Declaration of Louis Touton in Opposition to Plaintiffs' Motion for Preliminary Injunction Smiley v. ICANN (14 September 2001)

DECLARATION OF LOUIS TOUTON

I, Louis Touton, declare:

1. I am the Vice-President, Secretary and General Counsel of the Internet Corporation for Assigned Names and Numbers (ICANN). I have personal and first-hand knowledge of the facts set forth in this declaration and am competent to testify to those facts if called as a witness. I make this declaration in connection with ICANN's opposition to plaintiffs' motion for preliminary injunction.

2. The purpose of my declaration is to provide some background on the Internet's Domain Name System, ICANN's management of that system, and ICANN's efforts to expand the number of available domain names, including the new ".biz" names. I also discuss the harm that would result if a preliminary injunction is issued.

Background of the Internet's Domain Name System

3. The Internet is a network of networks that allows computers around the world to communicate with each other quickly and efficiently over a variety of physical links. These computers (and other devices) serve a variety of purposes: hosting web sites, handling e-mail, and providing access points for users. For the Internet to function effectively, each computer connected to the Internet must have a unique identifier, or address, so that information can be routed to and from that computer.

4. The unique identifiers used by Internet computers to route traffic and establish connections among themselves are lengthy numerical codes known as Internet Protocol (IP) numbers or addresses. For example, the IP number for the computer that hosts the Los Angeles Superior Court's Internet web site is 209.80.48.82.

5. Because Internet users cannot easily remember IP numbers, most Internet computers also have a unique, user-friendly address, called a "domain name," which corresponds to the computer's IP number. The domain name for the Court's web-site computer is "www.LASuperiorCourt.org".

6. User-friendly domain names would be useless without an effective way to translate domain names to the IP numbers that computers use to communicate among themselves. Such translation enables a user to access a website by typing the domain name rather than the IP number into a web browser.

7. Nearly all Internet computers translate domain names to IP numbers by using the Domain Name System (DNS), which the Internet engineering community devised in the early 1980s. The DNS consists of a hierarchical network of computers known as "nameservers." At the top of the hierarchy is a set of thirteen "root" nameservers, located at various sites around the world, identified by the letters A through M. After a domain name (with a corresponding IP number) is registered in the DNS, any computer with software called a "DNS resolver" can, through the hierarchical communication channels of the DNS, obtain the necessary information to translate that domain name to the corresponding IP number upon request. Thus, although some people might view themselves as "owning" a domain name, in fact, "ownership" of a domain name signifies only a right, for a period of time (usually renewable), to have the DNS translate that name to a given IP number.

ICANN's Role in Management of the DNS

8. Because the Internet arose primarily through research conducted or funded by the U.S. government, much of the DNS (including operation of the root-server system) historically was operated by government agencies or under agreements with them. In 1997, however, the President directed the Secretary of Commerce to privatize the DNS. After soliciting public comment, the Department of Commerce (DOC) issued in 1998 a "White Paper" stating that "the U.S. Government is prepared to recognize, by entering into agreement with, and to seek international support for, a new, not-for-profit corporation formed by private sector Internet stakeholders to administer policy for the Internet name and address system." A copy of the White Paper may be found on the Internet at a website maintained by the U.S. Department of Commerce at http://www.ntia.doc.gov/ntiahome/domainname/6_5_98dns.htm.

9. In response to the government's challenge, a broad coalition of the Internet's business, technical, academic and user communities formed ICANN in October 1998. ICANN is a non-profit, public-benefit (charitable), private-sector corporation, based in California but consisting of several internationally representative bodies (ICANN's board of directors, for example, includes nineteen individuals from eleven countries). ICANN has been recognized by the U.S. and other governments, as well as by technical standards-development bodies and other private-sector entities involved in the Internet's operation, as the global consensus entity to coordinate technical management of the DNS. ICANN's mission is to coordinate the operation of the DNS based on policies that are developed by the diverse stakeholder communities of the global Internet through a broadly representative, bottom-up, consensus-based process. In November 1998, ICANN and the DOC entered into a Memorandum of Understanding (MOU) in which they agreed to work together to manage the transition from government control to private-sector control of the DNS.

