Before the DEPARTMENT OF COMMERCE NATIONAL TELECOMMUNICATIONS & INFORMATION ADMINISTRATION
Washington, D.C. 20230
Improvement of Technical Management of Internet Names and Addresses Proposed Rule
Docket No. 980212036-8036-01
Dated: March 23, 1998.
COMMENTS OF PGMEDIA
pgMedia, Inc. <http://namespace.pgmedia.net>, respectfully submits these comments on the "Green Paper" <http://www.ntia.doc.gov/ntiahome/domainname/dnsdft.htm> released by the National Telecommunications and Information Administration ("NTIA" or the "Department") regarding management of the Internet Domain Name System ("DNS").
INTRODUCTION AND SUMMARY
NTIA's Green Paper proposes that certain key "centralized" functions for DNS administration be transferred from Network Solutions., Inc. ("NSI") and the Internet Assigned Numbers Authority ("IANA") to a new, not-for-profit corporation that will be governed by a broad cross-section of Internet stakeholders. This approach is clearly preferable to the current DNS system, under which NSI enjoys monopoly control of all generic top-level domains ("gTLDs"), including the huge ".com" registry, as well as the DNS "root" servers on which universal Internet connectivity is based. NTIA is plainly correct that competition in domain name registration services, on the basis of shared gTLDs, should be introduced as soon as possible.
The new Internet governance structure contemplated by the Green Paper would remove the anticompetitive incentives inherent in the current DNS system, under which potential registry competitors must depend on NSI, which controls the "essential facility" of the "A" root server, for access to "the dot." By contemplating a "set of authoritative and consistent roots" managed by a competitively impartial entity, instead of a single root server managed by a single commercial enterprise, the Green Paper would place the root server resources -- and with them the public interest requirement for universal resolvability of all Internet domains -- beyond the threat of anticompetitive conduct or catastrophic network failure. Such a development would mean that the customers of entrepreneurial registries such as pgMedia could, for the first time, have a fair chance to become a true part of the global Internet.
The Green Paper has a fatal flaw, however, in proposing that only five new gTLDs be opened for competition in the period through September 30, 1998. NTIA apparently believes that such a restriction is necessary (1) to assure technical "stability" of the Internet; (2) to assist trademark holders in "policing" violations of their intellectual property; and (3) to avoid "consumer confusion" from a proliferation of gTLDs. Even if the Commerce Department (or NTIA), let alone the United Stated government, had the legal authority to impose a limit on the number of new gTLDs -- which it does not -- none of these policy justifications has any validity.
As NSI itself has conceded, there "is no technical reason" why gTLDs need to be limited, because "[d]omains are equally scalable at all levels of the DNS. "The "dot" is, in reality, a simple 75K text file that can easily be expanded to virtually unlimited size, and at least comparable to the current ".com" zone file of 1.5 million entries. Thus, there is no stability reason why the Green Paper must artificially restrict the number of new gTLDs than can be added to the root servers, especially in the short run.
There is no difference between the Internet and any other medium of communications (TV, magazines, billboards, etc.) with respect to trademarks, because in each instance trademark owners have the resources, and the legal power, to assert challenges to their legal rights in any jurisdiction worldwide.
Consumers will not be more confused by different gTLDs than they are today by the massive proliferation of incrementally different second-level ".com" domains (e.g., flowers.com, florists.com and flowersforyou.com); rather, the introduction of new gTLDs (e.g., flowers.delivery or caruso.florists) could produce a new level of comprehension in consumer use of DNS.
In short, the Green Paper relies on misconceptions, unsubstantiated fears and chimeras in support of its crucial finding that only five new gTLDs should be added to the Internet root server system.
pgMedia strongly believes that, given the global nature of the Internet and the status of gTLDs as global communications resources, the U.S. government does not have the legal authority to restrict competition in top-level namespace on the Internet. There is no question that the Department of Commerce and NTIA lack any statutory authority to promulgate rules restricting the opening of new gTLDs, and that the National Science Foundation similarly has no authority to override ordinary U.S. antitrust rules applicable to access to essential facilities like the root servers maintained by NSI. Therefore, it is crucial that NTIA delete the third of the fourth functions it proposes be delegated to the "new corporation", namely to determine "the circumstances under which new top-level domains are added to the root system." Consistent with the Green Paper's preference for market-based solutions and bottom-up decision making, the NTIA should instead conclude that -- in the absence of an affirmative showing by the new corporation that technical considerations warrant a limit on gTLDs -- all gTLD registries which meet minimal technical standards must be added to the root system.
