Re: [rtcweb] How to multiplex between peers

Agreed. 3 options. I'd rephrase the third as package only what is available amongst existing protocol, and document what it doesn't do as a result.

But the current discussion is still in middle ground between all of them with no clear direction forward as to which it wants to do.

Regards

Keith

> -----Original Message-----
> From: rtcweb-bounces@ietf.org [mailto:rtcweb-bounces@ietf.org] On Behalf
> Of Harald Alvestrand
> Sent: 21 July 2011 12:06
> To: rtcweb@ietf.org
> Subject: Re: [rtcweb] How to multiplex between peers
> 
> On 07/21/11 12:44, DRAGE, Keith (Keith) wrote:
> > This discussion seems to be hovering in some sort of middle ground.
> >
> > Either 1)
> >
> > It has not decided what transport to use for real time data, in which
> case it should document the requirements for such a transport protocol,
> and then evaluate whether existing solutions meet those requirements,
> versus developing an entirely new protocol.
> >
> > Or 2)
> >
> > I has decided that RTP should be used, in which case it should identify
> the shortcomings (i.e. a set of requirements on how it needs to be further
> developed) and send these to the AVTCORE working group.
> Don't forget
> 
> 3)
> 
> It has decided that RTP should be used, with no changes and no new
> extensions, and is going through the ritual of swearing at the resulting
> shortcomings while figuring out which parts of RTP it can say "we don't
> need them, so it's OK to do things that mean we can't use them" to.
> 
> If this group gets lined up behind an 1-year dependency on another group
> before it can deliver its initial work products, this group is going to
> be irrelevant.
> 
>                Harald
> 
> _______________________________________________
> rtcweb mailing list
> rtcweb@ietf.org
> https://www.ietf.org/mailman/listinfo/rtcweb