Comments of the Internet Society

in regards to the

Department Of Commerce

National Telecommunications and Information Administration

15 CFR Chapter XXIII

[Docket No. 980212036-8036-01]

Improvement of Technical Management

of Internet Names and Addresses

("Green Paper")

17 March 1998

 

I. Introduction

The Internet Society (ISOC) wishes to thank the members of the team that produced the referenced document ("Green Paper") for their diligent effort against a formidable time constraint; and, for bringing forward several of the issues that are critical to the continued evolution of the Internet. We are confident that an international open process with no undue influence from any one entity to discuss and resolve these issues will result in conclusions and actions for the best interests of the Internet.

 

II. Executive Summary

The development and evolution of the Internet has been in process for a number of years. In all those years, it has employed principles of self-governance, out of which has emerged the concept of "rough consensus." Rough consensus is reached through a rigorous, openly vetted process that produces the best of alternatives; it does not mean unanimity, but it does mean broad based acceptance. Cooperation amongst the Internet stakeholders produces rough consensus and has brought the Internet to its present successful and robust state. The Internet Society encourages the continuance of this proven concept.

There are several key issues presented in the Green Paper for which ISOC has special concern or interest:

Central Authority - The Internet Society supports the concept of a central authority for the management and administration of: Internet address space, other protocol conventions essential to the use and operation of the Internet, domain names, and the root server system. We note the similarity between the proposal in the Green Paper for this central authority, and that proposed by Dr. Jon Postel of IANA <http://www.iana.org/iana/iana-plan-980113.txt> as developed through the normal process of consensus building which has been employed for the many years of Internet evolution.

The Internet Society expresses support for the proposal of Dr. Postel, and believes there are common grounds upon which these two proposals can coalesce. We believe it would be beneficial to the continued health of the Internet for rough consensus to result on this issue and we encourage all interested parties to commit to supporting the concept.

Registries - A registry is the database repository for one or more gTLDs. A registry is a monopoly. This is an unfortunate fact that exists, at least for now, due to technical limitations. As a result, a registry must operate under different guidelines than a normal business. The potential for abuse through price gouging, once registrants become "locked in" to the gTLD through the proliferation of its URLs, is significant.

Because of its unique situation, the registry function is a service function for the Internet and should be administered on a non-profit basis utilizing shared database technology. This allows multiple registrars to access and utilize the databases in the exercise of registering second level domain names. There should be no "ownership" of gTLDs, and they should be treated as an international resource, subject to the public trust.

There is no particular reason why there couldnít be multiple registries scattered throughout the world, each with a set of gTLDs having a shared database system with multiple registrars.

Registrars - True competition in the registration of domain names resides at the point where the user, or registrant, - the entity who ultimately requests the registration of a domain name - meets the Internet Domain Name System (DNS). That point is clearly the registrar function. Registrars may be located anywhere in the world; are independent businesses each competing for registrant business on the basis of price and value added service; and use the registries as a clearinghouse for assuring uniqueness in domain names. The registrar may certainly be a for-profit entity.

Dispute Resolution - As a result of the work of the International Ad Hoc Committee (IAHC) and the continuing efforts of the many Internet stakeholders working with the POC/CORE/PAB, a truly remarkable set of instruments to facilitate the resolution of intellectual property disputes, efficiently, effectively, and inexpensively has evolved. ISOC strongly supports this capability and urges international recognition and acceptance of its offerings as an option for resolving intellectual property disputes.

Governance - The Internet Society asserts its strong belief that if the Internet is ever to reach its fullest potential, it will require self-governance. If any one entity attempts to control or govern the Internet, it is likely others would protest. It then follows: if consensus is not reached, the Internet could very well become fragmented and, accordingly, cease to exist as we now know it. While we find much in the Green Paper with which we agree, its attempt to define details that would be better left to the process of bottom-up consensus building, appears to go against the principle of self-governance. The US government would be well-advised to leave the details such as: how many registries, for-profit or non-profit; the number of gTLDs; the number of gTLDs per registry; to be defined in a self-determination process such as that described in the History section.

