Agents Working Group Report -- DC 2002, Florence

	Chairs: John Kunze, José Borbinha


At the DC 2002 conference in Florence, October 2002, the
Agents working group met for 2 hours on each of 2 days.  The
first day began with a review of the action items from DC
2001 (Tokyo) and the group status report.  Discussion moved
on to the milestone of developing a list of CCP qualifiers;
the expectation was that this achieving this would be a
two-part task resulting in a proposal for a set of MARC
Relators (roles such as publisher, illustrator, editor, etc.)
and a proposal for structured values for CCP elements.

This led to discussion of two important DC Usage Board
decisions relevant for DC Agents.  First, approval of the
idea of letting the US Library of Congress "bless" and
maintain the MARC Relator terms (about 185 terms) as Dublin
Core Agent/CCP element qualifiers.  If this occurs, the DCMI
is likely to view this as partial completion of our
milestone.  Consensus among the assembled group was to
endorse this UB decision.

The second UB decision was to declare that dc:creator would
be a refinement of dc:contributor (dc:creator will still be
usable as a top-level element).  Consensus was to request the
UB members present (Andy Powell, Diane Hillman) to clarify
whether loc:creator or dc:creator refines dc:contributor.

Then Warwick Cathro gave some background on the Agent element
set notion, and this led to discussion of Robina Clayphan's
proposal regarding structured values (AgentDetail) for CCP
elements.  Rough consensus emerged that there was a need in
some cases to be able to describe agents (eg, affiliation,
email address) completely external to resource descriptions,
and in other cases to include some agent details in resource
descriptions.  A group decision was taken conduct an
investigation into requirements for an Agent Core (Warwick
volunteered).

The second day began with a very interesting presentation by
Robina Clayphan on the INTERPARTY project, which touches on
many topics pertinent to DC Agents (authority, existence,
identification, rights management, etc).  Then John Kunze
gave a summary of the work going on in the DELOS/NSF Actors
working group (which had just met in Florence).

Next, the group went back to work on the issues related to
appropriate places to include agent information.  Decisions
were taken to seek further articulation of the two main
opposing views (inclusion vs. exclusion of agent details from
resource descriptions).

In the remainder of the time, Stuart Weibel gave a very brief
recap of his current thinking on an Agent linking convention
proposal. The group discussed this and decided that the
proposal was essentially sound but that a simplified
convention should be articulated.


		DC Agents WG -- Decision Summary

Decision 1
----------
who:   Warwick Cathro, [log in to unmask]
what:  Will write up a feasibility investigation into the requirements
       for an Agent Core.
when:  progress report by December 20, final report by March 2003

Decision 2
----------
who:   Andy Powell | Diane Hillman
what:  Will articulate pros (Andy) and cons (Diane) of the concept of
       including URIs and email addresses in Agent/CCP values.
when:  in an email to dc-agents by December 1

Decision 3
----------
who:   John Kunze
what:  Will try to synthesize Andy and Diane's, revisiting Robina
       Clayphan's "structured value" AgentDetail proposal.
when:  by December 15

Decision 4
----------
who:   Stuart Weibel
what:  Will propose a 3-component (3 maximum) Agent linking convention
       (eg, appropriate literal, link, link type).
when:  by November 1