

If this provision is implemented, we posit that many new domain registrants will "voluntarily" wait 30 days before they put their new domains into production use. It is simply too big a risk to begin using a new domain if an individual or company knows that anyone can come along and challenge it in the first 30 days. This objection clause effectively imposes a 30 day waiting period while allowing the authors of the "Green Paper" to claim, "We stop short of proposals that could significantly limit the flexibility of the Internet, such as waiting periods or not allowing any new top-level domains."

It is our belief of the suggestions made within the Green Paper regarding jurisdiction in disputes, the most appropriate is to place jurisdiction within the registry's primary country of business. Two other suggestion were offered, both of which we feel have substantial shortcomings.

The first, to place jurisdiction where the registry database is maintained, does not take into consideration that given the global nature of the Internet and the need for robustness, registry databases may in fact be maintained in several countries, creating a difficult question of where jurisdiction lies.

The second, to place jurisdiction in the country where the "A" root server is located, is no appropriate. As there is only one widely recognized "A" root server at this time, jurisdiction over all domain disputes, no matter their country of origin or the country in which that registry conducts business, would reside within a single country. This would subject the entire world to the laws of a single country in the resolution of domain name disputes, a situation which we would do believe optimal nor appropriate.

By placing jurisdiction of disputes within the registry's primary country of business, there is far less question of jurisdiction. Furthermore, if consumers have sufficient concerns their rights will not be respected by a certain country or countries, they can choose to do business with a domain name registry in a country where they feel their rights will be appropriately protected.

VI. The Proposal

E. The Intellectual Infrastructure Fund

The Intellectual Infrastructure Fund is comprised of contributions from a wide variety of individuals, companies and organizations. Because these funds are an outgrowth of the world-wide growth in the Internet, it is not entirely appropriate that the U.S. Congress be granted fiduciary responsibility to allocate these funds. As the proposed new oversight organization will be representative of the worldwide interests of the Internet Industry as well as the Internet Community, we feel it most appropriate that the disbursement of grants for Internet technologies should be at the discretion of this new organization.

VII. The Transition

B. Competitive Registries, Registrars, and the Addition of new gTLDs

In terms of registries, the ISP/C has no particular objection to registries being non-profit or for-profit. We also have no particular objections to their being sole source registries as long as the terms of contract are revocable and renewable by the domain name oversight organization.

In terms of registrars, the ISP/C endorses the concept of competition among registrars as we support the expression of the free market, and we encourage multiple choices for consumers. The barriers for entry to become a registrar, however, should be much lower than for the registry entity itself, and the consumer market should be allowed to choose among registrars.

While we are an agreement with the majority of qualifications to be a registrar/registry as they follow common sense standards as to network design and operation, the ISP/C finds the requirement of 24 hour guards in particular to be costly and unwarranted. While it is certainly appropriate to require 24 hour security via an alarm company, staffing of 24 hour guards in two independent locations could prove cost prohibitive to many small businesses which may wish to provide registrar or registry functions and serve as a barrier to market, lessening the options in registrars/registries a consumer would otherwise have.

Given that in both the case of the registrar and the registry, both would be required to have two physically independent locations we feel that while desirable, there should be no requirement for 24 hour guards.


It is essential to proceed on this proposal with due speed, but only in such a manner which adequately represents all parties and treats them in a fair and equitable manner.

With few exceptions, the Green Paper provides the potential framework to offer open, representative, and balanced solutions to many of the difficult questions surrounding the issue of Internet governance and the removal of U.S. government involvement in these matters. The decisions made now have the potential of charting the course of the Internet and deciding the issue of Internet governance for the foreseeable future. It is important we choose fairly, carefully, and wisely.

On behalf of the Internet Service Providers' Consortium, we appreciate this opportunity to make this report of our remarks available to the public process. If there are any questions, please do not hesitate to contact us.

Submitted by:

Deborah A. Howard, ISP/C Chair of the Board and Executive Director


From: harry cole <>
To: NTIADC40.NTIAHQ40(dns)
Date: 3/23/98 4:28pm
Subject: Please Ensure Timely Implementation

March 23, 1998

To Whom It May Concern:

I am currently employing a web developer to design the web site for my
practice. The web site depends heavily on the new 7 Top Level
Domains, which are to be available by the new registrars. The CORE
group is responsible for these new domains, and the U.S. government
would be apt to follow their example. My web site developer suggested
I personally address you, as it could be that these domains are not
available for a long time.

I do not want a .com for my practice, as I am a family physician. I
have preregistered names in .info and .firm because they adequately
represent what I am attempting to provide with my web site.

Timely implementation of the new Top Level Domains is imperative.
Indeed, the U.S. government must realize this. Two years is too long
a wait. We were expecting to be up and running in February, 1998, and
now it is March. The further this time is delayed, the more money
will be lost and the more time will be wasted. My web hoster is
rightfully nervous.

I can understand your hesitation in creating the new names; one does
not want to plunge the Internet into turmoil. However, it is clear
that CORE has a solid solution in place and ready.

Respectfully yours,

Dr. Harry P. Cole, M.D.

"Healing is a matter of time, but it is sometimes also a matter of
Hippocrates, 400-460 B.C.

Get your free address at