sip@conference.ietf.jabber.com - 2003/03/17
[10:57] %% mass has arrived.
[10:57] %% mass has left.
[11:16] %% caryfitz has arrived.
[13:53] %% kmurchison has arrived.
[13:53] %% kmurchison has left.
[14:08] %% atanu has arrived.
[14:08] %% nermeen has arrived.
[14:08] %% atanu has left.
[14:15] %% rjs3 has arrived.
[14:16] %% atanu has arrived.
[14:16] %% atanu has left.
[14:19] %% lisaDusseault has arrived.
[14:19] %% es has arrived.
[14:19] <lisaDusseault> Hello I just volunteered to coordinate between the in-person SIP meeting and this jabber room.
[14:19] <lisaDusseault> Anybody in this jabber room not in the physical room?
[14:19] <caryfitz> yes
[14:19] <rjs3> yes
[14:20] <lisaDusseault> If you have questions to bring to the general group let me know. I'll try to transcribe but I predict my wrist will hurt if I do too good a job!
[14:20] <caryfitz> appreciate it
[14:20] <lisaDusseault> Rohan is going over draft status - three drafts in RFC-editor queue.
[14:21] <lisaDusseault> WG last call planned for JRosen's callerprefs draft immiedately after SF meeting.
[14:21] <lisaDusseault> Brian rosen is new editor for session-timer draft.
[14:21] <lisaDusseault> After his next version plan to advance that to IESG.
[14:21] <lisaDusseault> Guidelines draft might also go to IESG soon.
[14:23] <lisaDusseault> New topic: Non-INVITE transaction issues, Robert Sparks.
[14:24] <lisaDusseault> Non-invite transaction race condition was found.
[14:24] <lisaDusseault> [also you guys let me know what topics you're particularly interested in, as they come up, so I can decide where to type more and where to type less]
[14:24] <lisaDusseault> [thx]
[14:25] <lisaDusseault> In the race condition case, the UAS believes it has responded in time, but the UAC has a different idea of the result of the transaction.
[14:26] <lisaDusseault> The response can arrive after the timeout. THis is ironic when the "408 request timeout" arrives after the UAC timed out the request.
[14:26] <lisaDusseault> Even worse when proxies a "408 storm" descends on the originating client
[14:27] <lisaDusseault> [Robert pictures are very informative, his slides ought to be part of the meeting notes later]
[14:27] %% carlalf has arrived.
[14:27] <lisaDusseault> [if he hasn't already posted them]
[14:27] %% rjs3 has left.
[14:28] <lisaDusseault> If you end up with a transport error and move on to the next server in the SRV list, this is normal practice, but in the non-INVITE case you can't move on.
[14:28] <lisaDusseault> We need to push this back into the draft.
[14:29] %% fsolensky has arrived.
[14:30] %% rjs3 has arrived.
[14:30] %% mrose has arrived.
[14:30] <lisaDusseault> Should non-INVITE provisional responses be banned?
[14:31] <lisaDusseault> We'll harm recovery time for a lost response.
[14:31] <lisaDusseault> Timing diagrams are in the draft to show how this works.
[14:31] <lisaDusseault> You can get blacklisted, even worse than the race condition.
[14:31] %% hildjj has arrived.
[14:32] <lisaDusseault> There are two proposals [which are a bit much for me to transcribe]
[14:35] <hildjj> proposal a:
[14:35] <hildjj> -removes waste from the network
[14:35] <hildjj> - improves recovery from a failed node
[14:35] <lisaDusseault> Joe H is spelling me off transcribing (thanks Joe)
[14:35] <hildjj> - UAS NI transactions still have to complete immediately or risk losing the race
[14:35] <hildjj> - Does not address proxy doom
[14:35] <hildjj> Proposal B
[14:35] <hildjj> - 408's become useful again
[14:36] %% pgmillard has arrived.
[14:36] <hildjj> - Relieves proxies from timeout doom
[14:36] <hildjj> - Still needs to address harmful provisionals for UDP
[14:36] <hildjj> - Either deprecate UDP or take Proposal A 1-3
[14:37] <lisaDusseault> [Rohan mahy at mike]
[14:37] <lisaDusseault> [Now tim collins at mike]
[14:37] <hildjj> Tim: it's worth extending the discussion about the timeout values
[14:38] <hildjj> sparks: look in the draft for a beginning of the discussion. T1 are same on both sides.
[14:39] <hildjj> for non-invites disparate values of T1 are bad... leads to failure conditions
[14:40] %% TonyHansen has arrived.
[14:40] <lisaDusseault> [Cullen Jennings at mike]
[14:40] %% kibozer has arrived.
[14:41] <hildjj> cullen: we should come up with a resonable time. it doesn't have to be 64*T1.
[14:41] %% pgmillard has left.
[14:42] <lisaDusseault> [Dan Petrie at mike]
[14:43] %% pgmillard has arrived.
[14:44] <hildjj> T1 is supposed to be an approximation of network round-trip
[14:44] <lisaDusseault> [Henning Schulzrinne at mike]
[14:44] <hildjj> less worried about 1 in 10k problems
[14:45] %% pgmillard has left.
[14:47] %% carlalf has left.
[14:48] <hildjj> rosenberg: artificial constraints based on existing proxies
[14:49] %% carlalf has arrived.
[14:50] <hildjj> rosenberg: this is hard to do on the list, it's very detailed. can we take some people into a room with a whiteboard?
[14:50] <hildjj> sparks: send me a mail if you are interested.
