Gonzalo: do we want SIP extensibility at 10PM?

various opinions about this being too late for a can't miss all-hands
meeting, but also opinions that regardless, it's important, and we
need to talk about it now.

hands for have/not have roughly split.  meeting cancelled.



John Elwell: Updates to Asserted Identity
==================================================
open issue 2: allow p-asserted-identity in mid-dialog request?

Can include, but can't change.  Main use case is when transfer/etc
changes.  needed in UPDATE, only in update.

conclusion: leave as is, confirm on mailing list

open issue 4: need more discussion on proxy asserting identity in
responses/conditions under which they can do so


Sam Ganesan: UA Profile Datasets
==================================================

discussion of exclude mechanism

ekr: can't define a policy that prohibits the other end from showing
data to the user (or can't enforce it, at least)

fluffy: need clarification to figure out what it means, what can be
implemented, etc.


Sumanth Channabasappa: UA-profile Event Package for App-profile
==================================================

JDR: What is a SIP Trunk
==================================================

Lots of discussion that people who are selling this know what it is
already, and it's pretty much a marketing term now.  Question is how
to define something useful and consistent that doesn't conflict with
existing marketing.


Hans Erik van Elburg: Explicit Private Network Indication
==================================================

A lot of conversation about basic definitions and purpose.  

Need to be able to define private network and public network to make
progress on this.  Need to know how the originating proxy can
determine that it is a private call.  


Francois Audet: Feature referral in SIP
==================================================

Discussion over whether this makes REFER be a "SIP DO" method?  But is
this necessary in say a call center environment?  

Interesting comments made regarding possible extension to haptic
interfaces.


draft needs to reflect why this should be done in SIP and why it is
reasonable to do it with REFER.


Scott Poretsky: Benchmarking
==================================================

Rohan: importance of media in performance of SBCs (and others).  need
to consider difference between best practice and current practice

ekr: baseline assumptions that aren't specified by current draft are
missing and may be orders of magnitude significance depending on how
handled.

what exactly does 3261-compliant mean?


other comments about how to specify exactly what is being measured and
how


Alan Johnston: Secure media 
==================================================

discussion over how to combine security requirements with need to do
monitoring

poll: interest in this?  strong hum in favor