Re: [Stox] I-D Action: draft-ietf-stox-groupchat-05.txt

[Ben Campbell <[e-mail]>]

On Jun 19, 2014, at 5:01 PM, [e-mail] wrote:

>> The architecture treats the MSRP Conference server to be a monolithic thing. The result of this is, in the examples, all the SIP and MSRP traffic go to the sample place. That doesn't match the architecture from RFC 4975 and simple-chat, where the signaling (SIP) is separate from media (MSRP).  simple-chat talks in terms of a conference focus that handles the signaling side, and an MSRP switch that handles the media (MSRP) side.
> 
> For all practical reasons, they do end up in the same place. Otherwise, how would the MSRP switch communicate with the SIP server in order to share the information it knows about the participants or add/remove the media plane slots? Do we need another protocol or mechanism between the two? Even speculating about this possibility would introduce some unnecessary uncertainty for those contemplating implementing this specification.

MSRP and simple-chat allow them to be separate. If people implement stox-groupchat to assume that the focus and switch are always co-located, then they could not interop with a decomposed MSRP conference service. If you think that MSRP conference services should not be decomposed in practice, that might be worth a separate draft :-) (Note that "should not be decomposed in practice" is a very different thing than "have not been decomposed in currently available services.)

As far "unnecessary uncertainty" goes, all it needs is guidance to avoid any assumption that the focus and switch are at the same address. Implementing stox-groupchat with a non-monolithic msrp conference service does not require solving the "how does the focus talk to the switch" problem. I'm not suggesting anything more complex than saying "We are showing these co-located to keep the diagrams simple--but remember that the SDP may point to a different location for the MSRP service."