[14:50] <hildjj> sparks: small group is better, so only come if you have ideas
[14:50] <hildjj> moving to "referred-by" draft
[14:50] <hildjj> (sip jokes)
[14:51] %% mattb has arrived.
[14:51] <hildjj> no review yet on this draft
[14:51] <hildjj> chair calls for 4 volunteers
[14:51] <hildjj> 4 people volunteer
[14:52] <TonyHansen> /
[14:52] <TonyHansen> (sorry)
[14:52] %% michael has arrived.
[14:54] %% senthil has arrived.
[14:54] %% lisaDusseault has left.
[14:55] %% senthil has left.
[14:55] <hildjj> someone else who understands the issues better should probably take over. the SIP jargon is flying faster than I can parse.
[14:55] <hildjj> (and my battery is almost dead)
[14:56] %% hildjj has left.
[15:00] %% stephennadas has arrived.
[15:00] %% louie has arrived.
[15:05] %% kibozer has left.
[15:07] %% kibozer has arrived.
[15:07] %% fsolensky has left.
[15:08] %% mrose has left.
[15:08] %% mrose has arrived.
[15:09] %% stephennadas has left.
[15:10] %% randy has arrived.
[15:13] %% michael has left.
[15:13] %% kibozer has left.
[15:13] %% kibozer has arrived.
[15:28] %% nermeen has left.
[15:28] %% lisaDusseault has arrived.
[15:30] %% lisaDusseault has left.
[15:30] %% lisaDusseault has arrived.
[15:30] <lisaDusseault> Should I go back to logging here? I lost wireless for a while.
[15:31] <TonyHansen> please
[15:31] <lisaDusseault> Dean Willis just asked for a hum whether the ieprep draft should be taken on as a SIP WG charter item, with the caveat that ieprep requireemtns may be reexamined.
[15:31] <lisaDusseault> Consent was unanimous.
[15:32] <lisaDusseault> New topic: J Rosenberg, Caller prefs, 10 minutes.
[15:33] <lisaDusseault> Caller prefs has long been a solution looking for a problem which is why we've had this draft forever.
[15:33] <lisaDusseault> [Jon just dropped his microphone - they're teasing him about dropping his laptop last time]
[15:34] %% lisaDusseault has left.
[15:35] %% lisaDusseault has arrived.
[15:35] <lisaDusseault> We are now going to declare success on this draft and not accept any more use cases.
[15:36] <lisaDusseault> The rule matching has changed in the draft.
[15:36] <lisaDusseault> Many more changes are listed in a long slide which will be in the minutes.
[15:38] <lisaDusseault> Rohan is asking how many people read Jonathans draft or the use cases.
[15:38] <lisaDusseault> Rohan asks for 6 volunteers to review this draft.
[15:38] <lisaDusseault> [if you volunteer I'll tell the group]
[15:39] <lisaDusseault> Planning last call after volunteers review after this mtg
[15:39] <lisaDusseault> Any objections?
[15:39] <lisaDusseault> Sold.
[15:40] %% lisaDusseault has left.
[15:40] %% lisaDusseault has arrived.
[15:41] <lisaDusseault> New topic: Dean Willis congestion
[15:42] <lisaDusseault> Why are safe responses hard? Response must follow inverse path, but may be larger than request.
[15:42] <lisaDusseault> A "safe" UAS response is smaller than the request. Could send error if response cannot be smaller.
[15:49] %% sukanta has arrived.
[15:53] %% Cyrus has arrived.
[15:57] %% rjs3 has left.
[15:58] %% lisaDusseault has left.
[16:00] %% jasson has arrived.
[16:02] %% ulieberman has arrived.
[16:02] %% TonyHansen has left.
[16:02] %% ulieberman has left.
[16:05] %% takamiya has arrived.
[16:06] %% lisaDusseault has arrived.
[16:07] %% lisaDusseault has left.
[16:07] %% takamiya has left.
[16:07] %% lisaDusseault has arrived.
[16:07] <lisaDusseault> Obviously I am having trouble staying connected - the wireless keeps dropping me.
[16:07] <lisaDusseault> Rohan just started talking about interoperability.
[16:08] <lisaDusseault> He is asking for more implementations to test out new stuff to make sure it will work.
[16:08] <lisaDusseault> If you are interested in implementing S/MIME but think it is too hard see me or Jon.
[16:09] <lisaDusseault> As for TLS, we found out nobody did AES.
[16:10] <lisaDusseault> We were working on very basic TLS connectivity, which worked, but we didn't test any of the security features.
[16:10] <lisaDusseault> 5 minutes left for last issues.
[16:10] %% carlalf has left.
[16:10] <lisaDusseault> WG adjourned.
[16:10] <lisaDusseault> bye-bye.
[16:10] %% mrose has left.
[16:11] %% jasson has left.
[16:11] %% Joseph has arrived.
[16:14] %% mattb has left.
[16:15] %% Cyrus has left.
[16:15] %% sukanta has left.
[16:19] %% es has left.
[16:19] %% lisaDusseault has left.
[16:20] %% louie has left.
[16:29] %% Joseph has left.
[16:39] %% kibozer has left.
[16:40] %% randy has left.
[16:40] %% plenary has arrived.
[16:40] %% plenary has left.
[16:52] %% kibozer has arrived.
[17:38] %% sipping has arrived.
[17:38] %% sipping has left.
[18:21] %% jasson has arrived.
[18:21] %% jasson has left.
[18:43] %% kibozer has left.
[18:43] %% kibozer has arrived.
[18:59] %% kibozer has left.