Minutes, SIPPING WG, IETF 58

Notes by Brian Rosen and Francois Audet
Minutes edited by Gonzalo Camarillo
Meetings chaired by Gonzalo Camarillo, Rohan Mahy, Dean Willis

Session 1, Tuesday November 11, 2003 0900-1130
====================================

Topic: Agenda

Agenda accepted as previously posted.


Topic: Announcements and Status

Chairs mention that they will provide an updated list of milestones for the working group in the second session. MSCML will proceed as an informational individual contribution. There are still some security concerns with the Q.SIG work, but they will probably be resolved soon. The chairs mentioned the existance of draft-manyfolks-sipping-tiop.

Slides presented reviewing status of working group documents.
          pres-chairs-agendaandstatus-sipping-ietf58.ppt


Topic: Application Interaction

Relevant document:
        draft-ietf-sipping-app-interaction-framework

Slides presented
         pres-rosenberg-app-int.ppt

Discussion led by Jonathan Rosenberg

Issue: how to deliver DTMF. The current draft proposes two different mechanisms, one based on REFER and HTTP for presentation capable UAs and the other based on SUBSCRIBE/NOTIFY for presentation free UAs.
Discussion: It should still be possible to use KPML with other transport mechanisms. We need to think of the security implications of this work. Are servers supposed to be in the signalling path or can they obtain the scripts using some other means, such as the dialog package?
Conclusion: Strong consensus on documenting both mechanisms and provide recommendations on how to use them.

Topic: Event Package for DTMF Signals

Relevant document:
         draft-zebarth-sipping-dtmfad

Slides presented:
          pres-zebarth-dtmf-package.ppt

Discussion led by Joe Zebarth

Presentation: a mechanism to transport DTMF to implement prepaid services is needed soon (end of Q1 2004.) The mechanism proposed (an event package) is complementary to KPML because it moves the processing intensity from the Notifier to app server.
Discussion: The real difference between KMPL and this approach is the number of events per digit.
Conclusion: T1 writes a liaison to the IETF indicating their timeframe. If it is possible to meet their deadlines with KPML, the WG prefers not to duplicate standards track work.

Topic: Key Press Stimulus Protocol

Relevant document:

         draft-ietf-sipping-kpml

Slides presented:
          pres-burger-kpml.ppt

Discussion led by Eric Burger

Issue: is the SUBSCRIBE/NOTIFY mechanism OK?
Conclusion: no onjection to immediate NOTIFY with empty body.
Issue: three ways to specify digit maps.
Discussion: requiring three mechanisms is too much. A negotiation using an Accept-like mechanism could be appropriate.
Conclusion: continue discussions on the list.

Topic: Conferencing Design Team Report

Relevant Documents:
         draft-ietf-sipping-conference-package
         draft-ietf-sipping-cc-conferencing

Slides presented:
          pres-johnston-conferencing.ppt

Discussion led by Alan Johnston

Presentation: Framework needs one more revision to align with GRUU. Conference package needs some work to align with XCON.
Issue: side bars.
Conclusion: It will probably be an XCON document. The framework will talk about them.

Topic: Transcoding Design Team

Relevant documents:
         draft-camarillo-sipping-transc-3pcc
         draft-camarillo-sipping-transc-framework
         draft-camarillo-sipping-transc-b2bua
         draft-camarillo-mmusic-source-sink

Slides presented:
          pres-camarillo-transcoding.ppt

Discussion led by Gonzalo Camarillo

Conclusion: strong consensus on adding milestones for the framework, the conf bridge model, and the 3pcc model. Strong consensus on adopting the framework and the 3pcc documents as WG items for those milestones. The design team prefers to wait until the exploders work is completed before asking the WG to adopt the conf bridge model as a WG item.

