Minutes of the SIPREC WG Session at IETF 77,
Anaheim, CA, USA, 2010-03-23, 10.30-11.30 GMT-7
===============================================
Meeting chaired by Brian Rosen and John Elwell.

Minutes produced by John Elwell based on notes from Christian Schmidt and Roger Marshall.

Jabber log: http://www.ietf.org/jabber/logs/siprec/2010-03-23.txt

Topic - Administrivia (chairs)
==============================
http://www.ietf.org/proceedings/10mar/slides/siprec-0.ppt

No changes to published agenda.

Topic - Charter (chairs)
========================
The chairs summarised the charter for this newly-approved WG.

Topic - Requirements (Ken Rehor)
================================
http://www.ietf.org/proceedings/10mar/slides/siprec-1.pptx

During his presentation, Ken mentioned the fact that new use cases continue to be identified, 
and that it will be necessary to limit the scope of initial deliverables.

Somebody asked whether we have enough time or expertise to adequately treat security, and why
do we need this section? The chairs responded that security must be covered in all
deliverables, especially protocols, but also requirements. Andy Hutton pointed out that the
requirements are not addressing specific mechanisms.

From the chair, John pointed out that we have to identify the really important requirements 
- what are the short term and what are the long term requirements.

Topic - Architecture (Andy Hutton)
====================================================================
http://www.ietf.org/proceedings/10mar/slides/siprec-2.pptx

Concerning slide 5, media mixing bullet, it was questioned what is the problem with SDP
if streams from the two speakers are to be transmitted separately. Discussion pointed out 
that multiple m-lines could be used, although consideration would need to be given to the 
interaction between multiple m-lines for the different speakers and multiple m-lines for 
multiple media.

Hadriel Kaplan queried the exclusive usage of RTP as transport for recorded media, since 
it may not be a true RTP layer, literally just packet-tracing. He also pointed out that 
in the original draft there was a requirement to allow just packet copying, not 
translating/mixing. Also Cullen Jennings questioned why we need RTP. Andy Hutton said 
the requirement had been removed. This issue should be taken to the list.

Hadriel Kaplan also raised a question concerning notification requirements and the SIP 
layer. This should be taken to the list.

Somebody pointed out the similarity between this and lawful interception. The chair reminded 
participants of the charter statement on interception.

On the last slide, Hadriel Kaplan pointed out that some of the open issues related to the 
solution, and shouldn't apply to an architecture document.

It was noted that we may have to look at other groups for some aspects of the solution, 
e.g., MMUSIC, MEDIACTRL, XCON. An opinion was expressed that many of the requirements 
can in fact already be met by existing protocols. However, Brian Rosen, as chair, 
pointed out that this group is chartered for SIP-based media recording, so entirely 
different solutions are out of scope. Gonzalo Camarillo asked for any pointers to existing 
solutions to be posted to the list.

Summary (Chairs)
===============
John Elwell said that we should progress the requirements draft first, but was not sure that 
it is ready for WGLC yet. The WG has only been open for one week, and we need to get more 
list discussion, and to get consensus.  Much of the work has been going on within the design 
team, and discussion needs to move to the list. Let’s have some good discussion on the 
requirements and architecture documents. If it seems appropriate, we might hold an 
interim meeting (conference call) before IETF#78. The design team continues to hold a weekly 
conference call - others are encouraged to join, and the team should of course keep the list 
informed.