In the formation of the new IANA, the Green Paper calls for the creation of a new organization to represent the Internet user community. ISOC believes itself to be more representative than any other organization existing today. Equally important, it is composed of, and represents, a veritable whoís who of the Internet. Regarding the broader issue of governance, while IANA has been funded by the US government, it is chartered by the Internet Society through the Internet Architecture Board (IAB) in RFC 1601 <http://info.internet.isi.edu:80/in-notes/rfc/files/rfc1601.txt>. While it has accepted US government funding, it has nevertheless operated without other US government involvement or intervention for many years. ISOCís stewardship in this matter is a matter of public record representing a shining example of responsibility in fostering Internet self-governance. As with the Internet Engineering Task Force (IETF), IAB, and the Internet Engineering Steering Group (IESG), autonomy of operation is preserved.

Finally, ISOC believes it would be prudent for all stakeholders to agree to proceed under the authority of the current IANA, with the plan presented by the POC/CORE system. This does not preclude other systems which would conform to requirements established by IANA form being implemented, but consistency should be maintained. This action would be consistent with the aims of the Green Paper and provides a robust means to gain controlled experience along the lines proposed in the Green Paper.

 

III. History

There may be arguments about how old the Internet is, but certainly during the 1990s, the Internet has incurred tremendous growth and transformation. It has emerged from the academic and government environment, primarily within the United States, to a truly international network of tens of thousands of interconnected independent networks. It works because these networks cooperate; they agree to use standards, such as the TCP/IP protocol. The standards were developed through a grass-roots, or bottom-up process in an open forum where only the best ideas emerge.

The Internet reached its present robust state for many reasons: a brilliant protocol; an early founding period wherein the participants were virtually unnoticed; an environment extremely conducive to cooperation among and between the participants; the freedom to experiment in an uncontrolled environment; an open forum using a grass roots approach to standards development; indeed, a self-regulating and self-policing culture. It was from this background that the concept of "rough consensus" emerged. Rough consensus is actually derived through a very rigorous process that weeds out the weak proposals and encourages the best. It does not mean unanimity, but it does mean broad-based acceptance. This is the mode of operation for the IETF, the dominant standards body for the Internet.

Take away cooperation, and the dynamic, innovative Internet that we have come to know and admire will go away.

In the tradition of what made the Internet the phenomenal success that most agree it is today, the Internet Society, in response to a proposal by the IANA, initiated a process to enhance the Internet Domain Name System through the introduction of competition and an efficient, effective, and inexpensive system for dispute resolution between domain names and intellectual property owners. A committee was formed, the IAHC, with its members coming from a broad geographical spectrum as well as international, technical, business, and legal perspectives. Employing all the techniques learned from the history of the development of the Internet, the IAHC produced a plan which defined a structure and processes to achieve precisely what the Green Paper proposes to accomplish regarding domain names.

The IAHC plan, now known as the CORE/POC/PAB, or gTLD-MoU plan, is and has been open for modification and adaptation. It presents an excellent foundation from which cooperative effort, with the US government as one of the participants, could result in an international solution to the issues addressed in the Green Paper. It has the added benefit of having been produced in the traditions of the Internet and would be enhanced by wider participation as herein suggested.

 

IV. The Critical Issues

Among the most critical issues discussed in the Green Paper are:

    1. The concept of a Central Authority for the administration of Internet Protocol (IP) addresses and parameters, domain names, and the root server system;
    2. The concepts of Registries; whether they be for-profit or non-profit; "ownership" of gTLDs by registries; and, the use of shared database systems;
    3. The concept of competitive Registrars;
    4. Dispute Resolution mechanisms between domain name holders and intellectual property owners;
    5. The direct and implied issues of Internet Governance.