Expanding the Number of Domain Names Available for Registration in the DNS

10. The single most visible element of DNS technical management is the registration of domain names. The DNS uses a hierarchical naming system, with every domain name ending with a suffix, known as a "top-level domain" (TLD), appearing to the right of the last dot. Each TLD has a "registry" of so-called "second-level domain names" within it (for example, the .org registry contains the name "LASuperiorCourt"). Most of the TLDs correspond to the nations of the world, such as ".us" and ".uk", and are known as country-code top-level domains (ccTLDs). Other TLDs, such as ".com" and ".org", are known as generic or global top-level domains (gTLDs). Until recently, there were only seven gTLDs, and only three of these (.com, .net, and .org) were generally available to public registrations.

11. Even prior to the formation of ICANN, much of the Internet community came to believe that new gTLDs should be introduced. As the Internet has grown dramatically from thousands of computers to hundreds of millions, and has evolved from primarily a U.S.-based academic and research network to a ubiquitous global medium for commerce and information exchange, there has been strong motivation to augment and diversify the gTLDs established in the mid-1980s when the DNS was originally deployed. The June 1998 DOC White Paper reflected the widely held belief that increased competition within the DNS will serve to "lower costs, promote innovation, encourage diversity, and enhance user choice and satisfaction." At the same time, the almost uniform consensus in the community has been to proceed with caution so as to minimize the risk that new gTLDs will cause technical instability in the Internet, which now plays an increasingly critical role in everyday commercial, academic, and personal life.

12. The November 1998 MOU expressly contemplated that ICANN and the DOC would implement a process to study the possibility of expanding the range of available names. A copy of the MOU may be found on the DOC's website at http://www.ntia.doc.gov/ntiahome/domainname/icann-memorandum.htm. The MOU, however, contemplated that ICANN would manage a process to introduce new gTLDs into the DNS in a safe and fair manner, and ICANN has devoted a significant part of its activities to this issue since it was formed.

13. Because ICANN is a consensus-based body that relies on bottom-up policy development, the issues of whether and how to introduce new top-level domains were first taken up by the Domain Name Supporting Organization (DNSO), the ICANN global constituent body responsible for domain name policy issues. The DNSO organized a Working Group consisting of individuals with a broad range of expertise, which made recommendations after many months of study and deliberation. Upon reviewing the Working Group's report and public comments on it, the DNSO's executive body adopted the Working Group's recommendations that, "[b]ecause there is no recent experience in introducing new gTLDs . . . a limited number of new top-level domains [should] be introduced initially and that the future introduction of additional top-level domains [should] be done only after careful evaluation of the initial introduction" and "that several types of domains should be considered in the initial introduction." These recommendations were forwarded to the ICANN Board, which considered well over 1000 public comments on the DNSO's recommendations. In July 2000, the Board adopted the recommendations.

14. The recommendations adopted by the ICANN Board for a limited, but varied, initial introduction of TLDs reflect a consensus within the Internet community that mirrors the recommendations made by key organizations that have examined the question of how best to serve the public interest in a robust, growing Internet: In particular, the Board determined that a "small" or "limited" or "prudent" number of top-level domains should be introduced initially and that the effects of that introduction should be evaluated before proceeding with more wide-spread introductions of new TLDs. Since July 2000, ICANN's efforts have been focused on achieving this goal, which has been described as a "proof of concept" process to ascertain the effects of adding new TLDs and to determine how they may be added with the least risk of causing instability or unfairness.

15. At its July 2000 meeting, the ICANN Board established a procedure for selecting a limited number of gTLDs that would be introduced in a measured and responsible fashion. In particular, it was necessary to select a "registry operator" to manage each new TLD. The basic functions of a registry operator are to register domain names (which can number in the millions) within the TLD it operates and respond to millions of queries from computers seeking to translate those domain names to IP numbers. Registry operators traditionally accept domain name registrations from independent "registrars," which essentially act as middlemen between the registry operator and the many individuals and entities seeking to register domain names.

