*VidMid-VC Conference Call*
April 8, 2002
*Attendees*
Egon Verharen, SURFnet (chair)
John McNair, U. Tennessee - Knoxville
Steve Olshansky, Internet2
Ernst Heiri, SWITCH (Swiss Academic & Research Network)
Ann West, Internet2/EDUCAUSE
Mary Trauner, Georgia Tech
Art Vandenberg, Georgia State
Jill Gemmill, U. Alabama - Birmingham
Renee Frost, U. Michigan/Internet2
Brent Zionic, U. Colorado - Boulder (scribe)
Ken Klingenstein, U. Colorado/Internet2
Tom Barton, U. Memphis
Ron Tipton, U. Tennessee - Knoxville
Samir Chatterjee, Claremont Graduate University
Doug Sicker, U. Colorado - Boulder
Tyler Johnson, UNC-CH
*Discussion*
The discussion began with consideration of the April 16th deadline to have all NMI deliverables completed. Everyone agreed that we should try to stick to this date. Ken Klingenstein raised the issue of how to label the finished works. Because NMI does not officially endorse formal standards, they will not be labeled as such, but will be presented as recommendations or best practices from the working groups.
This led into the point that we would like to get implementations as widely distributed as possible. It was recommended that we have at least one more SIP testbed environment installed in addition to those at Colorado and Claremont, based on Microsoft's SIP solution.
Egon pointed out that every meeting we have in advance of the NMI release at the end of April needs to focus on the progress and status of the subgroup deliverables. He also advocated greater participation from the Access Grid list members. At the moment, it appears that the Access Grid is more focused on the encryption of the channel than they are on the authentication of the user. Ken stated that a goal of working together is to produce a set of patterned flow documents for the authentication process as done with XML through SAML for both SIP and H.323. This should also include some discussion of the "federated" model.
This led to some discussion of Samir's recently circulated flow diagrams. People who had a chance to look them over generally agreed that they looked good, and Ken thought we might want to focus on the message content a bit more; for instance, how the signaling works when it passes through a gateway from SIP to H.323.
Tyler spoke next regarding his latest drafts of the commObject spec and white paper, and reiterated his request for feedback once more. He pointed out that on Friday, April 12, a new technical advisory group composed of leading thinkers and developers in directory services will have their first call, the MACE-Dir Technical Advisory Board. Tyler recommended trying to pursue joint ITU/IETF standards tracking. Ken hoped that we might be able to solicit further feedback by making all the docs and the latest version of the commObject class available on the web site. Perhaps having a final vetting call for all the subgroups would help, and might be coordinated with the directory group meeting next Monday, April 15th.
Art was next with an update on the Resource Discovery subgroup, and mentioned a goal to carefully consider the solution represented in the Globus Toolkit. They are working to create a high level diagram that includes the minimum components of the enterprise directory, video directory, terminals, and gateways. Tyler had added a basic list of resources that need to be discovered: name & authN, dialing instructions, endpoint configuration information, user-defined resources, and the endpoint types. Tyler made the point that the sequence of discovery varies depending on the actions that are undertaken.
SteveO brought up the fact that document templates and file naming conventions had been agreed upon and standardized, and promised to send that information out to the document authors as needed. Once the white papers are completed they can be sent to him for a final formatting check before being distributed.
Monday May 6th from 1-5 pm there will be an open meeting on VidMid AuthN/AuthZ at the Internet2 Spring Member Meeting in Arlington, VA. From 8-12 there will also be a session on Shibboleth. As for the discussion format, this led us into how to talk about implementations.
Larry at Northwestern and Tyler at UNC have both begun working with the commObject class, but what about the status of other testing centers? NMI will soon announce the selected testbed participant campuses, and they are a logical target, but they are not up and running yet. Aside from that not much has been accomplished in regards to utilizing commObject, perhaps because everyone was waiting for the finalized version. Perhaps we can go beyond the NMI's eight testbed sites. We put together a list of prospective commObject testbed sites, and their respective LDAP infrastructures:
| U. Alabama-Birmingham | iPlanet | |
| UNC-Chapel Hill | OpenLDAP | |
| Ohio State | iPlanet | |
| Northwestern | (not sure, but it is already working) | |
| U. Tennessee-Knoxville | iPlanet & OpenLDAP | |
| SURFnet | OpenLDAP |
*New Action Items*
1. [AI] 8-Apr-02 (All): Review latest revision of commObject spec and
Interrealm Directories white paper (sent to this list by Tyler April 3,
and soon to be posted to VidMid site) and provide feedback to Tyler ASAP.
2. [AI] 8-Apr-02 (SteveO): Setup final VidMid internal vetting calls all 3
subgroups, week of April 15.
3. [AI] 8-Apr-02 (SteveO): post AuthN/Z minutes and latest rev of
commObject spec and whitepaper to VidMid site
*Old Action Items*
1. [AI] 28-Mar-02 (Tyler): notify everyone when the new commObject class
is ready for review. [DONE]
2. [AI] 28-Mar-02 (All): Send comments on Art's White Paper. [DONE]
3. [AI] 28-Mar-02 (SteveO) Clean up old AI's (in progress)