Anyone who dabbles in demonology will tell you that to know a thing's true name is to gain power over it. More importantly, anyone who trifles with trademarks will attest to the importance of defending the identity of your product or service.
That issue lies at the heart of the International Ad-Hoc Committee's February 4, 1997 Final Report which recommends that seven new top-level domains be added to the existing .COM, .EDU, .GOV, .INT, .MIL and .ORG and ISO 3166 two-letter country code domains. The IAHC-proposed new domains:
These domains are designed to be only the first wave of what, by the turn of the century, is envisioned to be a hundred or more new top-level domains, each offered by hundreds of competing registrars.
The IAHC proposal, like the IAHC itself (www.iahc.org), had its genesis in the National Science Foundation's March, 1992 announcement that it would begin outsourcing domain name and IP number registration. On January 1, 1993, it awarded a contract to run the Internet Network Information Center (InterNIC) to Network Solutions, Inc. (NSI) of Herndon, Virginia. That contract runs through September, 1998 and, among other things, gives NSI the exclusive right to provide second-level registrations in the .COM, .EDU and .ORG domains (the only non-ISO 3166 domains for which there is a commercial market).
In June, 1995, NSI (then recently-purchased by Science Applications International Corporation, a $2 billion consulting firm and defense contractor) announced that it would institute a policy of suspending the registration of any domain name whose validity was challenged on trademark infringement grounds. Although domain registration like IP network assignments had always been free to end users, in September 1995, at the direction of the NSF, NSI announced that it would begin charging $50 per year for registrations in all but the .GOV, .INT and .MIL domains. (thirty percent of the money collected was to go into a trust fund to be used for infrastructure expenditures "for the good of the Internet as a whole," although the mechanism by which disbursements would be made was never satisfactorily defined and, as of April 2, 1997, no distributions have ever been made.)
Reaction to these changes was immediate and vocal. Much of it took place on the newdom, com-priv, ietf and domain-policy mailing lists and on the comp.protocols.tcp-ip.domains newsgroup. Criticism centered on the increasing desirability of creating effective competition for the InterNIC and NSI and on the stupidity of NSI's trademark resolution policy.
The trademark issue is a subtle one. Few business people (and fewer ISPs) understand that trademark validity is linked both to particular products or services offered and, in most cases, to the particular geographical area in which those products or services are offered. Thus, for instance, Saturn automobiles and the Saturn game console are each trademarked internationally and neither infringes on the other because no reasonable person is likely to confuse the one with the other. The problem arises because all commercial registrations are currently limited to the .COM domain and only one company can own SATURN .COM. Under NSI's existing dispute resolution policy, should either Saturn maker challenge the other's existing registration, that registration would automatically go on hold until the dispute was resolved in a court of law or via private contract. (In reality, SATURN.COM is, in fact, on hold, and Saturn Technologies Inc. of Chalford, PA, which registered it in May 1994, has lost its 3-year-old web site until the dispute is resolved.)
The debate continued, as Internet debates will, with no resolution in sight until January, 1996, when Paul Garrin launched his name.space alternative registry, (easy enough to do, since the BIND database is just a text file,) with nameservers in five different countries. In April, 1996, Eugene Kashpureff set up his own nameservers, added .EXP .LTD .LNX .MED .NIC and .XXX to the standard top-level domains and began doing business as the AlterNIC. Even though Garrin's and Kashpureff's new domains were only visible to a tiny fraction of the Internet population, (those who set name.space or the AlterNIC as their primary or secondary nameserver), and even though they charged money to register in some of them, both found customers for their offerings.
The following month, Jon Postel, Director of the Internet Assigned Numbers Authority, issued a proposal for multiple, exclusive, competing top-level domain name registries he called draft-postel-iana-itld-admin-00.txt. It garnered a lot of reaction, some of it positive, much of it negative. In particular, the proposal's suggested $100,000 fee for each registry was heavily criticized.