16. In August 2000, ICANN issued a formal call for proposals from entities seeking to be selected as registry operators for new gTLDs, and it identified the criteria by which those proposals would be assessed. ICANN's goal was not to select the "best" application, but to select a range of proposals that, when implemented together, would allow the Internet community to assess the performance of various viable responses to the challenges posed by the introduction of new TLDs. Apart from the paramount concern for maintaining Internet stability, important factors included enhancement of competition for registration services, enhancement of DNS utility, increasing the diversity of registration services, and assuring fairness in the registration of domain names.

17. ICANN assembled a team of technical, financial, and legal experts to review and evaluate the applications for new TLDs, 44 of which were received by ICANN. The Internet community reviewed the experts' 326-page report and the applications themselves (which were posted on the Internet), engaged in a robust dialogue through ICANN's public forums (including over 5000 postings and an all-day open meeting), and intensely discussed the issues through ICANN's constituency and supporting-organization structures. After extensive review of the applications, experts' report, and public input, the ICANN Board discussed the applications in a several-hour open session on November 16, 2000.

18. This careful process resulted in the selection of seven proposals, including the proposal by JVTeam, LLC - now known as NeuLevel, Inc. - to become the registry operator for a ".biz" gTLD. ICANN then began negotiating agreements with the selected applicants to ensure introduction of the new gTLDs in a manner consistent with the consensus goals of the Internet community. ICANN posted drafts of these agreements on the Internet and considered numerous public comments before finalizing any agreements. Agreements with registry operators for three of the new TLDs (.biz, .info, and .name) have been completed, while the agreements with other four new TLDs (.pro, .aero, .coop, and .museum) are at various stages of completion.

19. Following completion of satisfactory agreements with ".biz," ".info", and ".name," ICANN recommended to the DOC that these new TLDs be added to the authoritative root-zone file. DOC accepted those recommendations and caused those three TLDs to be added for staging, testing, and evaluation, to be followed by full-scale deployment. The three TLDs, with skeleton contents, are now active in the authoritative root, and many steps (including the expenditure of millions of dollars) have been taken toward activating the TLDs to accept the millions of second-level domains that are expected to be registered.

Registration of Newly Available Domain Names

20. The first, essential phase in the introduction of a new TLD is the registration of domain names ending with the new suffix. To ensure that the registry operators fulfill their promises to serve the public through their operation of the TLD, each of the three finalized registry agreements includes performance and service-level specifications, as well as a schedule of maximum fees to be charged to registrars by the registry operators for initial registration, renewal, transfer, and other services. The agreements also include a Registry Code of Conduct that affirms the registry operator's role as a trusted neutral third-party provider of registry services to accredited registrars.

21. In addition, the introduction of new gTLDs has been designed to foster competition - and, as a result, the level of service to the Internet community - through the operation of multiple registrars that, for a fee, act for customers in gathering the appropriate data, properly formatting it, and inserting it into the registry. The June 1998 DOC White Paper expressly acknowledged the nearly universal support for a competitive registrar system. To accomplish this consensus goal, ICANN has developed an accreditation system for registrars. To become accredited, a registrar must enter into a Registrar Accreditation Agreement with ICANN and a Registry-Registrar Agreement with the registry operator for the TLD. Registry operators such as NeuLevel must give equal access and treatment to any ICANN-accredited registrar that seeks to enter into such an agreement for the purpose of being able to offer consumers the opportunity to register domain names in the new TLD.

22. One of the most significant issues raised by the introduction of new TLDs is the technical mechanism for start-up of the TLD. Throughout the years of discussions and deliberations within the Internet community regarding new TLDs, concerns have arisen from the fact that the introduction of a new TLD is likely to lead to a massive initial surge in registrations as people seek to register the most desirable domain names. This "landrush" must be handled in a manner that keeps the registry system operating in a stable manner while assuring a fair system of allocation among the many applicants for names. Because no gTLDs have been introduced since the 1980s, the Internet community has no real-world experience concerning how to meet the challenges of introducing a new, large TLD in the now-heavily-commercial Internet.

23. As noted above, the introduction of new TLDs is intended as a "proof of concept," so that the best approach to these and other problems is not decided in the abstract, but through close monitoring of different approaches in actual operation. ICANN already has created a TLD Evaluation Process Planning Task Force to craft a plan for monitoring the introduction of the new gTLDs and for evaluating their performance and impact on the performance of the DNS. It is anticipated that one of the focal points for this evaluation will be the registration process during the "landrush" phase.

