I consider this a cop-out. First, it admits to allowing a small number of humorous registrations as well as a number of defensive commercial registrations that may not meet the Eligibility Requirements, but nowhere does it admit to, or directly address, intentional gaming of the system by those such as the prolific Adrian Miles. They also state that [c]learly, some of these examples are not consistent with a rigid interpretation of the Eligibility Requirements. when clearly many of these examples don't fit even a loose interpretation of the Eligibility Requirements.|
Ben Edelman's study called 8% of the names into question. While some of these do fit within GNR's loose interpretation of its Eligibility Requirements as given above, many do not. There are no doubt other names that Ben did not list as he did not filter on those terms. So what is the extent of the .names registered that don't fit even a loose interpretation of the rules? 5%? 10%? I'd suspect more towards, or past, the latter figure, and it is likely to increase now that GNR has copped out.
To say that the DRP can be invoked to deal with this is absurd. First, who is going to pay over $1000 per name to ensure that such names are removed from registration? Second, to put this in context, the UDRP in its approximately two years of operation has heard complaints regarding approximately .01% of registered names, and in the vast majority of those cases the complainant had the additional incentive that they would wind up as the new registrant.
The reasons given for not pre-policing are also suspect. First, as I said earlier, if Ben (a single individual) can do such research using software, why can't GNR do so proactively? Sure, a suspect name in one language may be acceptable in another, one could use additional sorting criteria (such as incoming IP) to further pare down and flag non-conforming names. Only names that failed a series of such tests would need to be flagged for human review. This wouldn't get rid of all non-conforming names, but it might make it enough of a hit and miss proposition that those intent on registering non-conforming .names would give up and go elsewhere. Second, this idea that someone might register, say, dozens of .names for their friends is unlikely. Policing each individual registration as it comes in may well not be economical, but specifically disallowing multiple registrations (let's say, over 5, or 10) unless one goes through a human would also cut down on non-conforming names while remaining economical. Indeed, it would allow for upselling that could well be an additional source of revenue.
Someone intending to game the system might attempt to get around this by registering a few names at a time, going through different registrars, using different identities, different credit cards, but again, if it become difficult enough, it becomes uneconomical to game the system. Or are we to believe that we should throw up our hands in defeat, as GNR has done, and admit that it is economical to game the system and uneconomical to attempt to address that gaming? If that is so, we're in big trouble.
And are we also to believe that those who paid their $50k to ICANN in good faith to be considered for a new TLD should learn from this that one can mislead ICANN about one's intentions and get away with it (Afilias and NeuLevel also aren't blameless in this regard, so this is a disturbing trend)? The only lesson the New TLD Evaluation Task Force, or anyone else, can get out of all this is that to be successful under the ICANN regime at any level it is best to lie. -g