*MACE Conference Call*
June 16, 2003

*Attendees*

Bob Morgan (chair) - Washington
Diego Lopez - RedIRIS
Mark Poepping - CMU
Brian Gilmore - Edinburgh
Jim Jokl - Virginia
Steve Worona - EDUCAUSE
Michael Gettes - Duke
Steven Carmody - Brown
David Wasley - UCOP
Renee Frost - Michigan/Internet2
Keith Hazelton - Wisconsin
Scott Cantor - OSU
Neal McBurnett - Internet2
Steve Olshansky - Internet2
Ken Klingenstein - Colorado/Internet2
Ton Verschuren - SURFnet
Ben Chinowsky (scribe) - Internet2

*Discussion*

Ton noted that TNC 2003 had turned out to be "the best-rated TERENA conference ever." Excellent streaming and wireless facilities, and a good mix of networking people and users, were noted as key factors in this success. Presentations and other documents from TNC 2003 are available at http://www.terena.nl/conferences/tnc2003/programme/final-programme.html. [AI] Diego will work with Licia to plan a late October or early November TF-AACE meeting in Spain. Ken asked that the location be accessible via a direct flight from the US, so as to make it easier to get the desired broad representation from both academia and industry in the US.

The recent CAMP meeting was also a success. In accordance with the shift of focus represented by the name change from Base CAMP to just CAMP, the overall level of expertise was higher than in the past. Several new volunteers were recruited for various projects. Ken suggested that MACE revisit the idea of bringing a more topical focus to future CAMP meetings, either by having entire CAMPs devoted to particular topics, or by opening with an overview session, then offering multiple topic sessions (in serial or parallel), among which attendees could pick and choose. [AI] Renee will organize a conference call on how to structure future CAMPs. Bob suggested it would be helpful if CAMP presenters could make their documents available online before the meeting; [AI] Renee will look into getting CAMP presentations online ahead of time.

Ken reported that disappointment with progress with grids is fueling an effort by more traditional Computer Science people to convince NSF that "neither grids nor middleware infrastructure is where it's at -- we need more money for grad students." Better advocacy and the anticipated arrival of funds for the NSF cyberinfrastructure initiative (see http://www.cise.nsf.gov/news/cybr/cybr.htm) will hopefully improve the situation for NMI. Ken noted that the grid community has started to show a lot of interest in Shibboleth and federations; Ken and Bob will be at GGF8 next week (http://www.ggf.org/Meetings/ggf8/ggf8_reg.htm).

An Internet2 Middleware P2P working group is in formation; its work will center on Penn State's LionShare (http://p2p.libraries.psu.edu/), supported by Jabber (http://www.jabber.org/) for out-of-band communication. Funding proposals are being developed. There seems to be broad and growing interest in inter-enterprise authN for new contexts such as Jabber and videoconferencing, so there are many possible areas of work for the P2P WG, and possibly a need for other related WGs as well. Bob encouraged all to send the MACE list suggestions for the Middleware P2P working group charter. [AI] Bob will look into the status of work at UW on authentication in new contexts such as Jabber and videoconferencing.

There was a short discussion of an issue from VidMid: how to reconcile H.350 storing certs in the directory with Bill Doster's position that the cert should be stored in the OS, where the OS expects to find it? Bob observed that this is a bit like asking "should I keep my money in my wallet or in the bank?"; Jim suggested that the OS hold the user's certs and the directory hold others' certs.

A procedure for the urn:mace registry is emerging. Requests sent to mace-submit go to Keith, Bob, and Ken; Keith is responding to the requests and Lisa Hogeboom is maintaining the list of registered namespaces. MACE agreed to require that the very first thing after urn:mace be a naming authority; e.g., urn:mace:switch.ch is OK, but urn:mace:ch is not. [AI] Keith will set up a web page for the mace:urn registry.

Finally, Steven C. announced that the code for Shibboleth 1.0 is now finished; see http://shibboleth.internet2.edu/release/shib-v1.0.html. The Shibboleth team is hoping to launch InCommon later this summer. InCommon membership will be restricted to institutions, and membership will be thoroughly vetted. The pilot federation for InCommon is now called InQueue; the Shibboleth pilot sites will tranfer into InQueue. InQueue membership will also be restricted to institutions, but with lighter vetting. A "Club Roadkill" federation will have unrestricted membership. Nationally-based federations are also in the works, and Ken has suggested setting up federations based on research areas. The principal challenge over the next 6-8 weeks is to reach consensus on how to move the federations into production. Nate is working on a document describing the InQueue federation; [AI] Steven C. will send MACE the InQueue document once it's more complete. Ken noted that discussions continue with PingID (http://www.pingid.com), which aims to provide technology-independent support for federations. This is a hard problem. Right now PingID is looking at Shibboleth, Liberty, and federated Passport. PingID has joined the conference calls for FOO, which now has a web page: http://middleware.internet2.edu/foo/.

*Action Items*

[AI] Diego will work with Licia to plan a late October or early November TF-AACE meeting in Spain.
[AI] Renee will organize a conference call on how to structure future CAMPs.
[AI] Renee will look into getting CAMP presentations online ahead of time.
[AI] Bob will look into the status of work at UW on authentication in new contexts such as Jabber and videoconferencing.
[AI] Keith will set up a web page for the mace:urn registry.
[AI] Steven C. will send MACE the InQueue document once it's more complete.