Re: A proposal for solving non-muxed RTCP *and* ICE freezing

As far as I know, nobody does non-mux which supports ICE, DTLS and SRTP 
exchanges as per WebRTC standards. Having said that, I can't say my 
personal knowledge of what people do in the wild is exactly a 
comprehensive study.

When it comes right down to it WebRTC 1.0 does it so to achieve parity 
in capabilities, we support as well but we try to minimize the pain of 
exposure. If WebRTC 1.0 drops that as a requirement, I suspect our 
group's reaction would be to drop it as well.

I'm not sure there's a strong use case to need non-mux RTP/RTCP, except 
for "compatibility / parity" with WebRTC 1.0, or perhaps to allow RTCP 
meta data to flow through a proxy that knows the RTCP keying material 
whereas RTP keying material would be kept private. That latter use case 
was not a direct intended supported use case though in ORTC.

-Robin

> Iñaki Baz Castillo <mailto:ibc@aliax.net>
> May 5, 2014 at 7:59 PM
>
> Hi,
>
> I don't fully understand. WebRTC 1.0 clients (say browsers) MUST 
> implement rtcp-mux. Servers (which are not standardized by the WebRTC 
> 1.0 spec) may or may not.
>
> Anyhow, is there any media server that implements DTLS-SRTP and allows 
> non rtcp-mux (which means two separate ICE, DTLS and SRTP procedures)? 
> If not, which "interoperability" are we talking about?
>
> Thanks a lot for your replies. Regards.
>
> --
> Iñaki Baz Castillo
> <ibc@aliax.net <mailto:ibc@aliax.net>>
>
>