*VidMid-VoD Conference Call*
December 5, 2001

*Attendees*

Mairead Martin - (acting chair) U. Tennessee, ViDe
Ken Klingenstein - Colorado/Internet2
Ann West - Internet2
Steve Olshansky - Internet2
Jim DeRoest - U. Washington
Jill Gemmill - SURA, ViDe
Ellen Vaughan - Internet2
Mark Driver (scribe) - Internet2

*Agenda*

1. Chapel Hill VidMid-VC meeting report
2. DRM project
3. Draft Scenarios
4. VidMid-VoD information dissemination

*Discussion*

1. Steve O. described the Chapel Hill meeting as very successful. Involvement of NSF helpful for visibility and prioritizing VidMid-VC initiatives. Much time devoted to work plan and providing concrete deliverables. Dissemination of VidMid-VC info, goals and accomplishments important. Test-bed/Cookbook deemed necessary for 1) instruction on how institutions can get testbed programs started and 2) codifying work of group thus far. N.C. made excellent use of time and resources; impressive collection of brain power in one room. Thanks to Tyler Johnson for hosting. More details from Steve to be sent to list.

Discussion of overlap between VC and VoD: rights management, authentication, authorization and directory synergies noted. Is VC merely two-way VoD, or vice versa? Agreed that maturity and traction in VidMid-VoD will lead to engagements/projects between groups. Strong interest expressed re: Tyler's mapping of VC scenarios onto architecture reference model. Seen as good model for VoD group to organize work plan. Possibility of modifying Tyler's VC reference model for VoD use: extract VC endpoints and insert/overlay VoD endpoints. Several relevant components: identity directories, policy servers, DNS, etc. Shibboleth is viable as underlying security for both.

Ken gave Shibboleth update: code writing is progressing. Integration of 3 coding teams in next couple of weeks. Hope to have working demo for ALA event in January with access control mechanisms/screen sequence; alpha site ready in February. Ann West mentioned she had initial screens done.

VC issues shared in VoD re: rights manager, access control, enforcement, which servers accessed. CNI example: permission to access 20 second clip w/ plug-in structured for short edits. Will require significant work, possible N.C.-like meeting to hash out details.

Discussion of N.C. meeting proposal of object super class in VC group to support H.323 vs. SIP protocols. Resolution of competing signaling/videoconferencing standards through creation of object class overlay. Broadened to "communications super class" to encompass all media types.

Discussion of NMI release 1. Extra credibility for VoD. Possibility of using directories for non-people based entities. Internet2's eduPerson vs. eduOrg attributes. GRID efforts in this area; VoD may have to adopt by default unless scenario development indicates need for a different approach. Sparked discussion of database vs. directory issues.

Lightweight directory architecture suggested with more robust database backend. Discovery performed in directory. Policy contained in database. How to point from directories to databases. Bigger directories require more metadata; more hands-on for user if databases handle more. [A.I.] White paper suggestion: role of directories in VoD. Target date: March/April 2002. Mairead, Jim DeRoest take lead.

2. DRM Project. Group is interested in rights metadata. Ken has talked to Mellon about possible funding for a DRM implementation. Perhaps submitted to wrong area, with wrong spin or emphasis. Need Close coordination with Mellon on project/grant proposal. DRM proposal supported by CNI, Internet2, others? Discussion of Secure E-content Attribute Management (SEAM) project.

Mellon seems to like approx. 3 lead institutions w/ other associated members to balance distribution vs. accountability. Some discussion of possible roles by usual participants. Part of effort would include interfaces to parallel initiatives that are proprietary/unique. Research Channel mentioned, other commercial efforts should be monitored/engaged.

Time is of the essence. Mellon Foundation generally moves quickly (3 board meetings/year -- next: April). VoD is in talks; Ken hopes to firm up schedule by next week. Suggestion that draft proposal be submitted soon to receive feedback and align goals. URL for SEAM's DRM model at http://www.ait.utk.edu/projects/seam/seamintro.pdf. Get in touch with Mairead to get involved.

Related efforts in Internet2 and GRID re: similarities between federated administration (Internet2) and virtual organizations (GRID). Issues such as registries, unique names, security keys for assets, embedded credentials, alliances, purchasing clubs, access controls, etc. mentioned.

