On Tue, 13 May 2003, Douglas Campbell wrote:
> Andy,
>
> >>> [log in to unmask] 13/05/03 01:09:10 >>>
> I've put up a new version of the working draft at
> http://www.ukoln.ac.uk/metadata/dcmi/dcq-html/
> which I hope takes account of all the comments and discussions.
>
> Sorry, coming in late to this one - looks like the important aspects
> are covered, looks good. Just a couple of minor points...
>
> 1. - "In the context of this document, it should be noted that the
> resource being described is an HTML/XHTML Web page and the Dublin Core
> record describing that resource is embedded into it using HTML/XHTML
> <meta> elements."
>
> Perhaps add something like:
> "It is only possible to describe one resource using the method in this
> document, to describe multiple resources from a single Web page requires
> linking to a separate XML or RDF/XML description containing the multiple
> Dublin Core records, as described above."
Yes, OK - thanks.
> I don't know if describing multiple resources is acceptable or not, but
> I always wondered how you would individually describe a number of PDF
> files (or other non-HTML files) linked off a single page so they'd each
> get their own record.
I don't think that it is possible to do this using HTML-embedded
DC metadata.
> 2.2 and/or 3 - Sometimes I've seen extra layers of refinement used, eg.
> "DC.Creator.Name.FirstName" (or even sometimes it's just to put in an
> encoding scheme, eg. DC.Subject.LCSH or DC.Coverage.Temporal.W3CDTF).
> Is it worth mentioning the unpreferredness of this?
My view is that it is sufficient to specify what is allowed - on the
assumption that everything else isn't allowed. But I'm happy to add
something along the lines of the above if others agree this would be
useful.
What are other's views?
> 2.4 - Is it OK to be a little more specific of what it replaces [see
> <new>]:
> "Where the value of a property is the URI of another resource (as is
> typically the case with the dc:relation element for example) an
> alternative form of encoding using the XHTML <link> element is
> preferred<new> (rather than using the dcterms:URI encoding
> scheme)</new>."
Ok, yes - thanks.
> 4. - This section is clear for the changes required for HTML 4.01 - but
> was the <link> element around during earlier HTML versions, in
> particular I'm thinking of the href aspect in section 2.4 if you're
> using HTML 2 or 3?
>
> 5. - I assume it was intentional that you used a "." for AGLS ie.
> name="AGLS.Function"?
Yes - following their current conventions.
> 6. - AGLS reference - I think they now longer use the long name as they
> feel it is no longer just a Government thing, their website calls it the
> "AGLS Metadata Standard" - best to check with them though.
OK, thanks.
> Thanx,
> Douglas Campbell
> National Library of New Zealand
>
Andy
--
Distributed Systems, UKOLN, University of Bath, Bath, BA2 7AY, UK
http://www.ukoln.ac.uk/ukoln/staff/a.powell +44 1225 383933
Resource Discovery Network http://www.rdn.ac.uk/
|