On Mon, 30 Sep 2002, Wagner,Harry wrote:

> The only functional requirements I'm aware of are at:
> http://dublincore.org/groups/registry/fun_req_ph1-20011031.shtml, and did
> not get beyond the draft stage.

I would like to think that in DCMI terms getting to 'draft stage',
reflecting consensus of the list, and staying as a stable document for a
year is worth quite a lot ;-)

Eric Miller said:
>
> > Harry, Is there a page where the requirements along with
> > those discussed
> > on this list are being maintained so that we can better track the
> > direction of this work?
>

That would be good.... and I reckon it would be worth doing a snapshot of
requirements post-DCMI 2002 meetings. Not sure anyone has the will to
track a functional requirement document against the DCMI WG mailing list
as time goes by though? Volunteers welcome!

I think we all have to accept we are not in some formal software
development process here... with user requirements, functional
requirements etc. Such documents are merely one form of communication for
us.

Harry said:

> The focus
> for phase 2 is to enable the registry to interface with other registries and
> applications.  Part of that interface will be to provide information about
> DC terms and qualifiers in various encoding formats, including xsd and rdfs.

... not sure how you see the registry providing info about terms in rdfs
encoding? I would assume that would have to be by means of a rdf schema?

> It makes sense that these formats would be useful to applications.  It does
> not make sense (IMO) for the registry to serve rdf schemas to users via the
> UI.  The registries focus is to provide information about DC terms and
> qualifiers, not about encoding techniques.
>

I see rdfs as just another sort of information about terms.

>
> There are some additional issues here.  The only approved version of the
> terms and qualifiers is the enUS version.  We will either have to serve
> non-enUS versions with a caveat or get the non-enUS versions approved.  I
> can see problems with both approaches.

.... if we treat non-enUS versions as in effect 'annotations' then they
carry with them the trust associated with their creators as authenticated
by the DCMI Registry. As I see it this is all we can offer, there is no
single DCMI body that can 'approve' the many translations. Fine we can add
a caveat.

Rachel

---------------------------------------------------------------------------
Rachel Heery
UKOLN
University of Bath                              tel: +44 (0)1225 826724
Bath, BA2 7AY, UK                               fax: +44 (0)1225 826838
http://www.ukoln.ac.uk/