At Mon, 28 Oct 2002 22:21:47 -0500, Margaret Wasserman wrote: > > Well, I've certainly heard it, and it looks much better, to me, > than the "two-faced" DNS hack I've heard described in the past. > > I'd love to hear what the other DNS folks think, because maybe > this would be a promising approach to address one of the serious > problems with private addressing.
>From a strictly DNS technical point of view, Mark's proposal is fine. That is, the DNS protocol mechanics are not a problem (nor would I expect them to be, coming from Mark). There is, however, the issue of the local mapping table between scope id and DNS equivilent (my-site.example.net and my-link.example.net in Mark's example). As Mark mentioned, this is not zeroconf. Also note that this mechanism uses a new RR type. This is certainly doable, but it's a bit painful for address types due to the DNS additional section processing rules and the number of other pieces of deployed code that would need to be updated. Of course, one could back translate SA6 to AAAA, but that breaks end-to-end DNSSEC.... Dunno if this is starting to sound familiar to anybody else, but my ears are still ringing from the AAAA vs A6 debate, and I don't particularly want to relive that experience anytime soon. The above comments are only intended to discuss the DNS sub-thread, and should not be construed as a change from my previously stated support for Margaret's proposal to forbid use of site-local addresses on globally-connected networks. -------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
