Here at the Amsterdam meetings of the GNSO, the key issue for the first hours has been how to evaluate proposed strings. Everyone agrees that the acceptability of the applicant's proposed string should be performed up front. You wouldn't want to get all the way to the end of a costly and time-consuming process only to find that the string you proposed wasn't acceptable for some reason (can you say ".iii"?). The harder question is against what criteria you judge the proposed strings. You have social issues, such as .GOD or .XXX. You have trademark issues, such as .ATT or .APPLE. You have geopolitical issues, such as .ASIA or .TIBET. You have technical issues, such as .EXE. And you have issues of potential user confusion, such as .COMM or .INFORM. We're having a difficult time formulating any rules that would take into account all of these potential challenges...and that's before we've added the complexities of IDNs at the TLD level.

Some formulations of a rule are so broad that they would make every possible string subject to challenge. Others are so narrow that they leave legitimate complaints on the sidelines. We're trying to find a balance, and it's not easy.

From a process point of view, we've agreed that the requested strings in new TLD applications should be published as soon as the applications are received for public comment. Now, on what issues should we ask for comment?


Meanwhile, DSL Reports is reporting that Earthlink is running its own version of Sitefinder for those broadband users that use Earthlink's DNS. NANOG has a thread, and the DSL Reports comment board has some conflicting reports on what is actually going on.