By eliminating any restriction on gTLDs and permitting future limitations only based on documented technical constraints, the Green Paper would meet its professed principles of introducing DNS competition as fast as possible while preserving consensus-driven, technically-objective Internet decision making. DNS administration should be as open, objective and competitively neutral as is the IETF RFC process of Internet standards-setting. The only way to achieve this goal is to eliminate the political and competitive incentives inherent in today's NSI monopoly of the "dot" while making gTLD competition a matter of purely technical, not pseudo "policy", consideration. After that, the marketplace itself would determine -- as a result of consumer demand -- which gTLDs are desirable and which registries, and registrars, will succeed commercially.
I. THE NEED FOR UNIVERSAL RESOLVABILITY OF DOMAIN NAMES REQUIRES, AS NTIA PROPOSES, THAT ROOT SERVER FUNCTIONS BE COORDINATED
Universal resolvability can be achieved while allowing for competition among domain name registries, thereby benefiting both the content providers and the end-users of the Internet. Likewise, the true value of the Internet can only be realized if all addresses are universally resolvable. That is to say, pgMedia agrees with the Department that the value attached to having a network of networks is maximized only when the root server network is properly coordinated allowing any user of the Internet to seamlessly access any site on the Internet using that site's domain name. To achieve universal resolvability, domain names must be unique and the protocols used must be standardized.
A. The Internet Root Server (or The "Dot") Functions As The "Traffic Cop" of the Internet, Allowing All Name Service Providers to Look to Authoritative Servers for the "Zone" Files Necessary For a Seamless Internet
The "dot" file is a plain-text file used by the nameserver software to define the available top level domains ("TLDs") to the rest of the network. Like all other "zone" files it contains the hostname to IP address mapping that the nameserver uses to find and connect computers over the network. Specifically, the "dot" file lists each available TLD and the names and addresses of the various nameservers that carry the Second-Level-Domain ("SLD") listings under a particular TLD. The ROOT-SERVERS that run the "dot" file direct requests to the appropriate TLD server which sends the request on to the Internet service provider ("ISP") or network managing the listed SLD that then returns the IP address for the requested domain name. So, for example, a request for "PGMEDIA.NET" would first be directed to the "dot" file that would return the locations of the TLD server for ".NET." The request would then be forwarded to the ".NET" server which would return the location of the nameserver for the SLD server for "PGMEDIA.NET" which contains the name to IP number mapping for the pgMedia domain.
Universal resolvability of TLD's requires that there exist an authoritative file that will direct requests to the appropriate server for any particular domain name request. There is more than one possible solution to achieving this goal. The Department's Green Paper suggests that the "dot" file be controlled by one entity. However, due to recent improvements in network administration technology, it is no longer required that a single entity, alone, manage the "dot." Decentralized and synchronized administration of the "dot" file would be a far better solution so long as there is free and open access by registries to the "dot" file. Under this model, the goal of competition in the domain name registration market can be achieved while true universal resolvability can be maintained.
B. Root Server Functions For gTLDs Are Currently Monopolized By NSI, Which Has Led to Significant Competitive Disparities
NSI controls and profits from the administration of the gTLDs (i.e. com, net, org, edu, int). NSI also controls the "dot" that would permit a gTLD to be universally resolvable. The "dot" file is an essential facility because the only alternative available to a registry wishing to compete with NSI, without access to the "dot" file, would be to create a new Internet. pgMedia, through its name.space(SM) division has created a network of 13 nameservers in seven countries. name.space(SM) currently offers registrations in more than 530 TLDs, these names are fully resolvable by the name.space(SM) nameservers, but are not universally resolvable on the Internet because NSI has refused to make the simple text edit to the "dot" file that would allow that universal resolvability. pgMedia supports the Department's position that coordination of root servers does not require a single "root," but rather consistent technical standards for interoperability and symmetry of all servers so as to maintain universal connectivity on the Internet. Operators of "dot" servers merely need to agree to run exact versions of the "dot" file and to implement a secure system to keep the contents synchronized whenever it is updated.
