Way back in 2000-2001, ICANN approved a handful of new top level domains, and entered into agreements with their promoters. Three of the sponsored domains, are coming up for renewal next year, so they’ve sent in their renewal proposals. A sponsored domain is one that restricts who can register to members of a particular community, in this case respectively co-ops, museums, and the airline industry. Let’s take a look and see how they’re doing.

.coop

The first proposal is from .coop, for co-operative organizations. Unlike every other sponsored TLD, one can make an argument that .coop is a success. They have 5,600 registrations, which they claim is far more than any other sTLD. (I should be able to verify this in a few days, assuming the other sTLDs provide access to their zone files like they’re supposed to.) This is a tiny fraction of the total number of co-ops, which according to the original .co-op proposal, is 48,000 in the US and 750,000 worldwide. Many of those are tiny little co-ops that are too small to have any sort of domain, but it still seems pretty lame that with under 1% of the eligible registrants having signed up, they’re still ahead of all the other sTLDs.

In their registration proposal, DotCooperation basically asks to renew the charter as is. They have a few niggling wording changes to clarify who’s eligible to register, and they report that they were not able to comply with the exact terms of the escrow agreement, apparently due to their inability to find an escrow company in England where the registry is that will agree to a contract under ICANN’s California law. Other than that, they’re happy with the way things are.

I attribute the relative success of .coop to the fact that co-ops are by their nature businesses, so the whole domain is run in a businesslike way. It also helps that there are some very large co-operative businesses and a lot of co-operative banks, so raising money was probably not a problem.

The original .co-op proposal projected that by now they would have between 100,000 and 500,000 domains, so they have egregiously failed to execute their business plan. Nonetheless, it’s hard to see any alternative scenario doing better for co-ops. The domain is run by the long standing respectable industry organizations that appear to legitimately represent co-ops around the world, I’ve never heard any grumbling from co-ops about how lousy .coop is, so this is as good as it gets.

.museum

At the time it was initially approved, the overwhelming response to .museum was ``Why bother?’‘, and that question remains unresolved. The renewal proposal starts by telling us how wonderful the museum community is, and how the .museum domain was established just as museums were starting to address digital media, online exhibits, virtual museums, and the like. It then admits that the registration rules are based on the membership rules of ICOM, the International Council of Museums, which are unlikely to be updated until 2007 to admit virtual museums, but when they are, .museum will want to update its charter to match. Then it gets to the meat of the situation:

The quantitative involvement of the dotMuseum target community in the actions described above has to date been significantly less than had been initially estimated. The cost of operating the domain is not met by revenue generated from the sale of names in it, nor is there any expectation that this situation will change in the near term. Although MuseDoma undertook its action in the belief that the domain would become self-sustaining within a foreseeable span of time, the initial prognosis of impending registration volume has not been borne out in actual practice. The business model in which the terms of the domain’s operation were conceived therefore requires fundamental revision.

Several paragraphs later it becomes apparent that they’re blaming it on the registrars. The process of registering in any sTLD is twice as complicated as registering in a normal TLD. First you have to prove you’re eligible by presenting your credentials to the sponsor. Once they’re satisfied, they issue you a codeword called a Community ID (CID) which you present to a registrar along with the rest of the details to get your domain. I did not find this procedure unduly onerous when I got my .aero domain, but apparently it is crushingly daunting for the members of the museum community. They also believe that a lot of the potential registrants are from countries where they don’t speak English, and the registry doesn’t speak whatever they speak, so the registry refers them to a nearby ICOM national affiliate who probably does speak their language to help them explain to the registry how they’re a museum. The applicant also has to negotiate about what name it can register since .museum has well-intentioned but over-complex naming rules intended to make the names look like a directory. Once all that is done, the applicant can get its CID and spend $200 for its two-year registration. It’s not hard to imagine that somewhere in this process it occurs to the applicant that for ten bucks it can register in .org and be done with it.

