On Sat, Mar 07, 2009 at 07:40:46PM +0100, Patrik Fältström wrote: > The problem with writing exact objective rules is that with the 6000 > languages, and enormous number of codepoints, it is extremely hard to > create say a regular expression that we know is _absolutely_ correct > regarding separating the good TLDs from the bad ones.
I'm not sure the "good TLDs from the bad ones" is what the I-D in question is trying to separate. I have the impression, from reading it, that the I-D is trying to talk about the narrow, technical definition of the last label aside from the null-length root one, as it appears in the DNS -- what we call "TLDs", but only in the narrow, technical sense of "the label appearing in the DNS protocol". The document has nothing to say about languages, code points, good policy, &c -- it's very careful, in fact, to say that it's not tackling that question, and is only interested in the DNS definition of labels at a particular level. I appreciate that perhaps the other questions are important ones that need tackling. But we have a narrow, technical problem right now that needs to be addressed: at the moment, depending on how one reads RFC 1123, the IANA-operated root zone has in it labels that it should not, because it contains labels that are not "alphabetic". I would like to suggest that we need to solve that specific issue first, just so that we even have something else to talk about. If we can't solve that issue, then one might reasonably argue that the current IANA-operated root zone is in violation of RFC 1123, and try to prevent additional movement on internationalized TLDs that way. A -- Andrew Sullivan a...@shinkuro.com Shinkuro, Inc. _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
