On Tuesday 17 October 2006 16:32, Stephen Farrell wrote: > Scott Kitterman wrote: > > I'd be satisfied if the requirements draft were to say: > > > > The protocol MUST NOT require use of a new DNS RR type. The protocol MAY > > allow for optional use of a new RR type. > > But wouldn't that decide the issue in that the consequence of the above > would be that the protocol MUST mandate use of TXT (or some other > existing) records? Maybe I'm missing something though. > > If so, I think that silence on this in the requirements draft is best, > but keeping the issue open in the tracker so we do have the debate later > on.
On Tuesday 17 October 2006 16:43, Michael Thomas wrote: > First off, I apologize for an almost word-for-word repeating of myself in a > previous post -- it's fairly obvious that my mail reading habit is fifo > based :) > > My feeling is that deferring this argument until the design phase won't > likely hurt, while getting it wrong in requirements phase certainly > will. If we > can just be patient until then, I think there will be plenty of opportunity > to work this issue through. I personally agree that this is an important > issue, but I'm also very interested in real experience which will support > the various conjectures about how the discovery mechanism should work. OK. I'll sit down now. Let's just keep the issue open for now and move on. Scott K _______________________________________________ NOTE WELL: This list operates according to http://mipassoc.org/dkim/ietf-list-rules.html
