On Thu, 28 Jun 2001, Thomas Baker wrote:
> We see three sets of issues:
Good summary of the issues.
I appreciate that this is not the list to discuss this, but fwiw, here are
my personal opinions on these three issues...
> 1) Element refinements of Creator/Contributor/Publisher -- things like
> "Illustrator". We have been discussing a way to give approval to a
> sub-set of the MARC relator terms (see Rebecca Guenther's posting of
> June 22) -- an idea that was originally proposed in (I think) 1999.
> Other working groups could put forward other such refinements, but I
> think this Usage Board would prefer that we point to other
> namespaces for really specialized terms and limit the refinements in
> the DCMI namespace to a minimum number of particularly useful or
> salient terms.
I'd like to see MARC relator terms specified as refinements of
dc:contributor only. If the MARC terms do not include creator and
publisher, then I would like to see creator and publisher proposed as
refinements of contributor, i.e. we have dcq:creator and dcq:publisher
defined as being equivalent to dc:creator and dc:publisher but also as
refinements of dc:contributor. In effect we deprecate the use of
dc:creator and dc:publisher, recommending use of dcq:creator and
dcq:publisher instead.
This puts all the 'agent' elements other than dc:contributor at the same
level - and everything dumbs down to dc:contributor. No existing stuff
breaks and no new terms need adding to the dc namespace. ...and it moves
us away from the problem where one or more of the terms in the MARC list
can dumb-down to several terms in the dc namespace.
> 2) Types of agents, such as Person or Organization.
> We would welcome a proposal.
We more or less have a list don't we? I'd like to see this as a second
DCMI AgentType vocabulary rather than as additional terms in DCMI Type.
> 3) Properties of persons and organizations, including contact
> information. I don't believe anybody on the UB wants us to reinvent
> vCard as a DCMI vocabulary. If Person or Organization were
> recognized as DCMI types, however (signaling a broadening of our
> implicit scope), then this would make other properties of agents
> into candidates for DCMI terms.
One or more application profiles that use terms from dc, dcq and vCard
namespaces is what we want here isn't it?
But I guess I need to resend this message to the DC Agents WG?! :-)
Andy
--
Distributed Systems and Services
UKOLN, University of Bath, Bath, BA2 7AY, UK [log in to unmask]
http://www.ukoln.ac.uk/ukoln/staff/a.powell Voice: +44 1225 323933
Resource Discovery Network http://www.rdn.ac.uk/ Fax: +44 1225 826838
|