Last week we attempted to explain the origins of the often heated debate that is currently raging about the future administration of the internet’s domain name system (04/09/98). Today, in the second part of our three-part mini-series, we will look at the initial skirmishes among the various players that have led to the political endgame being played out now.

By Nick Patience

After the Internet Assigned Number Authority’s (IANA) director Jon Postel suggested in May 1996 that up to 150 new generic top- level domains should be added to the internet root spread across 50 registries and the proposal had been looked at by the board of the Internet Society (ISOC), IANA and some internet veterans formed the International Ad-Hoc Committee (IAHC), which, in December issued a proposal for adding not 150 but seven new gTLDs (firm, .store, .web, .arts, .rec, .info and .nom). The IAHC plans included one of the longest acronyms, the generic Top-Level Domain Memorandum of Understanding (gTLD-MoU) which is a detailed explanation of the need for more gTLDs and how it would go about creating and managing them (06/04/97). Those who agreed with it were encouraged to sign the MoU and a signing ceremony was held in Geneva, Switzerland on May 1 1997, at which the IAHC dissolved and morphed into the Policy Oversight Committee (POC). The gTLD-MoU also called for the formation of a Council of Registrars (CORE) comprising companies that wanted to register second-level names in the new top-level domains. CORE is a non-profit entity and would be the sole registry for the new gTLDs, while the registrar companies would be free to compete in the open market. Registries are the bodies that maintain the so- called ‘zone files’ for a particular TLD. Those files contain the name of each second-level domain (SLD) (ie, the ‘ibm’ in ibm.com) and the IP address of each SLD’s name server. Registrars are in the business of liaising with users, and submitting requests to the registry. Each registry in CORE had to fork out $10,000 to show they were committed, and to pay towards the cost of building a shared registry system, the contract for which was awarded to Emergent Corp. Applications were open for 90 days between July and October 1997 and some 65 companies signed up. The total now stands at about 80 after the application period was re-opened this year. It could be argued that it was the gTLD-MoU that finally caused a major rift between what could loosely be called the ‘old’ internet users and the new breed of commercial users, who spotted a lucrative source of revenue and profits. Many of the new players objected to the single registry idea of the MoU and argued that there should be competition in both the registry and registrar markets and that should include the breaking up of the monopoly held by Network Solutions Inc (NSI) in the .com, .net and .org name spaces. However, the MoU supporters argued that the registry should be a non-profit monopoly in order to maintain domain name portability; should one CORE registry go bust, the name could be retained and handled by another registry. Competing registries, the argument goes, would not permit this. The gTLD-MoU also includes a dispute resolution mechanism for disputes arising from conflicts between domain names and trademarks, developed by the World Intellectual Property Organization (WIPO) in Geneva. WIPO is one of the bodies with representatives in the POC; the others are ISOC, IANA, CORE, the International Telecommunications Union (ITU), Internet Architecture Board (IAB) and the International Trademark Association (INTA). It was this perceived bias towards ‘old- school’ internet bodies and Geneva-based entities that included a lot of government representatives that irked many of the new, purely commercial internet players.

Temporary monopoly

During June 1997 things seemed fairly quiet while the community absorbed the gTLD-MoU proposals and the likely impact and opposition started to gather momentum. At this point NSI, publicly, at least, was in no mood for sharing the .com, .net and.org registration business with anybody, especially the gTLD- MoU/CORE crowd. Then on July 1 the US government issued a request for comments (RFC) to learn more about the various schemas being bandied around. A few days later NSI announced that it had filed for an initial public offering, no doubt sensing that a monopoly position, albeit a temporary one, was an ample basis for an internet IPO (07/08/97). Even after the IPO, the company is still 80%-owned by Science Applications International Corp (SAIC). However, the filing did reveal a couple of blots on NSI’s landscape, namely an antitrust probe by the Justice Department and a similar case filed against it in March last year by New York-based PG Media Inc. Briefly, in March 1997 PG Media asked NSI to add some 300 new generic top-level domains (gTLDs) to the root, which NSI refused to do, saying that IANA was the only entity that could grant such a request. PG Media said NSI was infringing its first amendment rights as well as denying access to essential facilities, which is an antitrust violation, so it sued to get access to the root, as well for damages. All this time a group operating under the AlterNic banner readied its alternative registry system, establishing a rival root to that controlled by the NSF/NSI contract. On the weekend of July 12, AlterNic’s head, Eugene Kashpureff hacked into the InterNic site and redirected its traffic to AlterNic.net for about 48 hours as a protest against the NSI monopoly. He later publicly apologized to NSI for his actions, but was hunted down by the FBI and Canadian police, spent two months in jail in Toronto and was this year fined a nominal sum and put on probation for his actions. Then on July 17 NSI had one of those days you don’t want while readying an IPO. Human error by one of its operators caused a corruption in the root zone files, which was then duplicated around most of the 13 root server around the world, rendering a great deal of the web and millions of email addresses unusable for a large part of the day on the east coast and all day in some other parts of the world. NSI learned a lot from that and has since tightened up its operations (07/17/98). Around this time some groups began emerging in opposition to the gTLD-MoU position, but also wanting a voice in the future of the DNS, among these were a group sprouting out of the Association of Interactive Media called the open Internet Congress (OIC). Nothing much came of that, but the people behind it went to play a role later on down the line. At the end of July a meeting was held in Washington DC at which many of the major players came together for the first time. It was also the first time NSI said publicly that it was willing to share the registration business in .com, .net and .org (08/06/97). The overall tone of the meeting was that in order to avoid the US government taking over the entire process, the private sector must get its act together and present the feds with a sound proposal, convincing it that it can be left safely in private hands. Meanwhile, the government got a fairly paltry response to its RFC, perhaps reflecting the lack of understanding of the subject or its importance at that time (08/19/97). By deadline time the feds had only received about 400 responses. Back on the PG Media legal beat, it emerged later in 1997 that the NSF had written to NSI twice, in June and August, directing NSI not to add any more gTLDs to the root. So in September 1997, PG Media added NSF’s name to its suit. In November CORE said it was still confident its seven TLDs would be added to the root when it announced its intent to build a shared registry system. And at that point it seemed it had a very good chance of pulling it off, especially with the ISOC team working the Washington corridors of power, having raised $880,000 from registrars to build the registry system, in the hope that it would one day get the chance to use it (05/17/98). Christmas came and went and with CORE still confident of playing a major role, the US government issued its green paper on the administration of the DNS on January 30. Essentially a document to provoke discussion, it certainly achieved that objective. But it also ripped the heart out of CORE’s plan to add seven new TLDs. There should only be at most five new TLDs added for the time being, the paper suggested, with each one being administered by a separate registry. It was clear that CORE, like NSI would now be forced to give a lot of ground. Just how much ground was the issue that would be fought over through the rest of the winter, and into the spring and summer.

The third and final part of this series detailing the endgame politics that were thrown up by the green paper will appear next week.

รก