IETF
rtcweb@jabber.ietf.org
Friday, April 8, 2016< ^ >
martin.thomson has set the subject to: RTCWEB WG http://tools.ietf.org/wg/rtcweb/agenda
Room Configuration
Room Occupants

GMT+0
[06:30:32] Neustradamus leaves the room
[12:52:13] Meetecho joins the room
[12:53:45] Jonathan Lennox joins the room
[12:59:24] m&m joins the room
[13:00:18] <m&m> ¡buenos dias!  I'll be your jabber relay, if you have something you want said in the room, preface it with "mic:", por favor
[13:00:18] Dan Wing leaves the room
[13:00:19] Wolfgang Beck joins the room
[13:00:42] ken carlberg joins the room
[13:00:42] <Wolfgang Beck> muchismas gracias
[13:00:45] csp joins the room
[13:01:13] Maire Reavy joins the room
[13:01:27] <Jonathan Lennox> Alternately, eat our own dogfood, and try the Virtual Queue!
[13:01:35] m&m leaves the room: Disconnected: Replaced by new connection
[13:01:40] Victor Pascual joins the room
[13:01:51] m&m joins the room
[13:02:21] Paul Kyzivat joins the room
[13:02:33] <Wolfgang Beck> as a non-native speaker, it is nice when you hear your questions in perfect english from the jabber relay..
[13:03:51] Varun Singh joins the room
[13:03:57] Ray Atarashi joins the room
[13:04:17] ted.h joins the room
[13:04:23] Ben Campbell joins the room
[13:04:34] Randell Jesup joins the room
[13:04:52] ken carlberg leaves the room
[13:05:13] <m&m> JSEP - ekr
[13:05:43] Ray Atarashi leaves the room
[13:05:48] <m&m> or rather "JSEP - Justin"
[13:06:09] Ray Atarashi joins the room
[13:06:46] frodek joins the room
[13:06:58] <m&m> slide 1: Changes since IETF 94 (1 of 2)
[13:07:08] Ray Atarashi leaves the room
[13:07:23] <Randell Jesup> Wolfgang - need to loop in speech recognition (which we've been adding in Firefox), feed that through automated translation, then through WebSpeech to synthesize english - or use mic: if you're worried about that.  ;-)
[13:08:00] <m&m> slide: Added Rtp Transceivers, RtpSenders, RtpReceivers
[13:08:03] Maire Reavy_2 joins the room
[13:08:39] Maire Reavy leaves the room
[13:09:03] <m&m> slide: Removed OfferToReceiveX options
[13:09:33] Ray Atarashi joins the room
[13:09:39] <m&m> slide: General Bandwidth Handling
[13:10:00] Maxime Marchand joins the room
[13:10:04] Maire Reavy joins the room
[13:10:15] magnus joins the room
[13:10:18] Alissa Cooper joins the room
[13:10:22] <m&m> slide: How addIceCandidate can be used for end-of-candidates
[13:10:32] <Wolfgang Beck> I fink I have to speak ze englisch zen
[13:10:44] Maire Reavy_2 leaves the room
[13:10:55] <m&m> slide: Clarified how to handle various received attributes
[13:11:08] <m&m> slide: How attributes should be generated for bundled m= lines
[13:11:10] Chris Wendt joins the room
[13:11:42] Karthik BR joins the room
[13:12:22] <m&m> slide: ICE restart even if the ICE pokicy is more strict
[13:12:44] Pierre-Luc Gregoire joins the room
[13:12:52] Michael Froman joins the room
[13:13:28] <m&m> slide: Remove the 'public' ICE candidate policy
[13:15:27] <m&m> slide: Reference draft-rtcweb-ietf-ip-handling
[13:15:49] <m&m> slide: Clarify a=imageattr handling
[13:16:55] Ray Atarashi_6845 joins the room
[13:17:24] <m&m> slide: Add more detail on VoiceActivityDetection handling
[13:18:08] <Paul Kyzivat> I just lost audio/video
[13:18:20] <Wolfgang Beck> audio still fine for me
[13:18:31] Paul Kyzivat leaves the room
[13:18:51] Paul Kyzivat joins the room
[13:19:09] <Wolfgang Beck> (video is fine, too)
[13:19:16] <m&m> slide: No longer required to support RTCP MUX 'negotiate'
[13:19:54] Ray Atarashi_6845 leaves the room
[13:20:07] <m&m> slide: OPEN ISSUES
[13:20:20] <m&m> slide: #235 Nail down specifics of configurable SDP params
[13:21:55] <Randell Jesup> SDP munging Must Die!! ;-)
[13:22:40] <Wolfgang Beck> Using SDP was a bad idea in the first place
[13:23:14] <Wolfgang Beck> SIP SDP interoperability was lost long ago
[13:23:29] Ray Atarashi_2960 joins the room
[13:23:59] <m&m> reminder: if you want something said in the room by !yourself, please prefix it with "mic:"
[13:28:22] ken carlberg joins the room
[13:29:23] Chris Wendt leaves the room
[13:29:32] Chris Wendt joins the room
[13:30:19] Ray Atarashi_2960 leaves the room
[13:31:18] <Jonathan Lennox> We don't hear you, Randall
[13:31:22] Ray Atarashi_5301 joins the room
[13:31:35] <Wolfgang Beck> https://self-test.conf.meetecho.com/self-test/
[13:31:55] <Jonathan Lennox> I assume other remote people also did not hear Randell?