A. Central Authority (The New IANA)

The Internet Society expresses its support of the document <http://www.iana.org/iana/iana-plan-980113.txt> written by Dr. Jon Postel and as evolved through an international vetting process. We believe it is similar enough to that proposed in the Green Paper to warrant consideration as a basis upon which to arrive at strong international consensus.

ISOC does not believe it is necessary for the Green Paper to attempt to define the specifics of board makeup for the new IANA and that the board makeup should be determined through public discussion, electronic forum, and other iterative mechanisms to arrive at an acceptable resolution. The IETF has a long and successful track record of resolving complex and difficult issues through this kind of process and these deliberative processes can be used as a role model.

As the new IANA relates to TLDs, registries, and registrars, the Internet Society offers the following observations which have recently been discussed within the IAB and the IESG:

    1. All TLDs, generic and country-code based, should derive their authorization and existence from the new IANA. The general rules of RFC 1591 <ftp://ftp.isi.edu/in-notes/rfc1591.txt> would apply.
    2. The special gTLDs of .mil and .gov should be treated, for historical reasons, as additional country code TLDs for the USA.
    3. For generic TLDs, the IANA board would designate either one registry with responsibility for all gTLDs or several registries, each with responsibility for one or more gTLDs. The designation process would involve proposals and competition; the registries would be operated as non-profits clearly separated from any other components of the same organization; title to the data and databases would clearly remain with IANA (the registry would obtain no equity interest in the databases by consequence of assembling, holding, or operating them); and, the IANA contracts with registries would contain provisions for appeals, frequent review of registry operations, auditing and testing of ability to transition to other registry operators, and early termination of contracts for not operating the registry efficiently or in accord with the public interest. IANA should be able to require performance bonds and/or other mechanisms for ensuring appropriate behavior both during registry operations and in the transition from one registry to another.
    4. In general and in the long term, economies of scale are likely to tend toward a single registry, operated as above, for all or most gTLDs. However, this is not a requirement; the IANA board may make other decisions and arrangements. For a short transition period (see vii., below), it is likely that the registry for the existing gTLDs would be different from that for new ones.
    5. In general, all authorized registrars would be permitted to make registrations in all gTLDs, and it would be a contractual requirement on registries to accept such registrations. Final review of qualifications for registrars (and appeal from any delegated system) would rest with IANA and not with the registry. However, the IANA board may accept proposals for specific gTLDs with different rules about candidates for registration and registrars; as with any other gTLD, IANA shall designate an appropriate registry.
    6. The operation of the root domain and corresponding servers and supporting services is, as outlined in the GP, an IANA responsibility. Should IANA choose to delegate day-to-day operational responsibility to an organization with registry or registrar responsibilities, sufficient contractual and organizational boundaries would be required to avoid any plausible possibility of abuse of the relationship.
    7. For a transition period and in order to facilitate that transition, the US government is encouraged to contract with a qualified organization for the registry operation of any existing gTLDs (e.g. .edu, .gov) not handled elsewhere. Such a transitional contract should be short, should explicitly recognize IANAís oversight authority, and should contain the general provisions outlined above (specifically, non profit, public trust assumptions, IANA ownership of data, IANA appeal and audit, clear separation from any organization in the registry or root operations business - noting that several of these provisions are slight refinements on those proposed in the Green Paper).
    8. In order to support the speedy transition that is generally agreed to be preferable to a long and extended process, the qualifications for registrars established in the gTLD-MoU should be taken as a starting point. The existing registrars authorized by that process should be recognized conditioned upon immediate re-opening of the application and acceptance process. Once the IANA board is established it may supplement those criteria, or, as suggested above, establish different criteria for new gTLDs as needed.

