*MACE Conference Call*
October 7, 2002
Bob Morgan (chair) - Washington
Ken Klingenstein - Colorado/Internet2
David Wasley - UCOP
Ton Verschuren - SURFnet
Neal McBurnett - Internet2
Brian Gilmore - Edinburgh
Steven Carmody - Brown
Renee Frost - Michigan/Internet2
Scott Cantor - OSU
Nate Klingenstein - Internet2
Steve Olshansky - Internet2
Michael Gettes - Georgetown
Keith Hazelton - Wisconsin
Ben Chinowsky (scribe) - Internet2
The call opened with discussion of upcoming meetings.
- The Fall 2002 Internet2 Member Meeting is October 27-30. Renee noted that the middleware sessions (see the program at www.internet2.edu/activities/php/agenda-Fall02.php?session_event_id=122) are still "slightly subject to change", so contact her if you have concerns.
- Ton noted that he will be attending a meeting of the TERENA Task Force on Authentication, Authorisation Coordination for Europe (TF-AACE), November 26-27 in Stockholm. He would like someone from MACE to help him explain Shibboleth and related components -- especially Pubcookie -- with a view toward helping European developers understand what they need to do to ensure interoperability with Shib. Contact Ton if you're interested. Registration information is at www.terena.nl/tech/task-forces/tf-aace/workshop/registration.html; attendance via videoconference may also be a possibility.
- Bob suggested that MACE might want to be officially represented in the November 12-13 AAMC workshop, "Convergent Validity: A Forum For Academic Medical Centers to Discuss HIPAA Implementation"; for details see www.aamc.org/meetings/specmtgs/hipaagir/start.htm.
Much of the call was devoted to a discussion of issues raised by Nate's draft standards for Internet2 document process, format, and naming. Points made and questions raised included:
- The difference between an Internet2 working group and an Internet2 initiative needs to be clarified.
- The document needs to provide more detail on the duties of the document review panel and the external reviewers.
- Filenames should: 1) not change radically between draft and final versions, 2) use dashes as component separators and underscores to represent spaces within components, 3) specify dates to the day, rather than just to the month.
- There was much discussion of dates vs. version numbers. Dates provide more information than version numbers, but version numbers are more traditional. Bob suggested that the use of both is inevitable, so appropriate contexts should be defined for each; for example, "1.0" has a more definite meaning for a software release than for a document.
- Unambiguous names are good; alias names would probably add more confusion than convenience.
- These documents will have lots of metadata that won't fit in the filename, so where should it go? Somewhere in the document, somewhere in the Internet2 Digital Library database? David suggested that MACE seek input on this question from a librarian. Bob cautioned that these documents will get sent around in many formats, especially email, that won't include any metadata that's stored separately from the document.
- Lots of things seem to remain in draft format forever, so it would be good to have a timeline for docs to move to final. On the other hand, Bob pointed out that NMI releases will include things that are still in the draft stage, in order to get external review. David suggested that timelines be presented as guidelines rather than absolute requirements.
- SteveO suggested that MACE seek external review of the document-standards document; as Internet2 and NMI have specific and well-defined constituencies, this process need not become unwieldy.
- Internet2 no longer needs to correlate its naming convention with the Grid's, so the naming scheme used for NMI-R1 is being abandoned.
- Long lists of authors are bad. Bob suggested that documents list the editor as the author, and include extensive acknowledgements specifying who did what.
[AI] All who have suggestions on issues raised by the draft document-standards document will send them to Nate and cc the MACE list. SteveO noted that the naming convention in particular needs to be decided on by October 21 -- the date of the next scheduled MACE call -- in preparation for the October 25 release of NMI-R2.
Finally there were short updates on NMI, FOO, and DRM:
- Ken noted that NMI is now prepared to accept a wider range of contribution types, including a wider range of software and object classes. This will provide a way to incorporate some of the software being produced by the Grid, as well as items that haven't come through NMI-affiliated working groups. NMI-R3 will probably be the first to include this broader range of items.
- The first FOO call went well. Bob noted that the call included some individuals approaching this problem from a "large commercial organization perspective", as well as the usual suspects.
- Charters for the DRM technical and policy groups are still in the works; the technical group is expected to launch before the policy group.
[AI] All who have suggestions on issues raised by the draft document-standards document will send them to Nate and cc the MACE list.