[13:05:46] --- ggm has joined
[13:06:22] <ggm> CRISP is on!
[13:06:27] <ggm> ok. I lied. its almost on.
[13:07:14] --- dblacka has joined
[13:07:30] <ggm> agenda. is on the IETF web site
[13:07:36] <ggm> http://www.ietf.org/ietf/04nov/crisp.txt
[13:13:56] --- hardie has joined
[13:16:14] <ggm> April has the chair. 1hr meeting.
[13:16:20] --- ed has joined
[13:17:03] <ggm> two things to progress. Areg, and Andys newest efforts (update on)
[13:17:14] <ggm> current IRIS docs in RFC-ED q state.
[13:17:21] --- anewton has joined
[13:17:49] <ggm> IDN over EPP BoF today, 5pm. could relate to us. encourage people to go.
[13:18:04] <ggm> Shane/Engin. AReg status.
[13:18:46] --- kjd has joined
[13:18:49] --- vlevigneron has joined
[13:19:15] <ggm> changes since last IETF meeting being discussed.
[13:19:37] --- pawal has joined
[13:19:39] <ggm> 2 versions released, 07 and 08.
[13:19:49] <ggm> 6 to 7:
[13:20:31] <ggm> <name> element is back.
[13:20:58] <ggm> <findASNByNumber> search is in. -with specificity. AS query is out (not needed)
[13:21:21] <ggm> its subsumed by this function.
[13:21:58] <ggm> copied something from DREG. <language> element. -in <findOrganization> and <findContacts>
[13:22:16] <ggm> 6 to 7 versions slide 2.
[13:22:45] <ggm> taken out closest-match specificity query. not needed. covered by less-specific with allowEqualities set true.
[13:23:12] <ggm> <beginsWith> and <endsWith> restored for some <find*> ops
[13:23:28] <ggm> addded 'inverse' queries in (from WHOIS), needed functionality. to look by contact into resources.
[13:23:57] <ggm> added <findNetworkBySpecificity> search. finds multiple nets with same start and end addresses. other searches don't work well to figure out parentage
[13:24:04] <ggm> 6 to 7 slide 3
[13:24:42] <ggm> changed XML syntax to match changes. changes to <contact> result.
[13:24:48] <ggm> diffs 7 to 8.
[13:25:11] <ggm> added one more search, <findNetworksByNameserver> -corresponds to an inverse query in current implementation.
[13:25:17] <ggm> split into normative/informative structure
[13:25:26] <ggm> added examples (as presented at last IETF)
[13:25:39] <ggm> fixes into example request/response sections.
[13:25:43] <ggm> fixes in formal XML.
[13:25:52] <ggm> last issue. URI resolution.
[13:26:10] <ggm> says 'should start from iris.nro.net, follow referrals'
[13:26:23] <ggm> Issue is what WKS does to IESG. push to IESG, see what happens.
[13:28:07] <ggm> Ted: go forward with easy way to tell people how to manage referral problem. 'here is the root'
[13:28:22] <ggm> But people will want to set these up local to org, outside of global. eg private ones.
[13:29:30] <ggm> Given what we just went through pulling WHOIS stuff out, pulling WKS out of this doc, but in separate doc (informational) doing 'this is what NRO is going to do' doesn't build dependency on single root into protocol. if that would work for people, want to deploy. but pull away from proto doc and put in informational/BCP gets cake and eats it too.
[13:29:43] <ggm> Andy need to set context.
[13:30:08] <ggm> doesn't mandate client has to use this resolution method. without, can stipulate direct method. can come back later and define other methods.
[13:30:24] <ggm> gone round and round, how to do this, how do server discovery and come back to WKS.
[13:30:34] <ggm> Ted splitting into doc to read separately..
[13:30:36] <ggm> Andy they won't see it.
[13:30:55] <ggm> Shane. I really like the idea to put it into BCP/informational
[13:31:40] <ggm> Leslie off the cuff, yes probably makes sense. easier to say 'reporting current status' and change it -or make it IANA parameter
[13:31:54] <ggm> Markus or specify fallback mechanism
[13:32:31] <ggm> Engin. this is all we have. deliberate. so it can be made better later, makes sense to me to split docs, leave unspecified here. if needed continue to work on problem.