SINDI(TM), developed by pgMedia, is software technology designed to fulfill the function of keeping the name data synchronized in near real-time with security and authentication built in. This technology allows multiple registries to simultaneously submit unique SLDs to any TLD based on a first come, first served basis, and to easily survey the root file for existing TLDs. Through deployment of technologies like SINDI (TM), there is no need for monopoly control of the root server functions to continue to be awarded to NSI. Rather, registries around the world would be free to amend the root file based on market demand.
C. Transferring Control Of Root Servers To An Independent, Non-Profit Corporation, With Wide Representation Is A Valid Approach To Root Server Administration
pgMedia supports any solution that would open the domain name registration market to competition, so long as there is a non-discriminatory and simple procedure for adding new TLDs. As discussed above, there no longer exists an explicit technical need for the root file to be managed by a single entity. However, more importantly, there are serious and pressing problems associated with allowing the root server to be controlled by a private entity with a vested interest. NTIA's proposal to end NSI's monopoly control of the root server through creation of a "private, not-for-profit corporation . . . to manage the coordinated function in a stable and open institutional framework," is right on target. The root server is an essential facility, access to which is required for participation in the domain name business. The Department's proposal would transfer control of that essential facility to a non-profit organization created to serve the best interests of the entire Internet community, not just one company or set of companies.
pgMedia agrees wholeheartedly with the Department that there is no longer a need to propagate a domain name monopoly and the resultant monopoly rents paid by Internet users every day. In moving away from a privately controlled arrangement, an independent, non-profit corporation will remove the commercial conflict of interest that has been present to date. As a matter of policy, no registry should also control the root server, and NTIA's proposal to separate the two is admirable.
II. gTLDs ARE A PUBLIC DOMAIN RESOURCE WHICH MUST BE SHARED AMONG COMPETING REGISTRIES
pgMedia believes that generic-TLDs should be shared among the competing registries. gTLDs are not the property of any entity, including NSI. The technology exists today to allow such sharing and there is no public policy reason why such sharing should be restricted. Although there are stakeholders who advocate a branded model for TLDs, such a solution only benefits the registries that would then have monopoly control over that particular TLD. Rather, Internet users would benefit the most from an open system of TLDs whereby each registry could register new domain names under any existing or new TLDs.
A. NTIA Correctly Adopted pgMedia's Approach of "Shared" Registries, Under Which All gTLDs, Including Existing ".com" Registry, Will Be Open To Competitive Registration of Second-Level Domains.
pgMedia believes that all registries should compete on the basis of service and price, thereby benefiting the users of the Internet rather than developing a system of branded TLDs which would only benefit the registries by creating an artificial shortage in the generic name space. NTIA correctly recognized the importance of competition in the domain name registration industry in adopting a "shared registries" concept in its proposed rule. pgMedia urges the Department to retain this fundamental improvement in any further rule.
pgMedia has long supported shared gTLDs, and believes that gTLDs are a public resource that should not be within the commercial or competitive control of any one entity. The Department should resist creating new monopolies (in each new TLD) by giving a single registry the sole right to register under any one TLD. Such a proprietary system would be akin to geographic market allocation--each registry would own a series of TLDs over which it could exercise complete price and service control. To the extent that certain TLDs became recognized as the market norm for an industry or organization, as is the case currently with .com, users in need of domain names under those TLDs would be forced to deal at whatever price demanded. An open, market-driven system on the other hand would require different registries to introduce efficiencies, innovation and price reduction into the registration services market in order to compete and survive.
name.space has already made numerous achievements and improvements in registration services, even though its own registry does not currently enjoy the market benefits of universal resolvability. These include:
- Instant, form based registration
- Virtually instant updating of the zone files
- Address portability
- "Smart" Whois ("sWhois") -- a software application that allows cross-registry look-up capabilities
- Secure Internet Name Data Integrator (SINDI)
- Lower prices for users
These advantages are available today via the name.space (SM) system. Competitive gTLDs are a technologically feasible reality, as demonstrated by the over 500 gTLDs operated by name.space (SM). Shared gTLD registration is also technically possible and competitively preferable. Although companies like pgMedia have been able to make significant technological break-throughs in DNS administration despite the existence of a blanketing monopolist, it is simply indisputable that true competition in the domain name registry industry would result in even greater technological improvements, leading to better services and lower prices for consumers.
