I agree with Jane's idea of a broader base type; just not in touch with how to implement it and then getting lost in some really messy territory.
If I understand her point, we need to have a schema that allows a community to say e.g., the value of DC:date needs to be in yyyy-mm-dd format and not just a string. More problematic, of course, is if a community wants to say the value of dc:creator should be a complex type with <firstName> <lastName> elements. This gets us into the whole dumb-down discussion, which Roland and crowd has addressed so eloquently in their solution to qualifed dc in rdf (relying on rdf primities such as rdf:value, etc.).
I'm not sure that we want to or can imprint this dumb-down ability in an xml schema.
Jane, I've seen some code fragments suggesting how to do this. Perhaps we can develop a complete example....
xml for the "base schema" defining the elements with a proper base type
packaging schema with types restricted by a community (e.g., OAI who only wants string values)
example instance document
Carl
> -----Original Message-----
> From: Pete Johnston [mailto:[log in to unmask]]
> Sent: Wednesday, March 06, 2002 7:16 AM
> To: [log in to unmask]
> Subject: Re: Public Comment on DC-simple XML Schema declaration within
> OAI
>
>
> Hello Jane,
>
> Thanks for your comments.
>
> > My major concern with this schema is that you are limiting
> > implementors who wish to refine the base DC element types -
> > to restrictions of strings only.
>
> This "simpledc" schema has, admittedly, been driven (to some extent at
> least) by the specific requirements of the OAI folks. The current OAI
> schema for DC metadata uses a type of string for all elements
>
> http://www.openarchives.org/OAI/1.1/dc.xsd
>
> so I guess the OAI folks are happy with a base of xs:string and don't
> have a requirement for a broader "base" type.
>
> But leaving OAI aside for a moment, I think the intention was to model
> (only) 'Simple DC'?
> I guess this brings us back to the perennial "what is Simple DC?"
> question, which I'm not sure DCMI has a clear answer to - and
> partcularly, is it true that in 'Simple DC', the value of a DC element
> is a literal. I guess I was basing my assumptions on the "abstract
> model" described in section 4.1 of
>
> http://www.ukoln.ac.uk/metadata/dcmi/dc-xml-guidelines/
>
> <quote>
> A simple DC record is made up of one or more properties and their
> associated values.
> Each property is an attribute of the resource being described.
> Each property must be one of the 15 DCMES [DCMES] elements.
> Properties may be repeated.
> Each value is a literal string.
> Each literal string value may have an associated language
> (e.g. en-GB).
> </quote>
>
> I do take your point that other applications making use of
> elements from
> the http://purl.org/elements/1.1/ namespace may well require more
> flexibility in terms of typing. And when we were working on
> the schemas
> to support the Guidelines for DC in XML document, we picked
> xs:string as
> a fairly arbitrary "base" type to get something working, but I would
> have been quite happy to make that broader along the lines
> you suggest,
> because the aim there was explicitly to try to support the
> functionality
> you describe.
>
> However, for this simpledc schema, the intention was (I
> think!) only to
> model 'Simple DC'. If the summary above is a good representation of
> 'Simple DC' (i.e. values are literal strings) and if this schema is
> intended only to model Simple DC, then I was going to argue
> for sticking
> with a base type of xs:string - but I've just seen Roland's
> message and
> I think maybe we need to look harder at that... ;-)
>
> If this is not the correct model for Simple DC and/or the schema is
> intended as a more general basis for application profiles,
> then I think
> a broader base type is certainly required, as you suggest.
>
> Cheers
>
> Pete
>
|