Hi to those of you on meta2 just joining this conversation... It
arises from an off-list discussion of a draft of my paper on Dublin
Core for the UK's electronic journal, Ariadne
(http://www.ukoln.ac.uk/ariadne/).
The URL that Andy gives below is to the draft of the paper as mounted
at Newcastle, but the formal 'published' version will be appearing on
Ariadne in the next day or two
(http://www.ukoln.ac.uk/ariadne/issue5/), and if you want to read it
you would be better off using the Ariadne version, where links etc
will make a little more sense (it also gives you a chance to browse
the rest of Ariadne while you are there!).
Jon Knight wrote:
> On Tue, 10 Sep 1996, Andy Powell wrote:
> > It seems to me that we now have two opposing views on the use of SCHEME
> > and TYPE for sub-elements, as described in
> >
> > http://www.ncl.ac.uk/~napm1/ads/dublin.html#scheme
this is the DRAFT. You'll be better off reading
http://www.ukoln.ac.uk/ariadne/issue5/, where the article will appear
within a day or two.
> > and
> >
> > http://www.roads.lut.ac.uk/Metadata/DC-SubElements.html
> >
> > It'd be nice to all be doing things the same way :-) ... and I'd really
> > like to get on and add some DC to the UKOLN pages?
>
> It would be nice to have a one-and-only-one standard for these. I based
> our DC sub-elements document on a combination of the original DC report, the
> SGML implementation of DC from Warwick, Paul's document, various
> postings from meta2 and comments from readers. In most of these, scheme
> was the basic sub-element and type only appeared in a few of these. If
> others on meta2 prefer Paul's idea of using scheme only to point to names
> of existing coding schemes and type otherwise then I'm happy to go
> along. If I don't hear any objections in the next few days I'll make the
> change to my document (glad I stuck draft all over it!).
OK, this is the meat of what we were discussing. In the recent paper
by Jon and Martin, many qualifiers to the DC elements are put in the
TYPE, with SCHEME used only for a few. MY paper suggests a different
approach to SCHEME and TYPE (one I'm sure I picked up from earlier
writings on Dublin Core) whereby SCHEME only refers to existing
schema, such as IMT (Internet Media Type), ISO31, etc, and TYPE
refers to the more general qualifiers such as email, address, fax,
etc.
This seems to create a more consistent approach, which will hopefully
be easier for the general user to understand.
Anyway, take a read of the two papers, and let us know what you
think...
Paul
Paul Miller
Graphics & GIS Advisor, University Computing Service
University of Newcastle, Claremont Tower, Claremont Road, Newcastle
upon Tyne NE1 7RU. tel (0191) 222 8212/8039, fax (0191) 222 8765
e-mail [log in to unmask] WWW http://www.ncl.ac.uk/~napm1/
[log in to unmask] http://www.ncl.ac.uk/~ngraphic/
|