[13:33:40] <ggm> Andrei. if can leave WKS in document, AND do BCP for crisp in registries, would be fine
[13:34:06] <ggm> Cathy. another divergeance from DREG. 'we might get "its in DREG. why not in AREG too?" responses. don't have problem spitting it out.
[13:35:23] <ggm> Engin. had clear idea of URI resolution in domain stuff, this is new stuff.
[13:35:32] <ggm> Leslie chose name to persist under .arpa?
[13:36:09] <ggm> April: who to write? [ENGIN!]
[13:36:24] <ggm> so now this document, in good shape, and new small document, to go forward.
[13:36:51] <ggm> Cathy: arpa is infrastructure, this is directory service related. not sure it would be acceptable to IAB. if going to write, needs to know what he is going to write.
[13:36:56] <ggm> April separate discussion.
[13:36:59] --- yone has joined
[13:37:22] <ggm> April take to list.
[13:37:32] <ggm> April need to move forward. ready for WG last call?
[13:37:47] <ggm> Engin do we need to fully implement the thing and then call for it?
[13:37:49] <ggm> April no.
[13:38:38] <ggm> Engin then I think it can be pushed forward.
[13:38:49] <ggm> April aim for end of month for second document, then 2 weeks last call, then ship.
[13:39:13] <ggm> Shane Q is, do we care about LLA query? its not in here.
[13:39:24] <ggm> Tim I posted email about hour ago to ML on this.
[13:41:36] --- robertml has joined
[13:42:02] <ggm> in San Diego, seemed to be nominal support from 6-7 people in the room "sounds like good idea" but Andy asked if we have solid use cases, and I have to admit we don't. In light of the fact that its not trivial to implement, don't have real strong use case for it. posted logic on LLA on-list, more that it created a balence between ability to do range queries, and 'show me responsible person for this <node>' combine the two, have LLA query. no other query gets to that point. but since there hasn't been a real user community needs this, something we can have later. not going to scream and yell about last call if there isn't a strong use-case for it. I can see it, I think other people can see it, some talk on list, given a couple of arbitrary numbers, can yield results you DONT want. don't know we need to design this to stop dumb things, but unless there is a strong comment from the room to keep it, set aside.
[13:42:37] <ggm> April so hopefully will be at IESG next time we meet.
[13:43:36] <ggm> Andy: iris dchk draft, iris-lwz
[13:43:54] <ggm> domain availability check and ultra lightweight UDP protocol. so split it out.
[13:44:31] <ggm> dchk defines the specifics for domain availability, lwz is for any context, just a lightweight query mechanism.
[13:44:41] <ggm> small traffic on the doc: nits, spelling. (dchk)
[13:44:42] --- g.engin has joined
[13:45:07] <ggm> for lwz draft, implemented, XML wrapping layer turns out to be bad idea. requires entrypoint into upper layer, can't see with BEEP or other trad transport.
[13:45:21] <ggm> the 'get profiles' exchange is redundant given S-NAPTR.
[13:45:28] <ggm> want to change to simple binary wrapper on XML.
[13:46:06] <ggm> request packet. 1byte header, 2 byte max resp. size willing to accept, byte authority
[13:46:11] <ggm> response is the same, plus response
[13:46:32] <ggm> simple header, version bit. req/resp flag, flag to see deflate state, error bits.
[13:47:00] <ggm> Ed.
[13:47:12] <ggm> Q based on DNSOP issue with resp size.
[13:47:24] <ggm> 512 Bytes is hard limit?
[13:47:35] <ggm> Andy if no higher length specified.
[13:48:00] <ggm> April read draft? [3-4 hands, usual suspects]
[13:48:14] <ggm> Andy can change, turn around quickly. will send to ML. not expecting big objections
[13:48:28] <ggm> April so we can take to WG last call?
[13:48:29] <ggm> Andy yes
[13:49:32] <ggm> AOB....
[13:49:33] <ggm> done!
[13:49:40] --- anewton has left
[13:50:16] --- ed has left
[13:50:16] --- yone has left
[13:50:42] --- dblacka has left: Disconnected
[13:51:07] --- pawal has left
[13:51:28] --- g.engin has left
[13:52:51] --- vlevigneron has left
[13:56:26] --- robertml has left
[13:56:41] --- hardie has left: Disconnected
[13:57:21] --- kjd has left: Replaced by new connection
[14:16:21] --- ggm has left: Disconnected
[14:33:58] --- hardie has joined
[14:34:05] --- hardie has left