*MACE Conference Call*
October 30, 2006
Bob Morgan (chair) - Washington
Ton Verschuren - SURFnet
Neal McBurnett - Internet2
Leif Johansson - Stockholm/SUNET
Steve Carmody - Brown
Renee Frost - Michigan/Internet2
Scott Cantor - OSU
Keith Hazelton - Wisconsin
Tom Barton - Chicago
Steve Olshansky - Internet2
Nate Klingenstein - Internet2
David Wasley - independent
Rodney McDuff - Queensland
Ken Klingenstein - Colorado/Internet2
Ben Chinowsky (scribe) - Internet2
Bob noted that the Malaga TF-EMC2, CAMP and JRA5 meetings revealed "lots of progress in federation deployments of all kinds." Steve Carmody observed that publishers seem increasingly aware of the value of federations; Bob noted that ScienceDirect now offers several choices of federation. Overall, the level of activity is encouraging. It was also noted that SCHAC is making good progress, driven primarily by the need to make university credits portable among different schools.
Not all the federation proposals put forward in Malaga use the end-to-end approach Bob prefers; he pointed out that eduGAIN in particular is gatewayed. Diego objected to this characterization, noting that while eduGAIN offers gateways (federation-wide Bridging Elements) for those federations not willing to deploy an individual BE at each participating site, this is an optional behavior, not an architectural feature. Bob replied that the issue is that BEs, and not just SAML endpoints, participate in the sign-on flow -- not whether a given BE represents one site or many. Diego pointed out that a BE that speaks only for a single realm is in fact a plain SAML endpoint, and that such endpoint deployments are preferred by the eduGAIN team.
- The December 4 Grid-Campus Integration session is coming together; Open Science Grid, TeraGrid, and caBIG will all be represented. [AI] Ken will recruit an EGEE presenter for the Grid-Campus Integration meeting.
- A CAMP based on the Authentication Roadmap has been scheduled for February 7-9 in Tempe, Arizona (right after the Net@EDU Annual Meeting).
- There will likely be both a CAMP and an Advanced CAMP in June; topics have not yet been determined.
The group discussed how to respond to the NSF Software Development for Cyberinfrastructure (SDCI) solicitation (http://www.nsf.gov/publications/pub_summ.jsp?ods_key=nsf07503). Ken noted that the two main new pieces of this solicitation are VOs and workflow. There is significant VOs work happening in the UK and Australia; this has led to renewed suggestions that a Middleware Investors for Collaboration and Education (MICE) group be launched to coordinate this work internationally. Ken is open to suggestions for how to move forward in the VO space. For workflow, Ken and Keith noted that scope is a big question -- should this address just Grid job scheduling, or should it be broader than that? [AI] Keith will contact Ken about scoping a response to the workflow piece of the SDCI solicitation.
Finally, Bob asked the group for comments on his October 30 email about the significance for MACE of the recent surge of interest in SOA. On the one hand, modularity is good; on the other hand, SOA raises questions about the future for LDAP. Keith observed that despite the abundance of hype, there is an important core to SOA, of people trying to make things more modular and reduce the number of interface points. There was general agreement with this evaluation; MACE expects to be discussing this further in the near future.
[AI] Ken will recruit an EGEE presenter for the Grid-Campus Integration meeting.
[AI] Keith will contact Ken about scoping a response to the workflow piece of the SDCI solicitation.