Forwarded from Dan Brickley...
Tom
---------- Forwarded message ----------
Date: Wed, 1 Mar 2000 19:17:22 +0000 (GMT)
From: Dan Brickley <[log in to unmask]>
To: Thomas Baker <[log in to unmask]>
Excellent news, very exciting. It's as if the Warwick Framework never
went away :-)
Dan
On Wed, 1 Mar 2000, Thomas Baker wrote:
> Dear all,
>
> In the Usage Committee call yesterday, there was considerable support
> for the following way forward:
>
> -- We can vote on the 'semantics' of the Agent WG proposal -- even
> Affiliation, Agent Type, etc -- as long as we stop calling them
> qualifiers of CCP elements of the DCMES and recast some of them
> (Affiliation, Type, etc) as a ballot for top-level elements of a
> minimal, first-step Agent Core (or Authority Core).
>
> Some qualifiers -- such as Role, which refines the CCP elements
> themselves (an Illustrator ISA Creator) -- would remain in the
> ballot of DCMES. However, elements that are descriptive of Agents
> (an Agent HASA Affiliation) would be declared out of scope of the
> existing DCMES and moved to the Agent Core.
>
> As a community, we already recognize the notions of "core element
> set", "element", "semantic refinement", and "value encoding", so no
> new principles would be necessary for doing this.
>
> Eric is currently separating the Agent qualifiers along these lines
> in two a revised DCMES qualifier ballot and an Agent Core ballot.
>
> -- We would then recognize that the practical problem for implementors
> is to formulate good-practice solutions and workarounds for those
> who want to use elements from both the DCMES and DC Agent Core in
> records with a flat structure.
>
> This could take the form of a document that explains the problem;
> provides examples; offers standard encoding schemes or structured
> values using Agent Core semantics (or vCard, for example);
> illustrates and comments on alternatives in HTML, XML, RDF, and
> other database formats if necessary; and evaluates any syntactic
> solutions in the light of implications for dumb-down and long-term
> migration to more structured records. That document could become a
> DCMI Recommendation.
>
> By doing this, we would separate the semantics of Resources from the
> semantics of Agents, and separate both of these from any
> community-specific structured values used as values for CCP elements.
> We would also provide richly contextualized explanations of the
> practical tradeoffs involved in combining elements from the two
> separate cores in flat record structures.
>
> Such a solution helps separate "semantic" issues from "syntax" -- at
> least, more cleanly than in our current approach. It would also more
> cleanly separate the dictionary for the "pidgin speakers" of DC from
> the richer (but more syntax- and application-bound) constructions for
> "native speakers".
>
> Christel Hengel of Die Deutsche Bibliothek (copied to this message)
> tells me that she strongly supports the creation of an Agent Core (or
> Authority Core). Indeed, she had announced at DC-7 that she would
> convene a DCMI working group for an Authority Core and plans to move
> forward on this next week so that it can be formally announced at a
> German library conference on 20 March. I understand the group will
> involve (among others) a broad range of German libraries and OCLC, and
> it would start with authority data on persons -- an easier problem than
> for corporate entities.
>
> As far as I can tell, there is considerable enthusiasm on all sides
> about this solution to our current situation. Indeed, the creation of
> a minimal, first-step Agent (or Authority) Core and the creation of a
> DCMI Authority WG mutually reinforce each other and would give this
> solution some much-needed momentum.
>
> Tom
>
> _______________________________________________________________________________
> Dr. Thomas Baker [log in to unmask]
> GMD Library
> Schloss Birlinghoven +49-2241-14-2352
> 53754 Sankt Augustin, Germany fax +49-2241-14-2619
>
> _______________________________________________________________________________
> Dr. Thomas Baker [log in to unmask]
> GMD Library
> Schloss Birlinghoven +49-2241-14-2352
> 53754 Sankt Augustin, Germany fax +49-2241-14-2619
>
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|