Re: [rtcweb] Media neg-eo-tiation and signalling archa-tecture

We also followed the WhatWG proposal as well as possible in our implementation (that you can download and test here: https://labs.ericsson.com/apis/web-real-time-communication/downloads). We're using SDP offer/answers as proposed in the whatwg spec; however there is a separate SDP dialogue per stream (and not per session) to avoid glaring conditions.

The work also generated quite some feedback to whatwg: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2011-June/031950.html

Stefan

________________________________________
From: rtcweb-bounces@ietf.org [rtcweb-bounces@ietf.org] On Behalf Of Cullen Jennings [fluffy@cisco.com]
Sent: Friday, July 08, 2011 4:11 PM
To: Niklas Enbom
Cc: rtcweb@ietf.org
Subject: Re: [rtcweb] Media neg-eo-tiation and signalling archa-tecture

Makes sense - thanks. I'm trying to interpret the spec by looking at the code. I'll stop doing that :-)

On Jul 8, 2011, at 1:58 AM, Niklas Enbom wrote:

> Not sure in which thread I should answer... In Chrome we're trying to implement the whatwg proposal with one intentional discrepancy, and that's the string format pointed out by Cullen. See details in this thread: http://www.mail-archive.com/whatwg@lists.whatwg.org/msg26163.html. There's also non-intentional discrepancies that we know of, which we're trying to remove.
>
> Just be careful when looking at Chrome code to interpret the whatwg spec. We're trying to interpret that spec in the same way as anyone else, and the whatwg author is not involved in the Chrome implementation.
>
> Niklas
>
> On Thu, Jul 7, 2011 at 7:52 PM, Cullen Jennings <fluffy@cisco.com> wrote:
>
> I went and looked at some call flows to try and sort out what is going on in the current released chrome code and as far as I can tell, Bernard you are right, it is passing the data over the HTTP channel. An example of what I see is below
>
> As far as I can tell this is the JSON encoding of XEP-180 used in jingle and a separate JSON encoding of some of SDP. Not sure this matters much one way or the other but wanted to pass on what I had found and I will correct the slides at some point - my apologies for getting them wrong.
>
> Cullen
>
>
>
> {
>   "media" : [
>      {
>         "attributes" : {
>            "candidate" : [
>               {
>                  "component" : 1,
>                  "foundation" : 1,
>                  "generation" : 0,
>                  "ip" : "144.254.150.241",
>                  "name" : "rtp",
>                  "network_name" : "Intel(R) WiFi Link 1000 BGN",
>                  "password" : "OpawIyXLYcJkwxFa",
>                  "port" : "61880",
>                  "priority" : 1.0,
>                  "proto" : "udp",
>                  "type" : "local",
>                  "username" : "90B0CPvC8p8XPMFt"
>               }
>            ]
>         },
>         "label" : 1,
>         "rtpmap" : [
>            {
>               "103" : {
>                  "codec" : "audio/ISAC"
>               }
>            },
>            {
>               "104" : {
>                  "codec" : "audio/ISAC"
>               }
>            },
>            {
>               "9" : {
>                  "codec" : "audio/G722"
>               }
>            },
>            {
>               "102" : {
>                  "codec" : "audio/iLBC"
>               }
>            },
>            {
>               "0" : {
>                  "codec" : "audio/PCMU"
>               }
>            },
>            {
>               "8" : {
>                  "codec" : "audio/PCMA"
>               }
>            },
>            {
>               "99" : {
>                  "codec" : "audio/CN"
>               }
>            },
>            {
>               "98" : {
>                  "codec" : "audio/CN"
>               }
>            },
>            {
>               "13" : {
>                  "codec" : "audio/CN"
>               }
>            },
>            {
>               "127" : {
>                  "codec" : "audio/red"
>               }
>            },
>            {
>               "106" : {
>                  "codec" : "audio/telephone-event"
>               }
>            }
>         ]
>      },
>      {
>         "attributes" : {
>            "candidate" : [
>               {
>                  "component" : 1,
>                  "foundation" : 1,
>                  "generation" : 0,
>                  "ip" : "144.254.150.241",
>                  "name" : "video_rtp",
>                  "network_name" : "Intel(R) WiFi Link 1000 BGN",
>                  "password" : "G4dq4e3f/G7FDJ5h",
>                  "port" : "61879",
>                  "priority" : 1.0,
>                  "proto" : "udp",
>                  "type" : "local",
>                  "username" : "CBkZYp6/DTTMEFeM"
>               }
>            ]
>         },
>         "label" : 2,
>         "rtpmap" : [
>            {
>               "120" : {
>                  "codec" : "video/VP8"
>               }
>            }
>         ]
>      }
>   ]
> }
>
>
> On Jul 6, 2011, at 9:43 PM, Cullen Jennings wrote:
>
> >
> > On Jul 6, 2011, at 12:06 AM, Bernard Aboba wrote:
> >
> >>
> >>
> >> * The last slide now says that WhatWG proposal is low path; this is not true. The WhatWG proposal as I read it is mute on how the SDPs are exchanged, it just hands those SDPs over to the
> >>
> >> application (though high path is indicated: "Communications are coordinated via a signaling channel provided by script in the page via the server, e.g. using XMLHttpRequest.")
> >
> > Hmm - my understanding was only the ICE related information was communicated that way and the SDP related information when across the channel set up by ICE but I could be totally wrong. It's pretty vague in the spec - I got that idea by talking to some of the people involved. Anyways, glad to retract that WhatWG is doing low path and just change it to not clear.
> >
> >
> > _______________________________________________
> > rtcweb mailing list
> > rtcweb@ietf.org
> > https://www.ietf.org/mailman/listinfo/rtcweb
>
> _______________________________________________
> rtcweb mailing list
> rtcweb@ietf.org
> https://www.ietf.org/mailman/listinfo/rtcweb
>

_______________________________________________
rtcweb mailing list
rtcweb@ietf.org
https://www.ietf.org/mailman/listinfo/rtcweb