On Thu, Dec 07, 2006 at 03:09:41PM -0500, Noel Chiappa wrote: > > From: David Meyer <dmm@1-4-5.net> > > > discussion of what the follow-on's to the workshop might be on various > > lists. What I'm hoping we can do is consolidate that discussion on > > ram@iab.org. > > I would take a slightly different view, and think there is (non-overlapping) > role for both (and I hope the various I* bodies won't take offense at me for > differing a bit with this plan; I really do have everyone's best interests at > heart here.) > > > First, this is a big problem, and there are a number of different technical > areas; with a single forum, it might get kind of Babel-like. Second, and more > important I think, I find that in any wide-ranging discussion like this, the > routing stuff (which is usually the hardest problem) gets the short end of > the stick. (I have my opinions on why that happens, but I'll leave them be > for now.) And, I hope I don't need to point out, it was coming problems with > the routing which have led to this. > > IMNSHO, it's really unacceptable to work on this problem and not give a key > role to consideration of how the routing is going to work. And to make sure > the routing is going to work, we have to dive into the muck and tackle the > routing technical issues in some detail, to make sure it all really works. So > we have to drive pretty quickly to considering the technical details of how > the routing is going to work, I think. And that's something the people who > seem to have energy available to discuss things have been fairly loathe to do, > despite some not-so-subtle prods from me. > > > So I think keeping the RAM list focused on just the routing stuff is really > pretty critical; it's clear that that's a really substantial discussion. > Discussion of other related topics there (e.g. separation of location and > identity) would just distract from that, and that's where A-D list would be > useful. > > I feel that discussion of such issues as end-end naming, whether we have to > use the "jack-up" model (in which end-hosts remain totally unmodified), etc, > etc, really ought to be kept on A-D, to keep RAM clear for these routing > issues. > > To keep this from getting too long, I'll send out a separate message in a > couple of hours (just on RAM) giving a little more detail on the points I > think RAM needs to work on, and how I think it can best make progress on them. Thanks Noel. No disagreement on my part. --dmm
Attachment:
pgpgSaO2MhA8o.pgp
Description: PGP signature