[13:32:04] <Meetecho> I don't think Randell really needs to do the self test :)
[13:32:07] <Wolfgang Beck> I couldnt hear him
[13:32:09] <Victor Pascual> I did not
[13:34:29] <Wolfgang Beck> audio is good
[13:34:38] <Jonathan Lennox> Meetecho: odd bug, Randell is both talking and in the queue on the screen we see.
[13:34:51] <Meetecho> UI bug, nevermind that...
[13:34:57] <Jonathan Lennox> OK
[13:37:34] Ray Atarashi_4196 joins the room
[13:37:46] Ray Atarashi_5301 leaves the room
[13:37:50] Randell Jesup leaves the room
[13:38:52] Randell Jesup joins the room
[13:41:00] <m&m> slide: #72 Maintain DTLS role during SDP renegotiation
[13:43:52] <m&m> slide: #149 Limitation on SSRC change is in violation with RFC7160 regarding RTP PT change
[13:44:58] <m&m> slide: #223 Specify demux algorithm for RTP
[13:45:34] <m&m> slide: #162 LipSync groups violating RFC5888
[13:47:06] <m&m> slide: #206 support explicitly disabling VAD in RTPSender
[13:48:22] John Leslie joins the room
[13:49:00] Michael Froman leaves the room
[13:49:21] Michael Froman joins the room
[13:50:21] Ben Campbell leaves the room
[13:50:25] Randell Jesup leaves the room
[13:51:16] Randell Jesup joins the room
[13:51:34] <m&m> slide: #247 Document what should happen when there are no matching codecs in answer
[13:52:47] <Randell Jesup> Justin meant to say "answerer should set port 0"
[13:53:45] Ben Campbell joins the room
[13:55:58] <m&m> slide: #249 Clarify a=rtcp processing
[13:56:19] <m&m> slide: #250 explain how to roll back an ICE restart
[13:56:49] <Paul Kyzivat> Johathan: that is not what O/A means. If X offers A and B, and Y answers B, then Y MAY send B.
[13:57:18] <Jonathan Lennox> RIght, but Y can't send A if it wasn't in the answer.
[13:57:54] <Paul Kyzivat> Yes he can.
[13:58:04] <Jonathan Lennox> Oh, he can? Huh.
[13:58:12] <Jonathan Lennox> Been a while since I read 3264.
[13:58:53] <Paul Kyzivat> I'm pretty sure.
[13:59:22] <Jonathan Lennox> We should presumably make sure of this.
[13:59:55] Maxime Marchand_1737 joins the room
[14:00:05] Maxime Marchand leaves the room
[14:01:09] <Wolfgang Beck> Mhmm. RfC 3264, Section 7: Offerer Processing of the Answer...It MUST send using a media format listed in the answer,    and it SHOULD use the first media format listed in the answer when it    does send.
[14:02:18] <Jonathan Lennox> Wolfgang: to be clear, "it" there in that sentence is the answerer, yes?
[14:02:35] <Wolfgang Beck> it=the offerer
[14:02:45] <Jonathan Lennox> We're trying to figure out what the answerer can send.
[14:03:03] <Jonathan Lennox> Can it send with anything that was in the offer, or only the codecs that were also in the answer?