At the June 24-25, 1996 annual meeting of the Internet Society Board of Trustees, Postel presented a revised draft for their approval. Draft-postel-iana-itld-admin-01.txt exchanged the $100,000 per registry flat fee for a $2,000 per year fee and 2 percent of revenues. It kept the original draft's proposal that up to 50 new registries be created initially, each of which would have the exclusive right to register names in up to three new top-level domains (i.e._150 new domains spread among 50 registrars the first year). It also proposed that additional new registrars be added at a rate of 30 per year for a period of five years.
ISOC's Board approved Postel's revised draft in principle in its Resolution 96-05:
This was made evident at a July 31, 1996 meeting between Bill Manning, representing IANA and various interested parties, including Christopher Ambler and his financial backer John Frangie. The meeting was to discuss the criteria by which applicants to run new registries might be evaluated by the ad-hoc committee to oversee the registration process. Afterward, Ambler handed Manning a $1,000 check sealed in an envelope. Ambler maintains that it was clear to Manning that the IODesign check was for fees to be paid to IANA for its formal application to run a registry and that Manning watched him put the check in the envelope. Manning strongly disputes Ambler's account, claiming instead that the envelope was "stacked in to a manila folder" along with notes on the meeting and that he did not discover it until after Ambler and Frangie had left. Ambler does not dispute that Manning returned the check, unopened, to him via certified mail the following day. Postel also put out a widely-disseminated e-mail on August 2, stating, "The suggestion that the IANA is accepting money to reserve new top-level domain (sic) is completely false."
POP GOES THE IAHC
On October 22, 1996, Donald Heath, President of ISOC, issued a press release announcing ISOC's intent to form an International Ad-Hoc Commit-tee to "undertake defining, investigating, and resolving issues" raised by the second Postel draft. Heath promised that the Committee would be made up of "representatives of the large international Internet community." Less than a month later, on November 12, Heath issued a second press release announcing the IAHC's membership.
The Committee was made up of:
Just over a month after its formation the IAHC released a Draft Specification for Administration and Management of gTLDs. Dave Crocker attributes its aggressive schedule to pressure from unnamed "large ISPs" and certain, mostly unidentified "powerful individuals," specifically including Paul Vixie, maintainer of the Berkeley Internet Name Domain (BIND) software and its standard cache file. Regardless of the sources of the pressure to produce rapid conclusions, it is clear that fear of further and irreparable fragmentation of the root name space drove the process.
The Draft Specification completely reworked the Postel proposal. Instead of an initial 150 new domains, it proposed a mere seven. In place of exclusive registration rights, all new registrars would share registration authority for all seven new domains. The existing TLDs would be excluded and would, at least for the remainder of its contract with the NSF, remain the exclusive province of NSI's InterNIC. The initial group of new registrars would be chosen by lottery from among the pool of "qualified" applicants. A $20,000 partially-refundable application fee would be required. No special consideration would be given to existing operators of "experimental" registries. A Council of Registrars (CORE) would formulate the "necessary contractual, legal, oversight and public policy framework under which CORE and the individual Registrars must operate" and a Board of Trustees would oversee CORE. And all new domain registration requests would be subject to a mandatory 60-day waiting period to permit trademark-related challenges to emerge before the name was activated.
Needless to say, the December 19 release of the Draft Specification created considerable controversy. The official comment period lasted six weeks and, in that time, the IAHC's own mailing list received over 4,000 postings. Other parties submitted commentary and criticism via private e-mail, fax, surface post and telephone calls.
Much of the criticism centered on the closed and relatively secretive nature of the IAHC's deliberations, and on the aggressive nature of its schedule. Some made threats of lawsuits and predictions of dire consequences for the future of the Internet's stability if the Draft Specification were adopted. Others took issue with particular conclusions or mechanisms outlined in the Draft Specification, including the proposed applicant selection process and the screening criteria (especially the financial criteria). Still others, particularly Aveek Datta of Monolith Internet Services International, complained that the Draft Proposal failed to provide a mechanism for free registrations for individuals and non-profit entities. And, with the exception of the trademark lawyers on the IAHC, almost everyone hated the 60-day waiting period and demanded it be deleted from the final document.
HORTON HEARS THE IAHC
The IAHC reconvened in Switzerland in late January, 1997 to mull over the mass of input on the Draft Specification and produce from it a Final Report. That document, The Final Report of the IAHC: Recommendations for Administration and Management of gTLDs, was released four days late, just before midnight on February 4. It differed substantially from the Draft Specification.
