On Fri, Mar 06, 2009 at 03:54:58PM -0500, Edward Lewis wrote: > knowledge of IDNs and the BiDi issue. But when I talked with some folks > around the ICANN circus, there was conventional wisdom that there would > be no admission of a delegation beginning or ending with a digit to the > root zone. This bar was in place because of the BiDi issues as I > understand from Patrik Faltstrom's blog. > > And, from what I have heard, I believe "display issues" is at the heart > of the problem.
Ok, so the "no beginning or ending digit" rule is a rule about U-labels, and not necessarily about A-labels. So the outstanding question, relevant to the draft at hand, is whether the issue is important for A-labels as well. Since this has been a topic of some importance in IDNAbis, I'll ask the people over there about it. Patrik is not the only person in that community who's on top of this issue. [Aside: while I'm on the topic of IDNAbis, I've recently again been reminded that it'd be really, really good if people with DNS clue would have a look at that work to see if there's anything problematic.] > Still, seriously, all this aside, I doubt we will ever want "manpages.5" > as a domain name. But I have no difficulty imagining someone with more money than sense wanting .666 as a TLD. And with the latest ICANN policies, they'd have a stronger argument than they once did. The question is whether there are important technical reasons to say no. "666" could never appear as part of an IPv4 address, after all. A -- Andrew Sullivan a...@shinkuro.com Shinkuro, Inc. _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop