Wednesday, 28 March 2012< ^ >
Tobias has set the subject to: stpeter has set the subject to: XMPP WG | | IETF83 Meeting, March 28, 2012
Room Configuration

[01:08:35] MattJ leaves the room
[10:33:43] Tobias joins the room
[11:00:38] Kev joins the room
[11:31:53] fippo joins the room
[12:08:29] Zash joins the room
[12:08:46] MattJ joins the room
[12:42:49] Tobias has set the subject to: XMPP WG | | IETF83 Meeting, March 28, 2012
[12:42:49] Zash leaves the room
[12:45:39] joins the room
[12:54:11] winfried joins the room
[12:55:26] Zash joins the room
[13:02:01] Florian Zeitz joins the room
[13:03:52] <Kev> Has anyone got the feed URI to hand? I know I have it in my inbox somewhere, but..
[13:04:15] <MattJ>
[13:04:24] <Kev> Marvellous, thanks.
[13:05:01] Hirotaka Sato joins the room
[13:07:18] <Kev> Well, that was happy for a couple of seconds.
[13:07:21] <Kev> Can't get it back now, though.
[13:08:51] <winfried> HTTP/1.0 404 File Not Found (open of `')
[13:09:01] hildjj joins the room
[13:09:30] gshapiro joins the room
[13:09:35] gshapiro leaves the room
[13:09:36] Jonathan Lennox joins the room
[13:09:57] yone joins the room
[13:10:14] Ben joins the room
[13:10:36] Dan York joins the room
[13:10:55] alexey.melnikov joins the room
[13:11:40] m&m joins the room
[13:11:43] pcl joins the room
[13:11:44] <hildjj> sound check?
[13:12:01] <pcl> Joe: I can hear you fine
[13:12:01] <Kev> I can't open the stream :/
[13:12:08] <Florian Zeitz> Kev++
[13:12:27] gshapiro joins the room
[13:12:27] <hildjj> anything we can do to help?
[13:12:57] <Kev> I have no idea :)
[13:12:59] rbarnes joins the room
[13:13:02] <Kev> Let me try a different client.
[13:13:08] <Kev> is correct, right?
[13:13:18] <rbarnes> hello, my name is richard, and i'll be your scribe today
[13:13:24] <Kev> Nope, 404 in a different client too, as expected I guess.
[13:13:25] <Dan York> hello, richard
[13:13:29] <rbarnes> if you have any comments you would like relayed to the mic, please prefix them with MIC
[13:13:53] <rbarnes> that URI loads for me
[13:13:59] <m&m> this problem occurred yesterday
[13:14:03] redaka joins the room
[13:14:03] <m&m> /-:
[13:14:15] <MattJ> I'm on a conf call atm so can't test
[13:14:16] <rbarnes> oh, no, actually, 404 for me too
[13:14:23] <m&m> we are pinging someone in MOC to figure it out
[13:14:23] <hildjj> stpeter is working the audio stream in the background..
[13:14:23] stpeter joins the room
[13:14:24] <m&m> s/MOC/NOC
[13:14:29] <Kev> Thanks.
[13:14:32] <hildjj> we're going ahead, though.
[13:14:39] <pcl> notes are also going into etherpad
[13:14:41] <pcl>
[13:15:05] ogashiwa joins the room
[13:15:20] <rbarnes> for anyone in the room: note takers get applause! (but not beer)
[13:15:25] <rbarnes> chairs doing agenda review
[13:15:31] stpeter pings the secretariat
[13:16:01] <rbarnes> m&m presenting on e2e
[13:16:31] <stpeter> "Big Ideas" slide
[13:16:43] Dave Cridland joins the room
[13:16:44] stpeter notes to m&m that it's really helpful to number your slides :)
[13:16:58] <rbarnes> slide: Discovering Support
[13:17:06] waqas joins the room
[13:17:13] <rbarnes> slide: Encrypting ..
[13:17:18] Dan York agrees with stpeter on the importance of numbers!
[13:17:42] <Jonathan Lennox> /help
[13:17:42] <Jonathan Lennox> Hmm
[13:17:44] <rbarnes> slide: … Still Encrypting ...
[13:17:52] <Dave Cridland> Strictly monotonically increasing numbering of slides is even better.
[13:18:01] <Jonathan Lennox> Dave: sequential, even
[13:18:03] <stpeter> so is not working?
[13:18:15] <Kev> stpeter: Correct.
[13:18:17] <stpeter> sigh
[13:18:29] <rbarnes> slide: … Encrypted!
[13:18:39] <rbarnes> slide: Decrypting ...
[13:18:50] <Hirotaka Sato> seems room 253 has been dropped from
[13:18:50] <winfried> the url of the stream it links to is 404 ( )
[13:19:00] <Kev> Which 404s
[13:19:10] <rbarnes> slide: … Still Decrypting ...
[13:19:11] <stpeter> ok
[13:19:41] <rbarnes> slide: … Decrypted!
[13:19:45] <Kev> It worked fine for a few seconds when I clicked it first just before 1300UTC, then cut out pretty much on the hour.
[13:20:04] <Dave Cridland> SO it's not just me having audio problems, then.
[13:20:17] <Dan York> In one of the other chats (soc) it seems like they are having meetecho problems as well.
[13:20:23] <rbarnes> slide: Making a Keyreq
[13:20:30] <Jonathan Lennox> meetecho is different from the mp3 streaming
[13:20:43] <rbarnes> slide: Accepting a Keyreq
[13:20:46] <Dan York> understood! Just noting that there are remote difficulties
[13:20:59] <Ben> do we have a problem report in to someone?
[13:21:17] <Dave Cridland> Ben, Someone other than each other, you mean? :-)
[13:21:24] <rbarnes> slide: Denying a Keyreq
[13:21:29] <Jonathan Lennox> Do the other rooms have mp3 streaming working? Wondering if it's one room or all
[13:22:37] <Dan York> richard barnes at the mic
[13:22:43] <hildjj> rbarnes says: please don't use keys
[13:22:55] <Jonathan Lennox> don't *re*-use keys
[13:22:55] <gshapiro> "use" -> "reuse"
[13:22:59] <rbarnes> thank you, backup scribes
[13:23:05] rbarnes applauds backup scribes
[13:23:16] xnyhps joins the room
[13:23:23] <Dan York> richard barnes back at the mic
[13:24:37] <Dan York> Joe Hildebrand at the floor mic
[13:25:19] <Dan York> Martin Thompson at the mic
[13:25:31] <Dan York> richard barnes back at the mic
[13:25:44] <Dan York> rb: document should have a section on key scoping
[13:25:46] <Dave Cridland> You know, without any audio, I suspect that knowing *what* is said rates higher than who said it.
[13:25:49] <hildjj> seems like we're agreeing it's the sender's choice, unique per-recipient
[13:26:06] <hildjj> on how much key-reuse we allow
[13:26:12] <Dan York> dwd - yes, good point!
[13:26:21] <rbarnes> hildjj: +1, but we need to recommend stuff
[13:26:22] <stpeter> signing...
[13:26:33] <Jonathan Lennox> Dave: are you watching the etherpad?
[13:26:35] <stpeter> m&m says we're going to ignore signing for now
[13:26:39] <Kev> Audio back
[13:26:41] <Dave Cridland> Oh, there's an etherpad?
[13:26:48] <stpeter> do you guys care about signing?
[13:26:48] <Kev>
[13:26:50] nemo joins the room
[13:27:06] <Florian Zeitz> I care about signing. Encryption without signing is IMHO useless
[13:27:06] <hildjj> Audio is back, apparently
[13:27:09] <Dave Cridland> I hear voices!
[13:28:08] vincent.levigneron joins the room
[13:28:26] <stpeter> so do people care about signing without encryption?
[13:28:31] <Dave Cridland> Yes.
[13:28:58] <Dave Cridland> There's several cases where non-repudiation of what is said end-to-end is rather important.
[13:29:22] <Dan York> dan york (me) at the mic
[13:29:23] <stpeter> send text :)
[13:29:31] <hildjj> dwd: you got volunteered to help
[13:29:32] <Dan York> richard barnes back at the mic
[13:29:41] <Dave Cridland> I shall dutifully delegate that one to Kurt Zeilenga, but yes. :-)
[13:29:42] <Jonathan Lennox> How is the streaming latency?
[13:29:51] <Dave Cridland> Jonathan Lennox, 10 secs?
[13:29:51] <Kev> Jonathan Lennox: 30seconds or so, at a guess.
[13:30:49] Tobias leaves the room
[13:31:20] <Dan York> Joe Hildebrand at the floor mic
[13:31:52] <Kev> Well, that audio was nice while it lasted :)
[13:32:05] <Ben> audio gone again?
[13:32:10] <Dave Cridland> I still hear voices.
[13:32:14] <Kev> I hear nothing. Hmm.
[13:32:16] <winfried> still got audio here
[13:32:20] <Dave Cridland> They don't appear to be in my head.
[13:32:27] <Dan York> richard barnes back at mic
[13:32:39] <Kev> I wonder why I've lost it and can't get it back, then.