III. NTIA HAS INAPPROPRIATELY LIMITED THE SCOPE OF DNS COMPETITION OUT OF FALSE, UNSUBSTANTIATED CONCERNS REGARDING THE TECHNICAL STABILITY OF THE INTERNET
The Department proposed a transitional limitation on the addition of new gTLDs to a total fiveNTIA's basis for this decision was (1) concern regarding the stability of the Internet (2) the potential impact of more new gTLDs on the trademark dispute process, and (3) fears of exacerbated consumer confusion. The Department was wrong on all three issues.
A. There Are No Technical Restrictions to Increasing the Number of gTLDs
There are no technical constraints on the number of gTLDs that can operate simultaneously on the Internet, and absolutely no basis for any concern that expansion of the TLD name space will contribute even to transitional instability of the Internet. NSI readily agrees: NSI believes that there is no technical reason why the number of worldwide TLDs needs to be limited. Domains are equally scalable at all levels of the DNS. There can be as many TLDs as there are potential second-and third-level domains.
NSI Comments (August 18, 1998) at C-10.
As NSI recognized, Dr. Paul Mockepetris, the inventor of the Domain Name System ("DNS") is equally convinced that there exists no technical reason for limiting the number of new TLDs. "DNS is highly scalable. There is no technical limit to the number of new top-level names that could be introduced. The original designer of DNS, Paul Mockapetris, has verified the scalability of DNS. As well, America Online's eight million-plus third-level domains are also evidence of DNS scalability. There may be some practical and operational issues to the introduction of new TLDs, but we should be very suspicious of those that seek to control, limit and administer TLDs under the false pretense of technical feasibility."
Internet Domain Name System: Myths and Facts, NSI 6/18/97
pgMedia has already proven that such expansion of the TLD namespace is possible. The name.space registry (SM) has been servicing many hundreds of new TLDs for over one and one half years with no problem or interruption in service. The same cannot be said for NSI who's antiquated update procedures led to the catastrophic failure of DNS last July.
The technical explanation for the feasibility of TLD expansion is simple. Behind the current "root.zone" or "dot" file, which lists the currently recognized top level domains, is simply a 75K text file; updating the file for new gTLDs istechnically trivial and transparent to the end users--new top level namespaces will simply become available and functional just as any new country code or new second level domain becomes transparently available as changes are made to the DNS listings on a daily basis without interruption of service or other noticeable change.
For example, the current zone file for ".com" contains over 1.5 million entries and can scale indefinitely. The structure of all zonefiles is identical and used in the same way by the software. Thus, structurally, the "root.zone" file and the "com" file are identical and the number of possible top level domains is equally as scalable as is any other zone file in the existing DNS system. Therefore, to the extent the DNS can handle over 1.5 million registrations under the TLD ".com," there is no reason it cannot handle as many new gTLDs.
The record before the Department from the August comments does not support any technical rationale substantiating NTIA's conclusion that the number of gTLDs in the short run "should not be so large as to destabilize the Internet." As stated above, there is literally as much of a technical basis to limit the total number of gTLDs as there is to limit the number of entries under the ".com" TLD.
Hence, there is no justification for NTIA's proposals that:
- There be a completely arbitrary limit of five new gTLDs betweennow and 9/98
- The new non-profit corporation should set policies for determining "the circumstances under which new top-level domains are added to the root system"
Any gTLD administered by a technically competent registry should be included in the root server system. At most, any non-profit corporation established by the Department should set minimum technical criteria for gTLD registries, per existing RFCs covering name service.
B. Trademark Disputes Do Not Involve DNS Registries, and Do Not Provide a Technical Impediment to Allowing New gTLDs
pgMedia agrees with the Department's characterization that "trademark disputes arise very rarely on the Internet today," and that "management of the Internet must respond to the needs of the Internet community as a whole, and not trademark owners exclusively."
