*MACE conference call, February 12, 2001*

*Attendees*

Bob Morgan (chair)
Neal McBurnett
Michael Gettes
Jim Jokl
Renee Frost
Ken Klingenstein
Paul Hill
Keith Hazelton
David Wasley
Steven Carmody
Ben Chinowsky (scribe)

*Discussion*

The call opened with a brief review of the January 29 MACE call, on which no minutes were taken. On that call the main issue discussed was the need to produce materials that can be used to convey the basic ideas (or "mantras") behind I2-MI to vendors and campus security and policy decision-makers. The basic idea here is to write up some of the things that MACE members are saying over and over already. [AI] Michael will outline some ideas for conveying I2-MI's priorities and general approach to a wider audience. Bob suggested that I2-MI follow the successful information-sharing example set by the Joint Techs meetings, and organize a series of "Joint Architects" gatherings of middleware implementers and researchers. Michael noted Jack Suess' plans for a Middleware Day at UMBC, and Ken noted he has received many requests for help in organizing similar events elsewhere. [AI] Michael will reply to Jack Suess about his plans for a UMBC Middleware Day.

Today is the official release date for eduPerson v1.0; DoDHE is also moving forward. Michael met with John Fowler (Global Education and Research Infrastructure Architect at Sun Microsystems) at the JA-SIG meeting. The paperwork for Sun's hardware donation is done, but the accounting for the software is a little more challenging -- Sun usually does donations at list price, and at $10/DN this comes to $60m for the 6m DNs being provided. Keith has worked out a deal with Metamerge (formerly known as Architech; see http://lab.architech.no/) such that by the end of this week we'll be able to do PUTs over HTTPS. Keith is also working with Art Vandenberg on getting SURA resources.

HEPKI-TAG is launching biweekly mobility and profile-convergence conference calls. Michael is going to Mitretek tomorrow to discuss ongoing work on bridge software. Ken noted that Steve Worona is taking the position of director of policy and networking for EDUCAUSE; there was general agreement that this is likely to be good for PKI. Keith reported that on the PKI Labs calls "the buzz seems to center around policy and authorization languages." It was noted that the PKI Labs calls consist mostly of information sharing, brainstorming and discussion, and that the participants seem pleased with this use of the calls. Ken has proposed that the PKI Labs sponsor a research workshop later this spring.

The Shibboleth project is moving forward. Steven noted that the Shibboleth group has sorted its work into four "buckets", and that a leader has been chosen for each. An all-day face-to-face Shibboleth meeting is coming up at the Internet2 Member Meeting. David's message-flows document is starting to generate some list traffic. Work is starting on defining interrelations between Shibboleth and authentication via OKI/Stellar, JA-SIG/uPortal, OASIS, and campus-specific SSO mechanisms. It was agreed that while it would be useful for MACE-Shibboleth to help OKI/Stellar, JA-SIG/uPortal and OASIS reach consensus on a common SSO mechanism, such an effort should be considered a "sidecar" to Shibboleth, with a higher priority being to ensure that Shibboleth publishes specifications sufficient to enable SSO developers to adapt their systems for use with it. Ken noted that Ira Fuchs is eager to receive an SSO proposal. A big question is the availability of programmers to write the code for Web-based SSO and identity components of Shibboleth. [AI] Ken will send a note to Carl Jacobson to find out more about JA-SIG's Web SSO and identity requirements.

Ken gave a brief overview of plans for the Spring 2001 Internet Member Meeting. An Early Adopters session on making the business case for middleware will likely address some of the communication issues discussed in the last MACE call. MACE-Dir's recent and ongoing work will be showcased at the Member Meeting.

NSF has issued its middleware solicitation; see http://www.nsf.gov/cgi-bin/getpub?nsf0163 . Ken warned MACE to expect lots of discussion of the integration piece over the next 4-6 weeks. The funds for integration -- expected to be about $1.4m for the first year -- may go to one integrator or to several. Ken expects Microsoft and Boeing to make zero-cost bids for the integration piece.

Finally, MACE discussed recent activity around the emerging use of middleware to provide online-course and curricular tools for K-20. Ken noted that he had received mail from NSF identifying this as an extremely important problem space and asking what NSF could do accelerate progress in it. While there was general agreement with the NSF's sense of the importance of this area, there was disagreement on the extent to which wide-area deployment issues and entrenched attitudes among teachers (Steven: "I'm amazed at how many K-12 teachers think that curriculum is something that comes in a a box from Houghton Mifflin") will slow progress in this area.

As several MACE members will be unavailable in two weeks, the next MACE conference call will be on March 12.

*Action Items*

[AI] Michael will outline some ideas for conveying I2-MI's priorities and general approach to a wider audience.
[AI] Michael will reply to Jack Suess about his plans for a UMBC Middleware Day.
[AI] Ken will send a note to Carl Jacobson to find out more about JA-SIG's Web SSO and identity requirements.