It replaced the Draft Specification's Board of Trustees with a Policy Oversight Committee (POC) and made CORE a Swiss-chartered not-for-profit association. The POC was given the power to police the registrars, including the ability to remove them for cause. It was also charged with authorizing additional registrars and additional TLDs and granted authority to modify the minimum financial and business requirements for registrar applicants as circumstances warrant. The POC was to be made up of representatives from the same bodies which constituted the IAHC (and in the same numbers) plus two representatives from CORE and a non-voting member representing the International Telecommunications Union (ITU), which was to act as the Depository for Memoranda of Understanding (MOU). The MOUs, once signed, would create CORE, the POC and a new body, the Policy Advisory Board, which was chartered to provide the POC with input regarding general policy direction for gTLD additions, the oversight of CORE and any amendments to the MOUs.
The Draft Specification's mandatory 60-day waiting period was gone, replaced by a voluntary waiting period for those determined to provide themselves with maximum protection against trademark-related challenges to their domain name ownership. In addition, the Final Report mandated that all domain name applicants agree to binding arbitration of any ownership disputes and awarded jurisdiction over that arbitration to Swiss-based Administrative Domain Name Challenge Panels (ACPs) to be administered by WIPO.
There would be 28 new registrars in the initial group and they would be distributed evenly across 7 geographical regions:
Any lottery would be conducted on a regional basis, with no lottery taking place in any region with 4 or fewer applicants. It also set up different minimum business and financial criteria for contested and non- contested regions:
It required that all applicants be located in WTO member countries or in countries which were signatories to the Paris Convention for the Protection of Indus-trial Property. It required that applicants own an existing, reachable second-level domain and that the DNS record for that domain be error-free. It kept the Draft Specification's $20,000 refundable application fee and added a requirement that the applicant pay the non-refundable cost of a third-party credit and business reference check.
The Final Report also set some fairly stringent operational, technical and administrative criteria for applicants. Among other things, it required that they have staff members with demonstrated experience in DNS and router configuration, SQL programming and database administration and UNIX sysadmin duties. It demanded they possess a 24 x 7 x 365 help desk and, prompted no doubt by the legendary unreachability of the InterNIC, an automated call distribution system. It also mandated an applicant have multihomed or distributed Internet connectivity and what it called "robust" backup and disaster recovery procedures.
Unsurprisingly, the Final Report was roundly slagged by the alternative registry community. Most complained that it gave no special status to operators of existing experimental registries. Others protested the high bars to entry for smaller businesses which the financial and business criteria presented. Still others abominated the mandatory binding arbitration requirement, viewing it as an infringement on their right to file suit. Aveek Datta refused to support it because it still made no provision for free registration for individuals and non-profits. And, inevitably, some characterized it as an attempt to destroy the alternative registries.
On February 27, after weeks of fruitless threats and pleas, IODesign filed suit in San Luis Obispo Superior Court against IANA, the IAHC, ISOC, Jon Postel, Bill Manning, Joyce Reynolds, Nehal Bhau (all IAHC employees) and Donald Heath (in both his roles), and 400 unnamed John Does. The complaint alleged breach of implied contract, restraint of trade, unfair trade practices and unfair competition, as well as detrimental reliance, intentional interference with a contract and with prospective contractual advantage and trade libel. It requested declatory relief, injunction, damages and treble damages against the defendants.
Meanwhile, on March 3, most of the alternative registries banded together at an Atlanta organizational meeting to create the Enhanced DNS (eDNS) specification. This was an attempt to standardize their sundry proprietary DNS database extensions and to create a mechanism for the establishment of an alternative global DNS root nameservice.
Then, on March 20, PG Media, doing business as name.space, filed suit against NSI in the Southern New York U.S. District Court, naming IANA, ISOC, the IAHC and Jon Postel as co-conspirators (but not including them as defendants). PG alleged multiple violations of the Sherman Antitrust Act, as well as violations of New York State's Donnelly Act and sought an injunction requiring NSI to amend the root configuration file to include the name.space domains, damages of not less than $1,000,000, triple damages and costs.
