*VidMid Conference Call*
September 24, 2001

*Attendees*

Ken Klingenstein (chair) - Colorado/Internet2
Michael Gettes - Georgetown
Mary Trauner - Georgia Tech
Egon Verharen - SURFnet
Jim Whitlock - Buffalo
Mairead Martin - Tennessee
Steve Olshansky - Internet2
Ellen Vaughan - Internet2
Tyler Johnson - UNC Chapel Hill
Philippe Galvez - VRVS
Grace Agnew - Georgia Tech
Jill Gemmill - ViDe/Alabama at Birmingham
Art Vandenberg - Georgia State
Bob Morgan - Washington
Renee Frost - Michigan/Internet2
Phil Galanter - NYU
Ben Chinowsky (scribe) - Internet2

*Discussion*

The minutes of the previous meeting were approved without changes.

Due to concerns about travel in the wake of September 11, the Internet2 Member Meeting in Austin is being replaced by a Virtual Internet2 Member Meeting. Much of the call was devoted to a discussion of how to ensure success for the VIMM, given the extremely short time available in which to plan it. The main points made were:

1. Split screens are bad. Having a presenter or moderator control who has the floor, or using voice-activated switching, is good.
2. For slides, picture-in-picture should take precendence over following along on the web; but since it's hard to get presenters to all make the changes required for this to work (e.g. using 32-point type), it's good to have the slides on the web as well, as a backup. The idea of using a commercial web collaboration tool for slides was unanimously rejected. Mary wants to use a Polycom-specific feature for her slides; the group was lukewarm about this.
3. Egon suggested giving someone the authority to bump people off the MCU if they're causing problems.
4. People should not need to pre-register; the system should make it possible to dial in at the last minute.
5. Hand-raising mechanism: chat bad, voice good. A large majority saw the way that things work on VidMid's audio conference calls as a good model for video: everyone chimes in, and the presenter or moderator manages conflicts. Chat or email might be useful for the operators or for someone functioning as a dedicated question-taker, but not for the audiences in the sessions.

Ken noted that many of the 75 people on the vidmid list (which will be retained as an announcement list) have not signed up for either vidmid-vc or vidmid-vod. Egon and Mairead stressed that drafting workplans is the top priority for both groups; they are also making steady progress on the scenarios. [AI] The group will aim to have draft workplans to review at the October 2-5 Virtual Internet2 Member Meeting, to have all scenarios completed by October 15, and to complete a list of requirements drawn from the scenarios by November 15. Jill noted that Microsoft is forming a videoconferencing working group; she has a grad student looking into integrating this group into ViDeNet, and is looking for further contacts at Microsoft. Ken noted that Microsoft's recent embrace of federated administration may mean that they'll be more interested in Internet2's middleware work more generally. Ken also observed that between the sharply increased level of interest in videoconferencing as an alternative to flying, and the fact that videoconferencing is an important focus of the NSF Middleware Initiative, the pressure is on VidMid to deliver something soon in this area. [AI] Ken will send Philippe information on possible funding for integrating H.323 and SIP into VRVS.

Finally, Ken walked the group through a slide set describing a preliminary version of a Shibboleth-based architectural vision for Internet2 middleware. Response was generally positive, with an understanding that more time is needed for review by MACE and subsequent refinement on Ken's side, and review and discussion on VidMid's side. Tyler noted that Ken's slides don't say anything about building a VidMid testbed; there was general agreement that VidMid needs to start thinking about building a testbed. Ken noted that he intends the architecture he's developing to be able to bind to a variety of protocols, such as HTTP, LDAP and a variety of video-specific protocols -- in Tyler's words, "a PAM approach".

*Action Items*

[AI] 24-September-2001 - The group will aim to have draft workplans to review at the October 2-5 Virtual Internet2 Member Meeting, to have all scenarios completed by October 15, and to complete a list of requirements drawn from the scenarios by November 15.
[AI] 24-September-2001 - Ken will send Philippe information on possible funding for integrating H.323 and SIP into VRVS.
[AI] 10-September-2001 - Ken will send the revised why-video-needs-middleware document to Ben, who will edit and post the document and reorganize the site to reflect its structure.
[AI] 10-September-2001 - Jill will incorporate a report on the August 15-16 ViDe workshop into the VidMid workplan.
[AI] 10-September-2001 - All who want to participate in the VidMid subgroups -- videoconferencing (vidmid-vc@internet2.edu) and video on demand (vidmid-vod@internet2.edu) -- will send mail indicating the group(s) they want to join and the email address they want to use on its list, to Lisa Hogeboom (lhogeboom@internet2.edu) with a cc to Steve Olshansky (steveo@luminagroup.com).
[AI] 20-August-2001 - Mairead will write a very short summary of VidMid's defining-VoD discussion.
[AI] 20-August-2001 - Egon, borrowing freely from all of VidMid's scenario work so far, will rewrite the scenarios at http://www.surfnet.nl/innovatie/surfworks/showcase/h20.html in the agreed description-example-needs format.
[AI] 6-August-2001 - Samir will look into the availability and qualities of public SIP servers.