 

B. Registries

Principal stewardship for the gTLDs and related data will be seated in the new IANA which will authorize registries as executors of this international public trust. The registries do not own any of the data, any more than Trust Officers at banks "own" the assets they steward, or neonatal ICU nurses "own" the babies for whom they care. The fact that those professional stewards do not "own" the objects in their care does not at all diminish their moral and legal obligations to execute their charge to the best of their ability and with all due diligence. Further, the execution of that trust gives them no long-term claim on the objects in their care. They may be life-critical but are nonetheless transient caretakers.

The registries collect, manage, and keep safe, a data corpus which is the fundamental glue of the Internet, and their job is to ensure the accuracy and high availability of the data entrusted to their care.

It is the explicit, direct responsibility of the new IANA to execute this international public trust. In doing so, it may elect to delegate the operational execution of this trust to registries; but, the ultimate responsibility for the execution of that public trust cannot be delegated. To this end, the new IANA must maintain an absolute ability to secure the timely and accurate execution of that public trust, including the removal and reassignment of a registry.

Serving as a registry is a task of significant gravity and responsibility which must be executed with supreme professionalism and attention to the sole charge of the public trust: the safe and accurate stewardship of the data. Organizations who imagine serving as a registry should think hard about what they are signing up to do. This is serious, world-class, public trust computing where failures can not be tolerated.

Registries are not meant to compete, although having multiple registries throughout the world would not necessarily be undesirable. Certainly registries must conform to minimum standards of operation such as hours of operation, maximum turn around time for registration, etc., but their customers are the registrars and not the end user, or registrant. When one first registers under a gTLD, they pick the gTLD depending upon what most fits what they are trying to project. Thereafter, they will rarely change their gTLD, but may find it highly desirable to change registrars.

The ability to change registrars, which gives a degree of "portability" to a domain name holder, requires that registries utilize a shared database capability that assures uniqueness in the registration of domain names. As long as all registries have a shared database capability conforming to the requirements of the DNS, multiple registries could exist.

As the repository for the database for each gTLD under its stewardship, the registry is, in fact, a monopoly for each of the gTLDs it services. As such it must operate under a different set of rules than a normal for profit business.

If a registry could be "for profit" based, there is the risk of significant price gouging, once a domain name is registered. The nature of the Internet is such that domain names become embedded in countless URLs and other forms of notice. They get linked in cascading fashions which make it extremely difficult to change, once released. The switching cost, therefore, could become quite high and, thus, a for-profit registry could take advantage of that plight by raising prices significantly.

We recognize that there are significant costs associated with maintaining and operating a registry, but we also know that those costs can be identified and pricing to registrars can be established based upon those costs. Accordingly, it is clearly desirable that registries be non-profit, with a cost-based pricing mechanism to its customers - the registrars.

C. Registrars

Registrars are the interface between the end user, or registrant, and the Internet DNS. It is here where true competition that benefits the user community will be beneficial. With multiple registrars operating worldwide under a shared registry system, consumers have the ability to pick which registrar best fits their needs, whether that is based upon price, quality of service, language capabilities, hours of operation, convenience of registration, etc. The registrant is not confined to any particular registrar because each registrar can register domain names under the same gTLDs.

However, even in a system where there are multiple registries, it is possible that a registrant may be compelled to deal with multiple registrars if the gTLDs employed by the registrant are housed in different registries, and the registrar is not affiliated with the designated registry. Given that registries are a service to the Internet, and specifically to the registrars as a public trust, this concept would seem to strengthen the case for one registry.

ISOC does not take a position on whether there should be one master registry, or multiple registries; only, that if there are multiple registries, they must have shared database systems with multiple registrar capabilities and be non-profit oriented.

D. Dispute Resolution

The Green Paper does not define a specific methodology for the resolution of disputes between intellectual property owners and domain names, but suggests certain minimum standards. This is superficially appealing but is likely to increase the confusion in an already complex legal arena. Owners of global trademarks and major brand names are already protesting the complexity of dealing with national law differences and differences between various administrative systems for handling trademark-domain name disputes. The Internet Society strongly encourages the US government to look seriously at the results of the consensus building effort of the IAHC/POC/CORE/PAB - gTLD-MoU process that developed the dispute resolution mechanisms in conjunction with the World Intellectual Property Organization (WIPO).

