Re: On babies and bathwater (was Re: [rtcweb] Summary of Application Developers' opinions of the current WebRTC API and SDP as a control surface)

On Fri, Jul 19, 2013 at 6:10 PM, Cullen Jennings <fluffy@iii.ca> wrote:

>
> On Jul 19, 2013, at 5:56 PM, Peter Thatcher <pthatcher@google.com> wrote:
>
> >
> >
> >
> > On Fri, Jul 19, 2013 at 5:54 PM, Cullen Jennings <fluffy@iii.ca> wrote:
> >
> > On Jul 19, 2013, at 10:18 AM, Peter Thatcher <pthatcher@google.com>
> wrote:
> >
> > >
> > > It's interesting that most or your list of things that needed to be
> solved without SDP (simulcast, FEC, correlation of RTP streams with
> MediaStreamTracks, glare) still haven't been solved for WebRTC even with
> SDP, despite many months (years?) of effort.
> > >
> >
> > Peter, with the exception of Simulcast, which of these do you think has
> not been solved in SDP when not using bundle?
> >
> >
> >
> I asked about SDP, not WebRTC. The thing that is making this take a long
> time is you don't want to use SDP.
>
>
I'm not asking about SDP.  I'm asking about WebRTC.   It's the WebRTC API
I'm concerned with.

 > Has FEC been defined for WebRTC?
> yes, that is defined for SDP O/A
>

FEC still isn't defined for use with the WebRTC API.  I believe it's still
in the long list of things that can't be resolved until Plan A vs. Plan B
is resolved.

> Has glare been solved?
> yes
>
>
In the context of WebRTC, it wasn't at the last IETF.  What has changed
since then?


>  > Has mapping of RTP streams with MediaStreamTracks been resolved (with
> the exception of the "unity plan" which has not yet been approved)?
> This is obviously trivial if not using bundle. Any version of MSID would
> work. It's only complicated by bundle
>
>
Even without BUNDLE, it's still not clear what kind of SDP createOffer
should generate for multiple tracks.  (with the exception of the "unity
plan" which has not yet been approved).


> >
> > I think the whole list I gave is still unsolved/unresolved.
> I disagree. They are only unresolved because we want to add bundle and
> that was a major change to RTP resulting in a bunch of work needing to be
> done to see how that impacted SDP.
>
>
It thinks there's more than just BUNDLE, but you make a good point:
 finishing the WebRTC API currently requires lots of major changes to SDP,
which creates big delays.  In fact, it's the number one things holding back
the API from being done, because major changes to SDP take a really, really
long time.


>
>
> >
> >
> >
>
>