W3C home > Mailing lists > Public > www-webont-wg@w3.org > October 2002

Re: why 4.3 Structured Datatypes should wait

From: Jonathan Borden <jonathan@openhealth.org>
Date: Wed, 23 Oct 2002 22:43:51 -0400
Message-ID: <005b01c27b07$303c9760$7c674544@ne.mediaone.net>
To: "Dan Connolly" <connolly@w3.org>, <www-webont-wg@w3.org>

In short, this argument might just as well hold for RDF datatypes which also
seem to use URIs for XML Schema simple datatypes.

(this is a real problem though)

why not extend RDF datatypes to enable complex/structured datatypes ala XML
literals e.g.

ex:foo ex:prop ex:structured"<this>is a <simple/> structured
datatype</this>" .

where ex:structured is defined to be something like:

<!ELEMENT this (PCDATA|simple)*>

I could live with typed XML literals for the time being.

I do agree that the current situation with URIs for XML Schema particles is
problematic.

Jonathan


>
> In short: because the XML Schema WG hasn't
> yet decided how XML schema components fit
> into URI space.
>
> cf.
>
> * TAG interested in progress on URIs for schema components (NUNs)
> Dan Connolly (Thu, Oct 17 2002)
>
http://lists.w3.org/Archives/Public/www-xml-schema-comments/2002OctDec/0011.
html
>
> To elaborate a bit, yes, it would be nice to say...
>
>  @prefix : <http://example/vocab#>.
>  @prefix xs: <http://www.w3.org/???/xmlschema-components#>.
>
>  :description rdfs:domain :Product;
>    rdfs:range :ProductDescription.
>
>  :ProductDescription a xs:ElementDeclaration;
>    xs:name "desc";
>    xs:typeDefinition [ a xs:ElementOnly;
>      xs:contentModel [ xs:sequence (:name :photoref :blurb)].
>
> so that
>   <desc>
>     <name>SuperCam</name>
>     <photo ref="supercam1.jpg"/>
>     <blurb>best camera since sliced bread!</blurb>
>   </desc>
>
> was an element of the class :ProductDescription.
>
> In fact, there's a nifty UML diagram in the XML Schema spec,
>   http://www.w3.org/TR/xmlschema-1/#component-diagram
> and if our UML/OWL stuff works out, it should be a slam-dunk
> to model stuff like schema components and properties,
> and to say that XML Schema documents have the corresponding
> RDF graph semantics.
>
> But the bad news is: the current XML Schema specs
> don't determine URIs for schema components such
> as :ProductDescription. They get as far as
> a (namespace name, localname, whichsortathing) tuple,
> but they don't tell you the name for the thing.
>
> I suppose we could write an expression for
> "the element declaration schema component with
> localname 'desc' in the namespace 'http://example/vocab' "
> ala
>
>   [ a xs:ElementDeclaration;
>      xs:name "desc";
>      xsmap:namespace <http://example/vocab>
>   ]
>
> but... is that what you really want to do?
>
> So I'd like to keep 4.3 postponed; perhaps
> our requirements document should show this
> as a goal we didn't meet... or more straightforwardly:
> let the issues list cite this message, and
> let an update of the requirement document cite all the
> postponed issues, as we go to last call.
>
> --
> Dan Connolly, W3C http://www.w3.org/People/Connolly/
>
>
>
>
Received on Wednesday, 23 October 2002 23:02:50 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:57:53 GMT