There is a robust set of procedures to efficiently, effectively, and inexpensively resolve disputes between domain name holders and intellectual property owners in the form of online expedited mediation, online arbitration, and Administrative Challenge Panels (ACPs) to allow preemptive prevention of the use of internationally famous trade marks in domain names by other than the rightful owner. More specific and detailed information can be found at the gTLD-MoU web site at <http://www.gtld-mou.org/>, or directly to the dispute resolution section at <http://www.gtld-mou.org/#dispute>.

The proposal in the Green Paper to let "each name registry" establish its own dispute resolution policy will introduce confusion and perhaps chaos into an already complex issue, and such a system might unduly favor one region compared to others. ISOC strongly encourages the US government to remove that proposal from the Green Paper. Instead, the Green Paper should contain a recommendation to adopt a specific, existing methodology for dispute resolution such as that referenced in the gTLD-MoU, as a basis upon which to begin. Allowing each registry to establish its own dispute resolution policy will create a situation where one of the criteria for the selection of a registry by a customer could be the type of dispute settlement to which the customer will be subject. This could, in turn, create a strong incentive for each registry to adopt a dispute policy that maximizes its business, namely a policy that is exclusively domain-name-owner-friendly, and trademark-owner-unfriendly, rather than adopting an effective policy that will intelligently balance interests and maintain a stable commercial Internet space for global electronic commerce.

The gTLD-MoU dispute resolution mechanisms do not impede or otherwise preempt or override national laws or court systems and always allow for participants to revert to those national systems at any time. Indeed, the procedures fit very well the criteria defined in the Green Paper for desirable systems and were developed by a wide body of legal experts, business interests, trademark experts, domain name holders, and interested parties of varying backgrounds, from around the entire world. The resultant system is a comprehensive and robust tool for Internet stakeholder use! We recognize that, beyond national court systems, there are other established and proven dispute mechanisms available.

 

E. Governance

As stated in the Introduction section of this document, the Internet Society is strongly supportive of the US governmentís desire to withdraw from any direct involvement in the development, operation, or management of the Internet. We have a concern that the Green Paper leaves the door open for continuing controlling involvement by the US government in the regulation of the Internet for as much as two years. That time period could be significantly reduced by the US governmentís support of, and participation in, the continually evolving gTLD-MoU, which resulted from the one-and-a-half years of consensus building within the international Internet industry.

We are similarly supportive of the US government proposal to assist in the creation of a new IANA that is non-profit and governed by a policy oversight board which must be open, transparent and accountable to Internet participants. We are compelled to point out that an effort to reinvent IANA has been underway for some time as a result of the anticipation that US government funding of the existing IANA would cease in the fall of this year. In keeping with the processes by which the Internet has reached its present robust state, that is, bottom-up consensus building, a proposal has been evolving that was written by Dr. Jon Postel, the current head of IANA. The most recent iteration of that proposal can be found at <http://www.iana.org/iana/iana-plan-980113.txt>.

ISOC recommends the involvement of the US government, as one voice among many, in the evolution of the Postel proposal and points out that it is similar to that proposed in the Green Paper. We believe that most Internet stakeholders would be eager to rally around a combined effort consolidating these plans and resolving any differences. This could be done fairly quickly via electronic forum and any appropriate public meeting forum.

While the US government has funded IANA for many years, it has not interfered in the operation of the organization nor in the decisions IANA has made. IANA has a sterling track record of responsible operation and is recognized for its impartial, fair, and consensus building methods in executing its responsibilities. Further, IANA has operated under the auspices of the Internet Society, in a fully autonomous mode consistent with the IAB, IESG, and IETF, as noted in RFC 1601 <http://info.internet.isi.edu:80/in-notes/rfc/files/rfc1601.txt>.