[14:03:21] Ray Atarashi_4196 leaves the room
[14:03:21] Ray Atarashi leaves the room
[14:04:29] <Paul Kyzivat> I don't have the rfc in front of me. But from what you quote, the section is discussing the offerer, and so "it" applies to the offerer.
[14:06:49] <Wolfgang Beck> AFAIK the main point of O/A was to ensure that both sides use the same codec. This used to be a requirement for some media gateways which could only activate one codec at a time.
[14:07:32] Ray Atarashi joins the room
[14:07:35] <Wolfgang Beck> with plain sdp exchange, you could send using any codec the offerer understands
[14:09:01] <Victor Pascual> that's my understanding as well
[14:09:08] Ray Atarashi_3888 joins the room
[14:09:15] <m&m> slide: #239 When is the SDP sess-version in o= line updated
[14:10:12] <Wolfgang Beck> but it seem as if Paul is right (as usual in these matters): 5.1 ..If multiple formats are listed, it    means that the offerer is capable of making use of any of those    formats during the session.  In other words, the answerer MAY change    formats in the middle of the session, making use of any of the    formats listed, without sending a new offer.
[14:10:38] <Jonathan Lennox> MInd you, sending with a codec that wasn't in your answer is a bad idea, because as soon as you send a re-offer in the other direction, assuming you don't but that codec in your re-offer, you'll have to switch away from it. (Because that codec can't be in the answer to the re-offer.)
[14:12:54] <Paul Kyzivat> Interesting point.
[14:12:59] Jan-Ivar Bruaroey joins the room
[14:17:11] ken carlberg leaves the room
[14:23:27] Alissa Cooper leaves the room
[14:24:40] Alissa Cooper joins the room
[14:28:03] Tim Panton joins the room
[14:28:44] <m&m> slide: #261 document API points for RtpSender/RtpReceiver
[14:30:29] Chris Wendt_4962 joins the room
[14:30:45] <Maxime Marchand_1737> hummmmm
[14:30:47] <Paul Kyzivat> Jonathan: thinking more, your point about bad idea is not right. If the reoffer from Y includes only B, the reanswer from X MAY also include A.
[14:31:11] Chris Wendt leaves the room
[14:31:16] <Jonathan Lennox> Really? I thought answer codecs had to be a subset of the offer codecs.
[14:31:56] <Paul Kyzivat> no, You can mention others that you support, but *you* won't be able to send them whitout another O/A.
[14:32:10] <Jonathan Lennox> Oh, hm.
[14:32:42] Maxime Marchand_1737 leaves the room
[14:32:53] Maxime Marchand joins the room
[14:33:13] <Jonathan Lennox> You seem to understand this better than I do.  Can *you* comment on JSEP issue 247?
[14:33:15] <Jonathan Lennox> https://github.com/rtcweb-wg/jsep/issues/247
[14:33:59] <Randell Jesup> hum for remove munging.  Kill it with fire!
[14:34:00] <Paul Kyzivat> I haven't been following this stuff. *Where* should I comment on it?
[14:34:00] <Maire Reavy> hum to remove SDP munging please!
[14:34:18] <Varun Singh> hummmmm
[14:34:25] <Jan-Ivar Bruaroey> hummmmm
[14:34:27] <Varun Singh> for removing munging
[14:34:29] <Jonathan Lennox> Paul: On the GitHub issue tracker URL i pasted.
[14:34:33] <Michael Froman> hmm for removing
[14:34:39] <Jan-Ivar Bruaroey> for removing
[14:34:47] <Paul Kyzivat> OK. I'll take a look.
[14:34:48] hta joins the room
[14:36:43] <Wolfgang Beck> webaudio modem: https://cho45.stfuawsc.com/WebAudio-Modem/FSK/modem.html
[14:37:09] <Randell Jesup> Agreed, "need not"
[14:37:25] <Randell Jesup> "Are not required" - etc
[14:42:19] <m&m> slide: slalom down the RT mountainside
[14:44:26] Alissa Cooper leaves the room
[14:44:27] hta leaves the room
[14:46:44] <Varun Singh> :) the mountain becoming shorter.