24. In order for the "proof of concept" phase to be as useful as possible, ICANN endeavored to ensure that start-up protocols would be designed with great care to address fairness concerns, including those involving registration of names in which some claim intellectual property rights. ICANN's initial call for proposals specified review criteria, including whether the proposal provided for orderly and reliable assignment of domain names during the start-up phase, with appropriate protections for the legitimate interests of existing stakeholders and against abusive registration practices. The report by the panel of experts that reviewed each proposal specifically addressed the start-up plan of each applicant and, in addition, compared different responses to the "landrush" problem. Finally, each of the three agreements entered to date contains a detailed start-up plan.

25. Each of the three start-up plans calls for an initial phase, prior to the "landrush," to address trademark and other intellectual property claims with respect to specific domain names. For the subsequent "landrush" phase, each plan calls for a submission period from which all applications submitted will be aggregated and a random-selection process will be used to determine the order in which domain-name applications will be processed during the "landrush" phase.

26. In light of ICANN's consensus-based approach to the introduction of new TLDs, it is apparent that the Internet community strongly supports the testing of random-selection processes during the "proof of concept" phase. A pure "first-come/first-served" system for new gTLDs is generally viewed as having two significant drawbacks. First, using a first-come/first-served system for the introduction of a large TLD is likely to result in processing requirements exceeding the capabilities of the registry system, as thousands (or even millions) of requests are made for the most desirable names. Although in a post-landrush environment names are commonly registered on a first-come/first-served basis, recent experience has indicated that it is not feasible to use first-come/first served during an opening of a highly demanded new namespace. (For example, late in the year 2000, first-come/first-served opening of internationalized domain name space within .com, .net, and .org overwhelmed the registry system for those particular TLDs, and this is the most capable DNS registry system in existence.) In addition to the technical challenges, first-come/first-served during high-demand periods can also cause unfair allocation as those with expensive, high-bandwidth, mechanized-script query systems shut out access by others using conventional means. This has occurred recently as some companies have sought to re-register expiring names by making tens of millions of rapid-fire queries to register dozens of names. When this high demand results in ordinary customers being denied access to the registry system altogether, serious questions arise about the fairness of the allocation. Although the internationalized-domain-name and expiring-names incidents suggest the problems in dealing with the registrations of names with high demand using first-come/first-served mechanisms, these problems would likely pale in comparison to those that would arise in the introduction of a large TLD using those mechanisms.

27. Other alternative approaches to the "landrush" pose a number of different problems. (Indeed, after reviewing many proposed approaches to "landrush" allocation, every approach appears to have drawbacks.) For example, at least in a large TLD, it generally is considered neither feasible nor appropriate for registration of domain names to be based on subjective judgments regarding the relative merits of competing applicants or applications. Among other things, it seems unlikely that judgments about relative "desirability" of competing applications would be accepted without protracted and costly challenges.

28. Similarly unappealing to most Internet stakeholders is a system that would auction priority processing to the highest bidder. A few of the proposals received by ICANN offered a variation on this theme, in that they called for initial demand to be dampened through imposition of higher registration fees at the outset of the processing period. For the most part, however, the Internet community does not support a system that directly advantages those with greater resources, as auction schemes are likely to do. In one recent small-scale episode, an auction mechanism was used during 2000 for the registration of certain names within the ".tv" ccTLD (associated with the island nation of Tuvalu), which resulted in a dramatically increased cost to the public for registering those names (some in excess of $5000 per year). That registration mechanism also resulted in associated activities where names were registered abusively with the intent of ransoming them to trademark owners. Since domain-name registrations are intended to be accomplished in a manner that facilitates the Internet's usage and makes it affordable and accessible to all, the use of this auction system provoked a significant number of complaints (and some considerable outrage) in the Internet community.