However, pgMedia disagrees with the Department's determination that increasing the number of gTLDs would make it more difficult for trademark holders to protect their marks. The addition of new domain names would have zero impact on trademark prosecution. The addition of new TLDs would not slow the search of domain name databases for trademark violations any more than the addition of new domain names under an existing TLD would. Any claim to the contrary is merely an effort to drag TLD registries into a fight that belongs solely to trademark owners.
As is the case now, trademark owners should continue to be required to police and prosecute their rights under their own devices. There is no other communications medium where those involved in providing the communication are responsible for participating in trademark policing or prosecution. The proliferation of other communications mediums, such as television or radio, has never been slowed or diluted to allow trademark owners an easier time of protecting their rights. In fact, it is the ubiquity of the medium, in this case the addition of thousands of new gTLDs, that gives mark owners the opportunity to add national and international value to their brand names. Rather than complain about this windfall, and the offsetting costs associated with tracking the worldwide exposure to their brand, trademark owners should be encouraging the immediate transition to an open gTLD system.
C. The Addition of New gTLDs Will Not Impact Consumer Confusion
Even with a limited number of gTLDs, there currently exist an infinite number of lower level domain name combinations that can be registered and employed by interested parties. Registrants are free to use virtually any second, third, fourth, etc., level domain name, so long as it has not been previously registered. It is therefore inconceivable that the addition of new gTLDs will somehow make consumer use of the Internet more difficult to employ. There is no substantive difference between the existing possibilities for confusion associated with an unlimited number of lower level domain names (flowers.com vs. florists.com vs. flowers.foryou.com) and those that might be found under a rubric of varied TLDs (flowers.com vs. flowers.delivery).
Rather, the recognition of specific domain names will be a function of advertisement and public awareness, as it is now. A few short years ago, few consumers knew what an Internet address was. Today, many, if not most, could not even name the five basic gTLDs, or even realize that there are gTLDs available for every country in the world. How can the addition new gTLDs be any more confusing than the existing fact that .za refers to South Africa?
Despite the confusion associated with the existing system, through advertising and repeated use, consumers are becoming more and more comfortable maneuvering on the web and locating the destinations of their choice. So too, in a world of many gTLDs, hosts will advertise and consumers will browse, and means for reaching each other will evolve. To the extent such a transition may cause some initial disruption, it is far better that it occur earlier in the growth of the Internet than later, when habits and customs will be more entrenched. More importantly, any minimal confusion that might occur in a multi-TLD environment that doesn't occur now, is more than offset by the advantages to consumers that true competition in the domain name industry will bring.
IV. NEITHER THE UNITED STATES GOVERNMENT NOR THE DEPARTMENT OF COMMERCE HAS THE LEGAL AUTHORITY TO RESTRICT THE NUMBER OF gTLDs
There exists no statutory or other authority that would permit the Department of Commerce or the United States Government to limit the number of new TLDs. If anything, Congress has expressed a clear interest in leaving the Internet unregulated.
It is the policy of the United States-
(2) to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation. 47 U.S.C. section 230(b)(2) (emphasis supplied).
While the Telecommunications Act of 1996 did not specifically address the domain name issue, it is clear that the Congresses' twin goals of deregulation and new competition would be severely undermined if the Department were to propagate some sort of "transitional" regulation regarding the number of new gTLDs that may be added to the root file.
pgMedia is sympathetic with NTIA's concern regarding coordination of essential Internet functions, but the agency, and for that matter, the entire US government, lacks the power to control the number of gTLDs, even in the short run. gTLDs are not a United States', but rather a global resource. The U.S. lacks the authority, under the U.N. charter and customary international law, to unilaterally assert extra-territorial jurisdiction by restricting the number of gTLDs. As policy matter, exclusion of EU and other nations from policy process could lead to costly fragmentation of the Internet community.
Historically, NSI has operated root servers pursuant to a "cooperative agreement" contract with the National Science Foundation ("NSF"), but neither that contract, nor NSF's statutory charter, grant either NSI or NSF the legal authority to restrict gTLDs and exclude competitors from the existing root server system. NSF Cooperative Agreement No. NCR-9218742. However, that is exactly what is currently happening, and, unfortunately, is what the Department proposed in the Green Paper when it suggested that only five gTLDs be added to the DNS.