[13:32:48] <Dan York> joe queueing back
[13:32:49] <stpeter> we've been having significant audio problems here this week
[13:33:50] <Kev> Completely restarting the stream worked for me.
[13:33:53] rbarnes applauds m&m for a good initial document
[13:34:17] <Dan York> chair (Ben) asks what are next steps
[13:34:24] <Dan York> Peter St Andre at mic
[13:35:20] <Dan York> Peter: "Should we look at OTR as an option?"
[13:35:31] <Kev> I'd have thought that anything that can encrypt a body can encrypt a stanza.
[13:35:52] <Dave Cridland> Kev, In principle, yes, but the currently running code does not.
[13:36:06] <Kev> Right, the current running code is not the right way to do OTR(XMPP)
[13:36:12] <Kev> But as a bootstrap it's appealing.
[13:36:16] <rbarnes> NEW PRESO: XMPP DNA
[13:36:17] <Dave Cridland> Kev, Mostly because the currently running code is designed to operate over any IM service.
[13:36:21] <rbarnes> slide: Existing Work
[13:36:27] <stpeter> moving on the DNA
[13:36:43] mycom joins the room
[13:37:09] <hildjj> are you guys hearing the siren on the stream?
[13:37:10] <Ben> is the audio stream picking up the french traffic noises?
[13:37:10] <rbarnes> slide: The Problem
[13:37:26] <stpeter> there's a traffic authorization problem out on the street here :)
[13:37:33] <Dave Cridland> I can't hear it over my own traffic noises, at least...
[13:38:35] <rbarnes> slide: building blocks
[13:38:54] <rbarnes> slide: Approaching Trust
[13:39:12] <Kev> And I'm the only one the audio keeps cutting out on?
[13:39:26] <Dave Cridland> Kev, Mine's stable, now.
[13:39:32] <Kev> Lucky me.
[13:39:36] <rbarnes> slide: Proof types
[13:39:51] <stpeter> browser / audio player issues with the audio?
[13:40:05] <Kev> My client loses the stream and I have to reload it.
[13:40:10] <Kev> (Reload the stream, that is)
[13:40:10] <stpeter> sigh
[13:40:23] <rbarnes>
[13:40:35] <stpeter> Kev: if you can provide error reports of some kind (not necessarily on this session), that would be helpful
[13:41:37] <stpeter> BTW, /.well-known/ is RFC 5785
[13:41:58] <Kev> I'm not sure I can, other than "My client stops playing audio". I'll look for some debug.
[13:41:58] <Kev> .
[13:42:25] <Dan York> richard barnes at mic
[13:42:32] Tobias joins the room
[13:43:37] <rbarnes> slide: Approaching Delegation
[13:44:09] <rbarnes> slide: Signaling delegation
[13:44:13] <Dave Cridland> I would note that "Don't freak out" is clearly a downgrade attack. :-)
[13:44:18] <Kev> How useful is the server saying "don't freak out please"? Surely someone's either going to freak out or not.
[13:44:31] <rbarnes> cridland: how so?
[13:44:38] <Ben> Dave, Kev, are those for the mike?
[13:44:41] <Kev> Ah, sweet audio, how do I miss thee.
[13:44:44] <stpeter> BTW, prepend with "MIC" if you want your words relayed
[13:45:01] <Dave Cridland> Well, I'll wait for the questions bit at the end...
[13:45:14] <rbarnes> slide: Outcomes
[13:45:23] <stpeter> /e goes up to the mic, expecting to relay
[13:45:29] stpeter learns to type
[13:45:52] <Dave Cridland> stpeter, I'll give you a chance to do your British accent.
[13:46:08] <stpeter> yeah right
[13:46:44] <Dave Cridland> stpeter, You mean "Jolly good, old chap".
[13:47:02] <rbarnes> dan york at the mic
[13:47:31] <pcl> rbarnes at the mic
[13:48:03] <Dave Cridland> MIC: So I don't think that you need to change *anything* on the wire for XMPP - what we're changing here is the decision on how do we authorize the cert presented as being for the domain, and that's changing RFC 6125 rather than RFC 6120.
[13:48:27] <Dave Cridland> MIC: ... Well, not changing, but adding to.
[13:48:38] <stpeter> ok, that's better :)
[13:48:57] <Dave Cridland> All in a British accent, K?
[13:49:32] <stpeter> I'm goin' t' use a southern accent
[13:49:52] <rbarnes> york at the mic
[13:50:51] saghul joins the room
[13:51:39] <Dan York> Joe Hiledbrand from the floor
[13:52:08] <Kev> MIC: I think you do have to try all 20, or you have an attack.
