W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > April 2001

Re: Referring to resources in RDF attributes

From: Graham Klyne <Graham.Klyne@MIMEsweeper.com>
Date: Wed, 18 Apr 2001 18:38:58 +0100
Message-Id: <>
To: Dan Connolly <connolly@w3.org>
Cc: RDF core WG <w3c-rdfcore-wg@w3.org>
At 01:05 PM 4/18/01 -0500, Dan Connolly wrote:
>I suggest, instead:
>     <rdf:RDF xmlns:rdf="http://www.w3.org/2001/04/18-rdf-with-qnames#"
>              xmlns:rdfs="http://www.w3.org/1999/PR-rdf-schema-19990303#"
>              xmlns:eg="http://www.example.org/Set">
>       <rdf:Description rdf:aboutQ="eg:Set">
>         <rdf:type rdf:resourceQ="rdfs:Class"/>
>         <rdfs:subclassOf rdf:resourceQ="rdfs:Container"/>
>       </rdf:Description>
>     </rdf:RDF>
>things I changed
>   1. the namespace name. This isn't RDF 1.0 syntax.
>   It's convertable to RDF 1.0 syntax (using XSLT;
>   we should provide a transformation if we persue this)
>   but it's not RDF 1.0 syntax.
>   We might conceivably replace RDF 1.0 syntax with this
>   syntax, pretending RDF 1.0 syntax never existed,
>   but (a) I doubt we will, and (b) until we do,
>   it's useful to keep them separate.
>   (cf my "namespace squatting: please don't" message
>         to www-rdf-interest a while ago)

... and it's not as if there's a shortage of namespace names, eh?

>   2. <rdf:type resource="..."> -> <rdf:type rdf:resource="...">
>   If you think that the XML Namespace spec specifies
>   that these are equivalent, read it again.
>      cf. http://www.w3.org/2000/03/rdf-tracking/#rdf-ns-prefix-confusion
>   I intend to elaborate on this separately...

I agree... that was careless of me.

>   3. rdf:about -> rdf:aboutQ
>         (for lack of a better name).
>   We can specify (using, e.g. XML Schema) that this
>   takes a qname value. More on using XML Schemas
>   to specify RDF syntax separately...

I considered something like this, but wanted to focus on the core idea.

Thought:  using a different namespace and XSLT transformation, maybe 
there's no need to be absolutely conformant to the current 1.0 spec:  in 
"-rdf-with-qnames#" one might say that the attribute value syntax was (say) 
like N3:

Just a thought.

Received on Wednesday, 18 April 2001 15:04:33 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:53:48 UTC