Attending:
-------------
John Loughney
Gonzalo Camarillo
Jon Peterson
Brian Rosen
Joerg Ott
Jonathan Rosenberg
Dean Willis
Henning Schulzrinne
Rohan Mahy
Allison Mankin


Question: Are we trying to change something or just register our
dissatisfaction? It seems like we will be able to make some changes, and
more ammunition would help. Many feel that there is a lack of
understanding as to what the issues are and what the interoperability
problems will be. Even though we as individuals may have been telling
them this, there are many who either don't hear, or who may dismiss this
as internal disagreement within 3GPP. 

Question: Is it possible to relegate some of the "wrongness" to
optional-to-implement?

Question: Who is it that doesn't care about interop and why? It appears
to be operators who want to keep a fully closed network. Our challenge
is to design things such that networks may be closed, but that if they
are opened, that interoperability becomes possible. The trick is to do
this without telling them that heir business model is broken. Instead,
we may be able to "go up a level" and say "In the old system, you
achieved this goal in manner X -- and although the obvious corollary is
Y, the approach Z works better and here is why." 

Suggested Approaches:
1) Stuff in 3GPP that's basically wrong and violating spec, need to be
brought forward immediately.
2) BCP about "how you should be using SIP". There's no guidance on how a
service provider should use the protocol.
3) Provide guidance in a broader sense as to how we can have a happy
coexistence of different networks using SIP.


Comment: We've considered documenting the motivation for SIP design. We
could build from the "how the B2BUA breaks things". Need something on
the order of 10 pages, maybe a SIG-COMM position paper.

Comment: We could start with a liason which says "Please consider
changing these specific X things that you are doing because they will
hurt interoperability and have other issues as documented herein. It is
not in the spirit of our specifications to use them as you have
described.

Comment: We need to prioritize, put out the big fires in release 5, help
to bring educated voices to the fore in release 6.

Discussion of SDP editing case: This specific example does not help in
any significant way and breaks things you may care about. As an
alternative, there is a SIP-architecture compliant approach which meets
your goals that does not have the negative consequences . . . 

Response:

1) Quick-hit liason statement covering immediate issues
2) Longer term things

Liason statement issues:
1) Identify a list of issues with specific references to 3GPP documents
    -- start from Dean's List, develop explicit references
2) Turn this into a framework document using a design-team approach
3) 

Task assignments:

1) Dean will edit list of stuff and develop references
2) All will send issues to Dean as they realize them
3) Allison will send reference liason statement
4) Rohan will draft first part of liason statement over weekend
5) Allison will pursue ISOC membership for 3GPP
6) Dean to ask Andrew for guidance on dividing docs
7) We will attempt to meet sometime in Yokohama, TBD, perhaps late
Tuesday