Topic: Emergency Calls Design Team

Relevant document:
         draft-taylor-sipping-emerg-scen
Slides presented:
          pres-taylor-emergency.ppt

Discussion led by Tom Taylor

Issue: Intermediaries need to insert location information (e.g., phones without emergency support and phones not recognizing emergency calls.)
Discussion: Should this intermediaries be existing SIP entities (e.g., proxies or B2BUAs) or can they be something special?
Conclusion: more discussions needed.
Comment: NENA has two WGs working on these issues and will provide SIPPING with requirements.
Comment: hearing impaired users need to be taken into account.

Topic: Event Package for User Configuration Profiles

Relevant document:
         draft-ietf-sipping-config-framework
Slides presented:
          pres-petrie-config.ppt

Discussion led by Dan Petrie

Issue: can we use XCAP?
Discussions: There are other protocols that could do the job as well, and the initial idea was to be protocol agnostic.
Comment: Are we talking into consideration wireless issues?
Issue: which security mechanisms need to be mandatory to implement?
Conclusion: Organize a conference call and report to the list.

Topic: Intermediary Session Policies in SIP

Relevant documents:
         draft-ietf-sipping-session-policy-req
draft-hilt-sipping-session-indep-policy
draft-hilt-sipping-session-spec-policy
Slides presented:
          pres-hilt-session-policy.ppt

Discussion led by Volker Hilt

Issue: two-pass model vs. three-pass model.
Conclusion: two-pass model is the way to go.
Conclusion: strong negative on sending the documents to SIP. They still need more work.

Topic: Reason Header for Preemption

Relevant document:
         draft-polk-sipping-reason-header-for-preemption
Slides presented:
          pres-polk-reason.ppt

Discussion led by James Polk

Issue: use a single or multiple namespaces.
Conclusion: use a single namespace.
Conclusion: progress as an individual contribution.

Topic: Location Conveyance Requirements

Relevant document:
         draft-polk-sipping-location-requirements

Discussion led by Brian Rosen

Issue: should we use bodies or headers.
Issue: we need to analyze the security implications of this work.
Conclusion: strong consensus on adding a milestone for this work and to take this document as a WG item for that milestone.

Topic: Transfer Issues

Relevant document:
         draft-petrie-sipping-xfer-issues
Slides presented:
          pres-petrie-xfer.ppt

Discussion led by Dan Petrie

No conclusions.


Session 2, Thursday, November 13, 2003, 1530-1730
=====================================

Topic: Agenda

Jonathan Rosenberg will present NAT scenarios in MMUSIC instead.
Chairs presented an updated list of milestones for the WG.

Topic: Requirements for end-to-middle security

Relevant document:
         draft-ietf-sipping-e2m-sec-reqs
Slides presented:
          pres-ono-e2m.ppt

Discussion led by Kumiko Ono

Presentation: difference between e2m and m2e. e2m requires discovery.
Issue: manipulation of bodies by intermediaries.
Discussions: a re-direction model where the UA is the one that inserts the body might be an alternative, although it may introduce long delays.

Topic: Requirements for e2m and m2e

Relevant document:
         draft-barnes-sipping-sec-inserted-info
Slides presented:
          pres-barnes-sec-int.ppt

Discussion led by Mary Barnes

Issue: shall we look at m2m as well?
Issue: trust model.
Issue: mechanism to challenge intermediaries.
Issue: intermediaries manipulating message bodies.
Conclusion: further discussions in the mailing list needed.

Topic: Role(trait)-Based Authorization

Relevant document:
         draft-peterson-sipping-role-authz

Discussion led by Jon Peterson

Conclusion: strong consensus of making it a WG item.

Topic: On-demand Access Authorization

Relevant document:
         draft-trossen-sipping-ondemand
Slides presented:
          pres-trossen-on-demand-auth.ppt

Discussion led by Dirk Trossen

Discussion: XCAP meets many, but not all the requirements.
Conclusion: the work is relevant, but we need more convincing use cases.

Topic: Early Media

Relevant documentss:
         draft-ietf-sipping-early-media-00.txt
         draft-camarillo-sipping-early-disposition-00.txt
Slides presented:
          pres-camarillo-early-media.ppt

Discussion led by Gonzalo Camarillo

Conclusion: strong consensus on making the early media disposition type draft a WG item.