[14:48:52] Ben Campbell leaves the room
[14:48:53] Ben Campbell joins the room
[14:48:55] frodek joins the room
[14:49:13] ted.h leaves the room
[14:49:56] m&m leaves the room: Disconnected: Replaced by new connection
[14:49:56] Maxime Marchand leaves the room
[14:49:57] Maire Reavy leaves the room
[14:50:00] <Varun Singh> early media?
[14:50:06] m&m joins the room
[14:50:26] magnus leaves the room
[14:50:29] Maxime Marchand joins the room
[14:50:41] Jonathan Lennox joins the room
[14:50:49] Jonathan Lennox leaves the room
[14:51:01] <Tim Panton> I’m really edgy about this. From a software engineering POV, we are smearing the DTLS implementation all the way up and down the huge webRTC stack which has to imply verifying security is more difficult.   We are also exposing all the DTLS handshake to javascript manipulation, which is a whole order of magnitude easier than redirecting the media and then intercepting it on the fly.
[14:51:29] Maire Reavy joins the room
[14:51:58] <Tim Panton> Neither of these problems crop up if you put the packets into the ICE exchange.
[14:52:02] magnus joins the room
[14:52:17] Chris Wendt_4962 leaves the room
[14:52:17] <Tim Panton> - I care about the long term key.
[14:52:23] <Wolfgang Beck> today, clients can determine if the server fiddled with the fingerprint
[14:52:38] <Randell Jesup> tim: you want to put mic: in front of that?
[14:53:07] Maire Reavy leaves the room
[14:53:18] <Tim Panton> only if we are discussing this?
[14:53:44] Maire Reavy joins the room
[14:53:47] <m&m> Tim: seems to me we are … do you want it relayed?
[14:53:51] <Wolfgang Beck> browser certs are not certs that have been validated by anyone; server can easily fake them
[14:53:55] <Tim Panton> please yes.
[14:54:32] Karthik BR leaves the room
[14:54:43] <Wolfgang Beck> the more I think about it, the worse it seems to me
[14:55:49] frodek leaves the room
[14:59:03] hta joins the room
[15:02:04] Ben Campbell leaves the room
[15:02:42] Ray Atarashi_3888 leaves the room
[15:02:43] Ray Atarashi leaves the room
[15:02:43] Maire Reavy leaves the room
[15:02:46] <m&m> that's a wrap
[15:02:49] Jonathan Lennox leaves the room
[15:02:52] csp leaves the room
[15:03:09] hta leaves the room
[15:03:19] frodek leaves the room
[15:03:48] Wolfgang Beck leaves the room
[15:03:49] Paul Kyzivat leaves the room
[15:03:50] m&m leaves the room
[15:03:53] Michael Froman leaves the room
[15:03:58] Varun Singh leaves the room
[15:03:59] John Leslie leaves the room
[15:03:59] Victor Pascual leaves the room
[15:03:59] Maxime Marchand leaves the room
[15:03:59] Randell Jesup leaves the room
[15:04:00] Pierre-Luc Gregoire leaves the room
[15:04:08] Meetecho leaves the room
[15:04:28] Tim Panton leaves the room
[15:04:59] Jan-Ivar Bruaroey leaves the room
[15:12:03] magnus leaves the room
[15:17:22] Ben Campbell joins the room
[15:23:43] Ben Campbell leaves the room
[17:01:15] m&m joins the room
[17:03:53] m&m leaves the room: Disconnected: Replaced by new connection
[17:03:55] m&m joins the room
[17:09:38] Alissa Cooper joins the room
[17:10:51] Alissa Cooper joins the room
[17:11:57] Alissa Cooper leaves the room
[17:19:53] m&m leaves the room: Disconnected: Replaced by new connection
[17:19:54] m&m joins the room
[18:00:07] m&m leaves the room
[18:00:12] m&m joins the room
[18:01:01] Alissa Cooper leaves the room
[18:04:35] m&m leaves the room
[18:04:54] m&m joins the room
[18:28:55] m&m leaves the room
[20:12:51] Neustradamus joins the room
[20:19:44] magnus joins the room
[20:39:36] frodek joins the room
[20:44:42] frodek leaves the room
[21:44:02] hta joins the room
[21:48:10] hta leaves the room
[21:54:04] hta joins the room
[22:02:32] hta leaves the room
[22:10:12] magnus leaves the room
Powered by ejabberd - robust, scalable and extensible XMPP server Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!