SCORING AT HOME
As of this writing, the IAHC members are still hard at work, generating the relevant MOUs to accompany their Final Report. If all the relevant stakeholders sign those MOUs, a call for registrar applicants will go out seven days after the final signature. The application period for aspiring registrars will last another 60 days, to be followed by a 30-day evaluation period, during which the candidates will be audited and investigated to determine their eligibility. At the end of the evaluation period, a lottery will be held for all contested regions and the results will be announced the following day. Assuming that all the winners can meet that early, they will all sign the CORE MOU one day later and the IAHC will mutate into the POC the day after that.
Some indeterminable time later, the members of CORE will have solved the problem of sharing the central registry database in a secure, robust implementation and the new domains will open up to public registration.
What will happen to .COM, .EDU and the other, familiar TLDs after NSI's contract expires in September, 1998, is still very much an open question.
In the meantime, neither IODesign's nor PG Media's request for preliminary injunctive relief has been granted as of April 2. Both trials are liable to be a long while in coming, assuming that the plaintiffs pursue their suits.
Leveraging the publicity medium of the Internet itself, the eDNS supporters have managed to attract a fair amount of attention. Unfortunately for them, even their own surveys show that, more than a year after the formation of both name.space and the AlterNIC, well less than 1 percent of the total number of nameservers across the Internet can resolve their non-standard domains. Worse still, from their perspective, neither the InterNIC root nameservers nor the nameservers of most large ISPs will resolve them, so the number of Internauts who can easily access those domains is considerably smaller than 1 percent of the Internet's population. It looks grim for the eDNS community.
Maybe that's not a bad thing. Certainly, the prospect of a fragmented root is a frightening one. Should eDNS catch on, that would be the most likely outcome, barring the highly-unlikely event of a full retreat by the IAHC and a complete surrender by NSI.
WHAT DOES IT ALL MEAN?
When the IAHC announces it will begin taking applications for new registrars, many ISPs are likely to find that they qualify, should they care to apply. Those who don't make the cut off for the initial round won't get the benefit of publicity that the first group of new registrars will enjoy, but they also won't have to deal with the technical headaches of creating and testing working shared-registry code. And there will always be later rounds for those who think that being a registrar is likely to be either a license to print money (unlikely and becoming more unlikely as more competitors join the market) or as a useful adjunct to their businesses.
For those who plan to stay out of the market, the technical, administrative and support burden will be relatively modest. As new domains are added to the root, you'll need to download the latest BIND cache file and restart your name server. You'll have to update your user support pages to reflect the choice of registrars. You may need to re author or replace a FAQ or two. No extraordinary effort will be required.
The larger implications are the more interesting ones. Firstly, although everybody talks about LDAP (and nobody besides Novell does anything substantive about it) there is a real possibility that the expanded top- level namespace will evolve into an Internet-wide directory service over the next few years. As the number of new gTLDs expands into triple digits, there is every likelihood that it will come more and more to resemble a Yahoo!-style hierarchical directory and that those domain name holders with the appropriate affinity will voluntarily locate themselves in the most intuitively-suitable gTLD.
Secondly, the IAHC process points toward the need for the Internet community to anticipate and preempt the attempts of national, regional or local governments to impose governance structures on the Net by ourselves creating those structures in advance of need. We can have jurisdiction forced on us, or we can establish it for ourselves and thereby make it responsive to our needs, rather than those of technology- impaired politicians and jurists.
To name a thing is to gain power over it.
Ask any demonologist.
LATE BREAKING NEWS!
80 organizations signed the Generic Top-Level Domain Memorandum of Understanding (gTLD-MoU) in Geneva on May 1, 1997. Also on May 1, the IAHC was formally dissolved as an organization, to be replaced by the intial Policy Oversight Committee (POC), consisting of:
On May 6, 1997, the widely-hated lottery for the intial registrar selection process was dropped from the proposal.
On May 29, 1997, draft copies of the CORE Articles of Association, the CORE Memorandum of Understanding and the first draft of an application to become a gTLD Registrar were posted on the gTLD-MoU Website.
More information as it develops.
(Copyright© 1997 by Thom Stark--all rights reserved)