Their proposed solution is to let museum associations such as the ICOM national affiliates act as limited registrars so they can provide one-stop shopping, and ideally (from .museum’s point of view) hand out names as part of the membership package. This is not a bad idea, but .coop has the same problem and solved it by arranging for national associations to refer applications to friendly registrars, and also arranging to collect the qualification info at the same time as a registration. I see that .museum has done some of the latter, but it could perfectly well do the former. In any event, I have my doubts that .museum will ever get to critical mass. There are less than 500 American registrations representing maybe 200 museums, since a lot of them have multiple domains, e.g. ringling.historic.mansion.museum, ringling.mansions.museum, ringling.art.education.museum, ringling.education.museum, ringling.florida.museum, fsu.ringling.education.museum, ringling.state.museum, ringling.florida.state.museum, ringling.house.museum, and john-ringling.house.museum are all the same place.

The original .mus application forecast between 20,000 and 50,000 domains in five years, which they haven’t come close to meeting, even with all of the duplicate names. The real problem is still ``Why bother?’’ The idea may have been to make the domain a directory of museums, but we’ve learned from Yahoo, AltaVista, and particularly Google that if you want to search for something, there’s better ways than using domain names to do it.

.aero

Finally we turn to the egregiously unsuccessful .aero, which is supposed to be for (depending on who you ask) the airline industry or the air travel community, sponsored by SITA, the industry-owned co-op that originally ran the telex networks and now the computer networks that tie the airlines together. Despite grand plans to make the domain work like a database, reserving all of the two letter domains that correspond to airlines and the three letter domains that correspond to airports, the community hasn’t come on board. They claim over 29,592 domains, but I suspect that about 29,400 of those are the two and three letter reserved names, few of which have been picked up by the corresponding airlines or airports. Early on, .aero was extremely snooty about who they would admit, but by 2004 they were desperate enough that when as an experiment I applied for a domain for my two bit airline ticket web site it took perhaps a day for them to approve it and sell me a domain.

They also blame registrars for their problems:

The recent acquisition of a major .aero accredited registrar (holding the majority of .aero registered domain names) by a third party not authorized to register .aero names illustrated the vulnerability of the model. It has demonstrated that the benefits associated with an established organization acting [as?] a sponsor can easily be lost.

While I can see that this would be a good reason to fire whoever wrote the registrar’s contract (some combination of ICANN and SITA), and it reflects poorly on SITA’s ability to carry out their duties under Attachment 9 of the .aero agreement which tells them to select registrars who will do a good job, it doesn’t sound like a failure of the registrar process.

Also, in the past four years, SITA has been forced to turn down a number of applications from aviation entities in the least developed countries—simply because the current distribution model is unable to accommodate the settlement mechanisms traditionally used in the air transport community.

I don’t understand that at all. When I registered airinfo.aero, Domainbank mailed me a bill, and I sent them a check for $200, about as ordinary a payment scheme as there can be, even if it’s not the way you settle payments for airline tickets. I can’t see either of these as a reason to give up on registrars. The real problem is the same as .museum’s, that the domain will never be big enough for registrars to care about it, so they’ll never be able to find a registrar that will do more than a perfunctory job.

SITA says they want to sell direct and through other organizations as in their 2000 proposal, which said that they would distribute names through industry associations such as IATA and ICAO. (Since IATA is behind .travel, it seems unlikely they’d sell .aero with much enthusiasm.) Maybe it would help, maybe not, since my experience suggests that the difficulty of the registration process is not a bar to anyone who actually wants to register.

A more important question about this renewal is whether SITA’s stewardship of this domain has served the community it’s supposed to serve. Their proposal had an optimistic estimate of 259K names, and a pessimistic one of 94K names, which appear to have been off by about two orders of magnitude, but we’ve seen that’s par for the sTLD course. An important difference between .aero and the other two sTLDs is that SITA does not include more than a small fraction of the community that the domain is supposed to support. If you’re an airline or an airport, you know who SITA is, but for all of the other potential registrants, from pilots and flight instructors to aircraft maintenance companies to two-bit airline media, it’s SITA Who? There is supposed to be a Dot Aero council that oversees .aero, but they include only a fraction of the registrant groups, their deliberations if any have been in secret, and their success in persuading people to use .aero has been non-existent. My advice would be to declare .aero a noble experiment that failed, shut it down, encourage anyone who liked .aero to try .travel and see if it works any better.