*MACE conference call, March 12, 2001*
Bob Morgan (chair)
Ken Klingenstein (co-chair)
Ben Chinowsky (scribe)
Ken opened the call by re-raising a question that came up at the MACE dinner at the Internet2 Member Meeting in DC: how should OIDs be managed on the campus? Paul noted that at MIT this is done by keeping all the OIDs in a single file, with the three people who have write access keeping each other informed about changes. There was general agreement that about all that can be said is that a central registry of some kind is needed. [AI] Ken will produce a first draft of an addition to the OIDs doc, and pass it on to Ellen or Renee for polishing.
A new middleware group, tentatively named "Mid-Vid", is being formed to address issues around middleware and videoconferencing. The plan is to devote the first call to exploring scenarios of video apps we'd like to enable, the second call to pruning these to the two or three most important scenarios, and the third to discussing with MACE the middleware issues involved in those scenarios. Mary Fran Yafchak, Art Vandenberg, Bob Riddle, and individuals from the Netherlands and Australia will participate in Mid-Vid; April 17 is the tentative date for the first call.
Ken led a discussion of the idea of Internet2, EDUCAUSE and Globus, together with Internet2 Corporate Partners IBM and Microsoft, making a bid under MACE's leadership for the systems-integrator role under the current NSF middleware solicitation. Ken noted that the solicitation is unclear about the definition of the "testbed" the chosen systems integrator is to produce. Michael noted that among people he's talked to there seems to be a feeling that a) NSF doesn't know what it's asking for, so it would be a waste of time to submit a proposal, and b) NSF is looking for a comprehensive array of middleware services, so offering pieces isn't likely to gain approval. There was general agreement that if MACE takes on this role, it should keep its current focus on building up from legacy systems, rather than taking on the task of integrating disparate sophisticated applications, and that for the latter purpose MACE should rely on groups specializing in working with those applications. [AI] All will keep their eyes open for individuals and groups interested in participating in a MACE-led bid for the integrator role under the NSF middleware solicitation.
Michael asked if there was interest in MACE becoming a member of the DMTF, and there was. [AI] Michael will pursue MACE membership in the DMTF.
There was a discussion of the emerging need, given the participation of multiple organizations in projects like the PKI Labs and Shibboleth, for rules to govern intellectual property in Internet2. Ken noted that UCAID is preparing an intellectual-property document, the central idea of which will be that "universities shouldn't have to pay twice, to invent and to use". Intellectual property issues will be discussed at the next UCAID board meeting in mid-April. [AI] Bob will look for IETF intellectual-property material that might be helpful in formulating an Internet2 intellectual-property policy.
Ken reported that Jim Farmer has been hired by JA-SIG to deal with XML, and asked if Internet2 should have someone in Washington working on XML issues. It was agreed that this is not necessary; MACE will consider Farmer its main contact for XML issues.
Ken reviewed a call for demonstrations from the Interop iLab, relayed by Ted Hanss. No interest was expressed; [AI] Ken will inform Ted that MACE is not interested in doing a demo at Interop.
Finally the group reviewed progress on various middleware projects. Bob reported that MACE-Shibboleth had achieved the "beginnings of consensus" in DC; Steven Carmody has documented this emerging consensus in http://middleware.internet2.edu/shibboleth/docs/carmody-shibboleth-flows/draft-carmody-shibboleth-flows-00.html. There was also consensus that it's OK for Shibboleth to move faster than OASIS; Shibboleth can switch to using OASIS's SAML tags once SAML is ready. Bob noted that the online-course scenario still has "a lot of thrashing around to be done"; Ken noted that this will be addressed on a conference call with IBM tomorrow. IBM's Mike Nelson has given Ken the impression he'll back Shibboleth "to the hilt".
Michael got good feedback on DoDHE at the Member Meeting; projects that will depend on it are starting to emerge. HEPKI-TAG is starting a new series of cert-profile calls this Thursday, and is close to issuing a formal dc naming recommendation. In HEPKI-PAG, David is going through the draft campus CP line by line with CREN counsel Dan Burk. The PKI Labs are planning a workshop; Ken noted that Tim Polk is eager to support this. Michael believes that a "critical mass" of schools interested in participating in the Mitretek bridge trial has been achieved, but suggestions for further participants are welcome; participants should be using standard CA software.
[AI] Ken will produce a first draft of an addition to the OIDs doc, and pass it on to Ellen or Renee for polishing.
[AI] All will keep their eyes open for individuals and groups interested in participating in a MACE-led bid for the integrator role under the NSF middleware solicitation.
[AI] Michael will pursue MACE membership in the DMTF.
[AI] Bob will look for IETF intellectual-property material that might be helpful in formulating an Internet2 intellectual-property policy.
[AI] Ken will inform Ted that MACE is not interested in doing a demo at Interop.