Re: [rtcweb] [MMUSIC] BUNDLE: RTP only?

A question: 

Since we have a clear way of de-multiplexing RTP and DTLS, is it necessary to entangle that (solved) problem with the A/V mux issue? I realize that SDP has the tendency to "deeply intermingle" just about everything, but at the RTP/DTLS level, we're in much better shape.

Bernard "Do we need to boil the entire ocean, or just a lake or two?" Aboba

On Feb 13, 2013, at 13:49, "Eric Rescorla" <ekr@rtfm.com> wrote:

> I would phrase this somewhat less strongly: less ports is better than more ports,
> so all things being equal, a multiplexing mechanism that keeps you down to
> one port is better than not. That said, if we decide that plan A is dead and
> we are going to plan B, I can live with having data on a separate port.
> 
> -Ekr
> 
> On Wed, Feb 13, 2013 at 12:42 PM, Randell Jesup <randell-ietf@jesup.org> wrote:
>> On 2/13/2013 3:37 PM, Cullen Jennings (fluffy) wrote:
>>> I think EKR expressed that he felt it was important for data channel to be included in the multiplexing.
>> 
>> And I agree(d), strongly.  We want 1 port to minimize NAT issues and startup times.
>> 
>> 
>>> 
>>> At the interim, during my BUNDLE presentation, I indicated that we need to make a decision whether the solution (whatever it will be...) shall support bundling of non-RTP media. Not saying it can't be changed, but as currently documented all alternatives only support bundling of RTP media.
>>> 
>>> There might have been some comments from the mic, but as far as I know there was no decision (if there was, please point me to it).
>>> 
>>> So, my question is: do we keep the scope to RTP media?
>> 
>> -- 
>> Randell Jesup
>> randell-ietf@jesup.org
>> 
>> _______________________________________________
>> 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