> From [log in to unmask] Fri Mar 8 07:09 MET 2002
> MIME-Version: 1.0
> Date: Fri, 8 Mar 2002 14:00:09 +0800
> From: Simon Cox <[log in to unmask]>
> Subject: Re: DC-ARCHITECTURE Digest - 6 Mar 2002 to 7 Mar 2002 (#2002-31)
> To: [log in to unmask]
>
> The business of mixing "simple content" (e.g. text) with elements is quite
> fraught in in the derivation mechisms within W3C XML Schema.
> As Pete points out below, mixed="true" does not constrain the position or
> number of text nodes.
> There is also some real confusion - even amongst the authors of the spec -
> as to whther or how you can start with a "mixed" supertype, and retstrict
> that down to a specific simpleContent derived type.
> i.e. - if all the basic elements are specified to be type="text"
> - i.e. mixed -
> then it is probablly not possible to derive a type by restriction whose
> simpleContent is a "date".
Huch! I thought it's just my meager knowledge of English to understand, whether
that is possible or not by the w3c-xml-schema-spec.
rs
>
> _____
> [log in to unmask] CSIRO Exploration & Mining
> 26 Dick Perry Avenue, Kensington WA 6151
> PO Box 1130, Bentley WA 6102 AUSTRALIA
> T: +61 (8) 6436 8639 F: +61 (8) 6436 8555 C: +61 (4) 0330 2672
> http://www.csiro.au/page.asp?type=resume&id=CoxSimon
>
> >
> > Date: Thu, 7 Mar 2002 13:29:42 -0000
> > From: Pete Johnston <[log in to unmask]>
> > Subject: Re: Public Comment on DC-simple XML Schema
> > declaration within OAI
> >
> > Carl,
> >
> > > 1. Do not spec an exact list of sub-elements. Allow that to
> > > be open territory. 2. Mandate in the schema that each element
> > > have a dumb-down value. Sigge's earlier mail stated this with
> > > the tag <simpleValue> but I don't think that this syntactic
> > > sugar is actually necessary.
> >
> > I think some sort of explicit marking of either the "complex value" or
> > the "simple value" may be required because the text content model (I
> > think) allows "mixed content" so there could be any number of
> > text nodes
> > interspersed amongst child elements within a DC element.
> >
> > I'm torn between your option 2 and option 3.
> >
> > I'm just a little bit worried that if we go the option 2 route, we're
> > attempting to reinvent (parts of) RDF a bit hastily.
> >
> > If we can satisfactorily address the definition of a second
> > schema which
> > uses import or redefine to provide the "tightening" of a loose base
> > content model (which we have now) for e.g. OAI's
> > "simple-DC-values-are-strings" model, I think that feels a better
> > solution to me.
> >
> > Pete
>
|