Topic: IPv4/IPv6 Translators in 3GPP Networks

Relevant documents:
         draft-elmalki-sipping-3gpp-translator
Slides presented:
          pres-camarillo-v4v6.ppt

Discussion led by Gonzalo Camarillo

Discussion: This work should take place in SIPPING with support from v6ops.
Conclusion: group of volunteers will help with this work: Marcus Brunner, Roni Even, Keith Drage, and Dirk Kutscher.

Topic: Exploders

Relevant document:
         draft-camarillo-sipping-exploders
Slides presented:
          pres-camarillo-exploders.ppt

Discussion led by Gonzalo Camarillo

Discussion: this is definitively interesting, but the WG is too busy to tackle more work right now.
Conclusion: continue mailing list discussions.

Topic: Dialog Event Package

Relevant document:
         draft-ietf-sipping.dialog-package
Slides presented:
          pres-mahy-dialog-pkg.ppt

Discussion led by Rohan Mahy

Presentation: many changes introduced in the last revision of the document. It needs substantial review.
Discussion: Rohan Mahy will send in two weeks an example of how people can present that they are busy doing something.
Discussion: need to clarify what people mean by "hold indicator".
Conclusion: further discussions to the mailing list.

Topic: SIP Torture Test

Relevant document:
         draft-ietf-sipping-torture-tests
Slides presented:
          pres-sparks-sipping-torturetests-ietf58.ppt

Discussion led by Robert Sparks

Presentation: lots of changes introduced in the last revision. Review needed.
Conclusion: we gathered volunteers to review all the messages.

Topic: SIP Load Management

Relevant document:
         draft-sparks-sipping-load
Slides presented:
          pres-sparks-sipping-load-ietf58.ppt

Discussion led by Robert Sparks

Conclusion: strong consensus on the fact that this is interesting work. Let us keep in on the list for now.

Topic: Even Notification Throttling

Relevant document:
         draft-niemi-sipping-event-throttle
Slides presented:
          pres-niemi-throttle.pdf

Discussion led by Aki Niemi

Discussion: this is important work for SIMPLE.
Issue: should we send this to SIP?
Conclusion: Folks should send their concerns to the list if they believe this should not be sent to SIP. If concerns are small enough, this will be moved SIP. Otherwise, it will stay in SIPPING longer.

Topic: RTCP Summary Report Delivery

Relevant document:
         draft-johnston-sipping-rtcp-summary
Slides presented:
          pres-clark-rtcp.ppt

Discussion led by Alan Clark

Discussion: scope should be clarified. Does this only apply to VoIP?
Discussion: do entities need to report about both directions, send and recv?


ACTION ITEMS

1) Receive a liaison from T1 and see whether or not KPML can meet their deadlines.

2) Chairs to discuss with ADs adding milestones for the transcoding framework, the conf bridge model, and the 3pcc model.

3) Chairs to discuss with ADs making draft-camarillo-sipping-transc-3pcc and draft-camarillo-sipping-transc-framework WG items.

4) Dan Petrie organizes a conference call on XCAP usage for configuration.

5) Chairs to discuss with ADs adding milestone for the location conveyance requirements.

6) Chairs to discuss with ADs making draft-polk-sipping-location-requirements a WG item.

7) Chairs to discuss with ADs adding milestone for role based authentication.

8) Chairs to discuss with ADs making draft-peterson-sipping-role-authz a WG item.

9) Dirk Trossen to provide the WG with use cases on on-demand access authorization.

10) Chairs to discuss with ADs adding milestone for early media disposition type or use the general early media milestone for this.

11) Chairs to discuss with ADs making draft-camarillo-sipping-early-disposition a WG item.

12) Chairs to make sure that Marcus Brunner, Roni Even, Keith Drage, and Dirk Kutscher help with the IPv4/IPv4 translators in 3GPP networks work.

13) Chairs to have volunteers review the SIP torture tests.

14) Chairs to check whether or not there are relevant concerns on the Event Notification Throttling work. Otherwise, send it to SIP.