[AI] Michael: contact Henning Schulzrinne and Doug Sicker about how SIP might be used to implement the use cases.

[AI] Peter: will initiate discussions of the use cases in the XMPP forums.

[AI] Peter: will review the SAML 2.0 specs, and initiate discussion around Jabber and SAML on Jabber forums.

[AI] Klaas: when he receives it, he will forward the write up on how A-select might be Jabber enabled.

[AI] Brendan, John Paul, and Klaas: will circulate the use cases in their organizations to see what kind of responses/issues are raised.

[AI] Michael: will circulate the use cases to the other working group chairs for feedback.

[AI] Neal: will review the use cases and determine if there is anywhere appropriate he can forward them.

I2IM Conference Call July 9, 2004

*Attendees*

Michael Gettes, Duke

Brendan Bellina, Notre Dame

John Paul Robinson, UAB

Peter St. Andre, Jabber Software Foundation

Klaas Wierenga, SURFnet

Terrie Clark, Internet2

Jeanette Fielden, Internet2

Neal McBurnett, Internet2

Steve Olshansky, Internet2

Discussion

Brendan has circulated the pseudonymous use cases to the list for final call. No changes were received so the use cases are accepted by general agreement. The next step is to tease out the requirements into a companion document and start discussions to figure out what technologies are needed to support the scenarios.

Peter can take the use cases and circulate them within the Jabber community to initiate discussions on how they might be implemented. Michael can approach people in the SIP space about the use cases.

A common thread in the use cases is SAML. The list can be used to share information on how SAML can be used in SIP, XMPP etc. and to develop a plan for moving forward. Information can be transferred into a document format as needed down the road.

John Paul indicated he is trying to analyze if what he is considering doing is a different approach from what others are doing. A generic anonymous handle may have broader applicability. He is working on a description of his approach and will send it to the list to generate discussion as it progresses.

Steve indicated that it may be possible to arrange space to experiment with a Shibboleth enabled Jabber server with Internet2.

Klaas has commissioned a study of how A-select (www.a-select.org) could be connected Jabber.

What does it mean to Shibboleth enable Jabber? How might one engineer the client and server in the XMPP universe to get a SAML assertion XML and send it along? Peter indicated that those conversations haven’t occurred yet in the Jabber universe, and that getting that conversation going is the next step. Peter forwarded to the group the address and subscription information where discussion occur within the Jabber community so people can participate in those discussions if they wish.

Michael requested that people circulate the use cases on their campuses to get feedback on what kind of issues they raise, and does it address expressed concerns. Brendan indicated he’s spoken with the messaging services team at Notre Dame about the cases but their main concern was demand for the service. He will also forward it to the campus CTO for comment.

Neal asked about the persistence of the id in the pseudonymous use cases. If needed could you go back to institution and make case for revealing the id? Michael indicated that it was feasible from a technical perspective but there are a number of policy issues regarding legal discovery and usage of such information. SAML 2.0 may have some things built into it to address the notion of eduPersonTargetedID, so whether that would survives or morph into something else is unknown at this point.