innovation in metadata design, implementation & best practices

Usage Board telecon report - Wednesday 28 January 2009 - 2100 UTC

This report:
Agenda was:

Present: Tom Baker, Julie Allinson, Pete Johnston,
         Akira Miyazawa, Stefanie Ruehle, Joe Tennis
Regrets: Andrew Wilson

Mailing list: 

Review of DCMI Libraries properties

We discussed Stefanie's draft note to Mary Woodley and Corey
Harper. Stefanie will send after incorporating comments
from Tom.

ACTION 2009-01-28: Tom to send comments to dc-usage.
-- DONE - see [2]


Review of Accessibility properties

Andrew to report on progress with AGLS in the February call.

Review of Scholarly Works Application Profile (SWAP)

Discussed draft review [1]. Noted that Julie has been making
corrections at [2] and will convert the AP into a HTML document
we can archive as a snapshot.

Accepted Julie's addition about the kev:ctx format as a
datatype, under "Metadata Terms Referenced".

Discussed the issue raised by Pete of two STs
(Expression/Entity Type and Expression/Type) with the same
property list constraint (a list of one member, dc:type) [3].

Noted that the DSP spec [4] on which the review criteria is
based is work in progress. Nonetheless, the consensus on
the call was that the discrepancy between the current DSP
specification and the review criteria is awkward. It would
be problematic to say the profile "conforms" if we need to
qualify that judgement by explaining a point on which it does
not conform. Rather, we should be in a position to say that
it either conforms with the draft DSP spec or does not.

There was agreement that we should publish the review with an
explanation of the problem uncovered and send this as input
to the SWAP and Architecture communities. Several possible
courses of action seem possible (see thread at [3]):

-- the DSP spec could be modified to remove the cause of the 
   discrepancy, as Mikael Nilsson has suggested [5]

-- the two STs in question could be collapsed into one single
   "weaker" ST [6]

-- the SWAP community could change the profile to not use
   the same property list constraint in the two STs (e.g., by
   using rdf:type for one of the two).

Pete noted a deeper issue in the DSP spec with regard to
possible overlap of sub-property list constraints, though
that is not an issue with SWAP.

ACTION 2009-01-28: Pete to draft a simple explanation of the 
issue for inclusion in the review, citing discussion on the list.


Profile Review Criteria

Discussed [1]. Agreed that the text should read:

    These tables may include the following information:
    * [OPTIONAL] Minimum and maximum numbers of times...
    * [OPTIONAL] The type of value surrogate...
    These tables must include one of the following property constraints:
    * [OPTIONAL] Property List Constraint...
    * [OPTIONAL] Sub-Property Constraint...

ACTION 2009-01-29: Tom to propose
corrections to the profile review criteria
for discussion on the next call -- probably as an erratum of the
current version instead of a completely new version.


Next telecon - sometime between 9 - 20 February

    See poll at

    -- Libraries terms (Stefanie)
    -- Accessibility properties (Andrew)
    -- SWAP
       -- Revised profile (Julie)
       -- Revised criteria (Tom)
       -- Reviewed review
          -- Description of DSP/SWAP discrepancy (Pete)