29. The widespread appeal of random-selection processes appears to arise from the fact that, in theory, it does not advantage any applicant over another. Every application has a fair chance. This, undoubtedly, is the main explanation for the fact that most applications for new TLDs that ICANN received during the year 2000 proposed some form of randomized selection. It also helps explain why every selected large TLD employs some type of randomized process during "landrush". Among the selected large TLDs, however, there are a variety of implementations that will be used. As noted above, at least in theory all approaches have their own drawbacks, including vulnerabilities to "gaming" by unscrupulous applicants. The purpose of the current "proof of concept" is to ascertain how these theoretical problems translate into practice, so that the optimum mechanisms for introduction of future TLDs can be developed base on actual experience.

30. One theoretically appealing mechanism for fair, randomized selection during landrush is to limit each applicant to a single application. This approach, however, is not feasible at reasonable cost to the public. Ensuring that each applicant submits only one application is not readily done by automated means. In view of the clever (and sometimes unscrupulous) approaches to obtaining advantage that appear whenever highly demanded domain names made available, it appears clear that some type of manual review of applications would be necessary to ferret out straw-man applications and the like. Given the very large numbers of applications are likely to be involved, my estimate is that having a reasonably rigorous manual review to limit each applicant to one application would raise registry registration fees from $5 - $6 per year to several times that amount, and even then, some duplicate applications inevitably would slip through. Again, this approach would not achieve the basic goal of making the Internet easily accessible and affordable to all.

31. The schedule for ".biz" registrations was announced to the Internet community on May 15, 2001. These announcements were available on ICANN's website, as well as NeuLevel's website, and were republished in a number of news articles. Beginning in late May, NeuLevel accepted submissions of trademark claims. Beginning in July, NeuLevel opened the application process. As was announced in May, NeuLevel has accepted applications from all applicants who choose to submit them, without attempting the technically impractical task of preventing multiple applications from the same applicant, with each application subject to a $2.00 processing fee. All applications received by September 17, 2001 will be organized into batches, which then will be processed in random order. Domain names registered through this process will become operational in early October.

32. The introduction and operation of the ."biz" gTLD and other new TLDs will not increase the revenue that ICANN receives. Although registry operators contribute to meet ICANN's cost-recovery needs, the total amount of those contributions are based on ICANN's expenses and is apportioned among all TLD registries. As a result, the number of TLDs and the number of domain names within each TLD affect the proportional contribution paid by each registry operator but have no impact on the total contributions received by ICANN. ICANN receives no part of the $2.00 application fee or other fees charged by NeuLevel in the domain name application and registration process.

33. On the other hand, the introduction and operation of the ".biz" gTLD will represent a momentous step forward in a long process of enhancing the utility of the Internet to the public throughout the world. By providing additional choices of names that can be registered, the .biz TLD will allow the Internet to grow and newcomers to obtain easy-to-remember names to support their Internet activities. By providing an alternative to the .com TLD, the introduction of the .biz TLD will increase the competitive choices of businesses seeking to have a presence on the Internet, and will also allow them to reap the benefits of competition through a variety of domain name options and services. Much of the community has been eagerly awaiting the introduction of a new gTLD since long before ICANN's formation in 1998, and many throughout the community have been complaining that there are not a sufficient number of easy-to-remember domain names available in the presently available gTLDs. Through ICANN, the community has expended immeasurable effort over the past three years to facilitate the achievement of competition and expansion in a careful and responsible manner. Interruption of the process at this juncture would have a significant adverse effect on the ability of the worldwide Internet community to enjoy the full advantages of that medium. It likely would create concern and confusion among those who have attempted to obtain new TLDs.

34. Indeed, I am extremely concerned that any delays associated with the roll-out of ".biz" could have significant impacts on the roll-outs of the other new TLDs. At a minimum, such a delay would create confusion as to whether the schedules that have been adopted for the other TLDs have been (or might be) changed. Internet users rely heavily on the stability of the Internet, and changes to the Internet need to be made prudently, cautiously, and predictably. A change, at this late date, in the start-up of the ".biz" TLD would cause great concern as to whether future changes would be made in this manner. The harm to the public of disrupting the long and careful process of introducing new TLDs would be, in my estimation, quite substantial.

I declare under penalty of perjury under the laws of the State of California that the foregoing is true and correct. This declaration was signed on September 14, 2001, at Marina del Rey, California.

__________________________
Louis Touton