[13:54:08] <Dave Cridland> MIC: If, before the TLS certificate is sent, you say "Hey, look, I'm about to send you a bogus cert and it's OK" over an insecure channel, that seems like a potential avenue for a downgrade attack. I imagine it's not the only path, and might even not be worse than existing ones.
[13:54:25] <Kev> MIC: Yes, I think that's right, If you can compromise the stream such that auth fails, you can compromise it such that everything fails, so it's just a different path to DoS.
[13:54:42] <Kev> ("That's right" being "I'm wrong")
[13:55:47] <Kev> (That is - I retract my previous comment)
[13:56:20] <Dan York> richard barnes at mic
[13:56:40] <Dave Cridland> MIC: I'll also note that existing running code always continues to handle the stream looking to authenticate via other methods. That is, our existing running code is basically doing DNA via traditional dialback as it is.
[13:56:58] winfried leaves the room
[13:58:08] mycom leaves the room
[13:59:52] <Dave Cridland> I'm actually willing to implement at least the .well-known mechanism now, without XMPP-level changes.
[14:00:11] <rbarnes> put'em all in the document, let the reader sort it out
[14:00:21] <rbarnes> stpeter coming up to present on XMPP Address Format
[14:00:23] <Dave Cridland> (And thus put in some feedback ino this stuff via implementation practise)
[14:00:27] <hildjj> dwd: thank you. that's likely useful.
[14:01:12] <rbarnes> slide: Open Issues (I)
[14:01:37] mycom joins the room
[14:03:13] gshapiro leaves the room
[14:05:29] <Dave Cridland> MIC (at some point): My proposal was closer to saying that I didn't think we needed to worry very much over resource parts, with the exception that chatrooms were important. That is, resource parts in practise are never typed by users, and so need little fiddling with to handle normalization etc.
[14:06:00] <hildjj> dwd: i think we got enough of the flavor of that, with the time allowed.
[14:06:06] <hildjj> let's discuss at precis
[14:06:16] <hildjj> ok?
[14:06:20] Hirotaka Sato leaves the room
[14:06:30] さとうひろたか joins the room
[14:07:48] <rbarnes> slide: Open Issues (3)
[14:08:27] <rbarnes> stpeter done
[14:08:39] alexey.melnikov leaves the room
[14:08:46] Dan York leaves the room
[14:09:37] <rbarnes> adjourned!
[14:09:38] Jonathan Lennox leaves the room
[14:09:45] vincent.levigneron leaves the room
[14:09:45] m&m leaves the room: Disconnected: session closed
[14:09:55] <Kev> Thanks scribes.
[14:10:19] yone leaves the room
[14:11:02] Ben leaves the room
[14:11:08] <rbarnes> no problem
[14:11:09] nemo leaves the room
[14:11:52] pcl leaves the room
[14:12:25] saghul leaves the room
[14:13:09] mycom leaves the room
[14:13:52] rbarnes leaves the room
[14:14:22] mycom joins the room
[14:14:39] hildjj leaves the room
[14:21:34] Zash leaves the room
[14:22:59] vincent.levigneron joins the room
[14:24:12] Florian Zeitz leaves the room: offline
[14:26:09] redaka leaves the room
[14:29:09] mycom leaves the room
[14:32:22] rbarnes joins the room
[14:32:33] ogashiwa leaves the room: Computer went to sleep
[14:33:05] m&m joins the room
[14:33:31] rbarnes leaves the room
[14:33:50] rbarnes joins the room
[14:34:22] rbarnes leaves the room
[14:34:55] Jonathan Lennox joins the room
[14:35:12] nemo joins the room
[14:36:09] さとうひろたか leaves the room
[14:38:26] m&m leaves the room
[14:38:32] Jonathan Lennox leaves the room
[14:39:31] stpeter leaves the room: Disconnected: connection closed
[14:56:02] Tobias leaves the room: Disconnected: connection closed
[15:00:23] Ben joins the room
[15:02:43] Ben leaves the room
[15:15:09] nemo leaves the room
[15:21:55] Dave Cridland leaves the room
[15:21:55] Dave Cridland joins the room
[15:33:39] leaves the room
[15:49:45] xnyhps leaves the room
[16:17:10] vincent.levigneron leaves the room
[16:38:45] ralphm joins the room
[16:38:45] fippo leaves the room
[17:14:51] ralphm leaves the room: Disconnected: conflict
[17:15:15] ralphm joins the room
[17:42:53] waqas leaves the room
[19:20:20] Kev leaves the room
[20:17:37] Tobias joins the room
[20:18:09] ralphm leaves the room
[23:28:23] Tobias leaves the room: Offline
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!