What content would be compelling for grant? SHOAH holocaust victim testimonials likely candidate because of Mellon Foundation's perceived humanities focus. SHOAH technical limitations may pose problem. More emphasis on capturing testimony than cataloging, hence little metadata available now. Technical infrastructure modest; unknown policy decision points and servers, high-performance connectivity issues as well.

Show list of possible content to Mellon to test waters. Need not be video only; text and audio permissible to supplement video content. Tie in other Mellon content and/or humanities-oriented community. MIT OKI courseware management suggested. JSTOR, ARTSTOR other candidates.

Discussion of security problems of portals. Where to store attribute release: at portal itself, or portal is transparent. Same issues with Shibboleth and DRM. Is architecture to enable portals and courseware similar? Present DRM courseware issues to Mellon? [A.I.] Mairead will follow up with answer on listserv.

3. Scenarios. Call participants were more middleware-oriented than VoD-oriented so this section is abbreviated. Questions about usefulness of scenarios for N.C. VC discussions. Agreement that they were good references, but a little too H.323-focused. Tyler's overlay of scenarios on VC model deemed extremely valuable. VC scenarios require more specificity to aid this effort. As mentioned earlier, a similar effort would be helpful for VoD as well. Scenarios would also be good for Mellon proposal. Scenarios in general had utility as jumping off point for extraction of specifications. Interoperability scenarios envisioned for H.323 and SIP; similar issues for VoD and DRM communities and standards.

4. Dissemination of VoD info. Discussion on ways to advertise the existence of VoD group and its efforts so as to gain momentum and avoid duplication/leverage synergies of other efforts. Increasing call attendance seen as hopeful side-effect. Mention of Educause 2002 call for proposals and creation of ViDe track at Educause to expand awareness. Would include video and middleware components. Ken mentioned there might be too much territory to cover on this track. He mentioned monthly "thought for food" sessions where discussions often get pushed to other lists. Videoconferencing of various events in general should also increase interest. Expansion to broader community, Educause 2, national and regional conferences. Opportunities on ViDe side to disseminate info.

Ken referenced post-9/11 virtual Internet2 member meetings (a.k.a. vimmlets) via videoconference as dissemination vehicle. Next vimmlet scheduled for March.

Attendance on VoD conference calls expected to rise once momentum increases. Wait for verdict until after holiday season has passed and school is back in session. DRM is a key research area, people may be over-extended. Overlap not yet recognized with other projects. Often takes one committed individual/group to sponsor integration initiative. Some may be pursuing less ambitious projects. VoD still having somewhat esoteric discussions; once implementation begins people will jump on wagon. Setting modest goals not appropriate; must address tough tasks (i.e. position paper for directories).

Jim DeRoest mentioned grant from public broadcasting consortium to U. Washington re: DRM. Parallel effort in School of Information Sciences with no knowledge of Jim's group. Also, 'Convergence group' at Carnegie Mellon. Illustrates need for better dissemination of VoD initiatives so efforts can leverage one another in same institutions. Perhaps focusing on different aspects of problem or time schedules (i.e. conceptual model vs. implementation stage), but improved coordination would behoove all.

METS has gap in area of research re: DRM. Need strategy for continuing VoD momentum. Other dissemination avenues: campus middleware architects conference in Tempe early 2002. ViDe group to augment video side of issue.

NMI involvement -- advertise momentum. Backing of CNI and Internet2. Educause is early sponsor (can put in publication). Must identify and define who VoD is.

[A.I.] from North Carolina meeting: question about how Educause will disseminate information on NMI. Educause track for next year?

Ann suggested discussing after knowledge of Educause involvement in VoD. Know by end of next week. Must determine fit of NMI with VoD and determine avenues of dissemination.

*Action Items*

[A.I.] VidMid VoD scenarios to list
[A.I.] Send update of DRM to list
[A.I.] Send Internet2 ViDe group invite
[A.I.] Work on Mellon proposal (comments to Mairead requested)
[A.I.] NMI2 document with VoD mention - year 2 solicitation
[A.I.] Collaboration on directory vs. database issue
[A.I.] Find interesting content for Mellon proposal (& others: NSF, NMI2?)

Next call: Wednesday, December 19th, 12:00 noon EST