The cooperative agreement between Network Solutions and The National Science Foundation leaves NSI with the power to implement new naming schemes or systems without seeking the approval of the NSF. NSI has improperly used this NSF agreement to perpetuate its monopoly and to prevent competition so as to continue its substantial profits derived from domain name registrations. According to NSI's own figures there are some 120,000 new domain name registrations per month equaling $ 12,000,000 in gross revenues per month.
pgMedia believes that the arguments advanced by NSI have been calculated to protect this lucrative monopoly position for as long as possible. By restricting the creation of new TLDs the Department of Commerce would be an unwilling accomplice in this bilking of the Internet community.
Congress has not granted either the Department of Commerce or NTIA the authority to constrain competition in the Internet. Congress created NTIA for the purpose of managing federal government's radio spectrum and advise the President on telecommunications policy. 47 U.S.C. section 902 (b). In addition, NTIA is given authority to "conduct studies and make recommendations concerning the impact of the convergence of computer and communications technology." 47 U.S.C. section 902 (b)(2)(M). Neither this nor NTIA's authority regarding telecommunications policies give it the right to restrain competition by restricting the number of gTLDs that may be added to the DNS. Similarly, the general regulatory authority granted to the Department of Commerce by Congress neither directly grants the Department regulatory authority over the Internet, nor expressly grants the Department the right to restrict competition. Therefore, any rule promulgated by the NTIA limiting the creation of new gTLDs would be ultra-vires.
VI. ANY LIMITATION ON THE NUMBER OF NEW gTLDs, BY EITHER THE NSF, THE DEPARTMENT OR ANY OTHER GOVERNMENTAL AGENCY, THAT IS NOT BASED ON TECHNICAL NECESSITY, WOULD BE VIOLATIVE OF THE FIRST AMENDMENT OF THE UNITED STATES CONSTITUTION
The Internet is "a unique and wholly new medium of worldwide human communication," Reno v. ACLU, 117 S.Ct. 2329, 2334 (1997), that "represent[s] an extraordinary advance in the availability of educational and informational resources to our citizens," 47 U.S.C.A. section 230(a)(1). As such, the Internet as a method of communication receives heightened protection under the First Amendment from restriction of the content it transmits. Reno, 117 S.Ct. 2329, 2334.
And, a domain name is more then the address for a site on the Internet. The domain name may be imbued with an idea or indication of the material to be found at the particular Internet site. pgMedia agrees with the Domain Name Rights Coalition ("DNRC"), that, at a minimum, "domain names are the identifiers used to designate and locate expression and communication." Limiting the number of TLDs limits the way in which content providers on the Internet can express these ideas. In the name.space (SM) system a candidate for political office may register their name in the following fashion "gore.for.president;" while in the current system, or the system proposed in the "Green Paper," this simple form of political speech would be impossible.
It is well settled law that for the government to limit a fundamental right, such as speech, it must first demonstrate 1) a compelling governmental interest that would be served by the regulation, and 2) that there exists no less restrictive method for achieving the governmental objective. See Sable Communications, Inc. v. FCC, 492 U.S. 115, 126 (1989). Absent any technical constraints on the expansion of the top-level name space, there is no compelling governmental interest that would be served by preventing pgMedia from offering domain name registrations under its expanded list of gTLDs. Moreover, even if there were a compelling governmental objective to be served by regulating the top-level name space, limiting the expansion to only five TLDs would by no means be the least restrictive alternative available. Therefore, any attempt by the Department to limit the expansion of the top-level name space would violate the First Amendment.
For the reasons stated above, pgMedia urges the Department to (1) abandon any effort to restrict the number of gTLDs that may be added to the DNS; (2) continue to support the creation of a non-profit organization to serve as neutral administrator of the root server, thereby separating the root administration services from registry services; and (3) continue to support a policy of shared rather than proprietary gTLDs, whereby true competition can work to lower domain name prices and improve services.
Paul Garrin, President
11 East 4th Street, 2F
New York, NY 10003