While we donít believe that it is necessary to wait until the new IANA has been duly constituted in order to begin adding new gTLDs to the root servers, we do believe that the definition of gTLDs, registries, numbers of gTLDs, and like detail should not be defined by the US government, nor by any one institution. This is counter to the principle of self-governance, for which the US government has so often expressed support. Once the new IANA has been created, its processes will be in the best position to make those determinations.

We do not see a need for the creation of a new membership association. The Internet Society has existed for approximately six years. It is more representative of the Internet user than any other organization. Any individual or organizational entities are completely welcome to join. We are open to new members, and democratically run. ISOC has a Board of Trustees elected by the world-wide individual members. It has a set of bylaws <http://www.isoc.org/isoc/general/trustees/bylaws.shtml>, a set of written rules and operating procedures, and (most importantly) a history of democratic operations. In addition to having thousands of individual members from over 150 countries, ISOCís over 140 Organization members make up a veritable whoís who of the Internet. ISOC is also the organizational home of the democratically run IETF, IAB, IESG, and Internet Research Task Force (IRTF) - the standards setting and research arms of the Internet community with bottom-up consensus building through thousands of participants. ISOC has demonstrated leadership in many areas, including organizing the effort to introduce competition into the DNS, one that has garnered more industry support than has any other proposal.

 

V. Conclusion

Fortunately, the USG is not looking for industry unanimity in how the Internet should be governed, because unanimity is not likely to happen. If it is willing to accept rough consensus, then it should look towards the gTLD-MoU plan as the role model. That plan, which created POC/CORE, resulted from a robust process that was thoroughly vetted in multiple public forums, and was more publicly scrutinized than any issue ever introduced into the Internet. The gTLD-MoU has more than 200 signatories (there were 217 as of 24 February 1998; the list continues to grow) and there are currently 88 registrars established. CORE has plans soon to open the application process for registrars and 22 new potential registrars have already indicated their intent to apply.

Those who could not, under the gTLD-MoU plan, carve out a piece of the Internet to call their own will never agree to true competition that leaves the end-users in the driverís seat. But who should the US government look to benefit, if not the end-users? It is not the role of the US government to protect the vested interests of those who would turn the Internet into their private profit generator. Rather, the US government should support industry efforts such as the gTLD-MoU, to transition to true industry self-governance.

We are in full agreement with the stated goal of the US government: " . . . to seek as strong a consensus as possible so that a new, open, and accountable system can emerge that is legitimate in the eyes of all Internet stakeholders." We submit that to ignore the efforts of major Internet stakeholders, in the form of the gTLD-MoU plan, does not lend itself to attaining a strong consensus.

We agree that the US government should not " . . .cede authority to any particular commercial interest or any specific coalition of interest groups." However, we caution that the proposed new membership associations, while in the process of evolving, may wind up controlled by a handful parochial interests. We submit that the existing user and registrar groups (ISOC and CORE) should be assisted and strengthened.

Concerns have been raised about anti-trust litigation. We suggest that the US government in cooperation with other governments, quickly draft anti-trust assurances that will support industry self-governance supporting the will of the world-wide Internet community.

We believe it would be instructive and beneficial for the entire process to proceed, under the current IANA, with the introduction of the POC/CORE system. Specifically, this would mean the introduction of new gTLDs into the root servers, the implementation of the already developed and tested shared registry system of CORE, and the use of the now existing dispute resolution procedures developed in the gTLD-MoU process. We recognize the possibility that other systems which would conform to the requirements of IANA may emerge and this action would not preclude their implementation. We simply believe this step would provide a robust means to gain controlled experience along the lines proposed in the Green Paper.

 

Contact Information:

Internet Society

Donald M. Heath

President/CEO

12020 Sunrise Valley Drive

Suite 210

Reston, VA 20191-3429

TEL: +703 648 98 88

FAX: +703 648 98 87

EMAIL: info@isoc.org

http://www.isoc.org/