[rdfweb-dev] RDF all the way on XML.com

Libby Miller Libby.Miller at bristol.ac.uk
Fri Aug 22 10:37:53 UTC 2003


hm...

<R r="http://monkeyfist.com/" id="r1">
  <PV p="http://foo.com/#Publisher">Monkeyfist Collective</PV>
</R>

actually I find that rather hard to read :)

I can certainly imagine that if that had been chosen as an xml syntax
there would still have been complaints; although it is clearer about the
underlying model which might have speeded up understanding of RDF, it
doesn't 'look like XML' in that it doesn't represent the relationshops
between things in an xml-y way:

<Person>
	<name>Libby</name>
	<mbox rdf:resource="mailto:libby.miller at bristol.ac.uk"/>
</Person>

<R id="r1">
  <PV p="http://www.w3.org/1999/02/22-rdf-syntax-ns#type" v="http://www.w3.org/2000/01/rdf-schema#Resource" />
  <PV p="http://xmlns.com/foaf/0.1/name">Libby</PV>
  <PV p="http://xmlns.com/foaf/0.1/mbox" v="mailto:libby.miller at bristol.ac.uk" />
</R>

See what I mean? think XML people would like the latter any more than
the former?

I think the underlying issue is that it's hard to encode a graph in a
tree syntax, since trees are a subset of graphs and not the other way
around.

At this stage however, it's not possible to change the syntax, unless
the RDF working group is reformed with a different remit. In its current
incarnation it was created to improve what was already there, and
explicitly not to create a new syntax. This was because people and tools
are now (and were then) using RDF, and to pull the rug out from under
their feet was thought to be the wrong thing to do.

http://www.w3.org/2001/sw/RDFCoreWGCharter

There is one real problem I can see with the syntax, which is that you
can't process arbitrary RDF with some generic XML tools, such as XSLT.

There are several efforts in progress to get various XML tools talking
to RDF so that this issue is minimised. My favourite is to coopt the
XPath syntax so that it can be interpretted to navigate the structure of
RDF documents, and then be used in XSLT and XQuery. There have also been
some discussions about processing RDF queries using XQuery.

Here is some work in progress along these lines:

http://sourceforge.net/projects/rdftwig/
http://rdfweb.org/people/damian/treehugger/

Dave's right in that www-rdf-interest at w3.org

(http://lists.w3.org/Archives/Public/www-rdf-interest/, although W3C is
having trouble with it's mailing lists at the moment due to the SoBig.F)

would be a better forum for these discussions; it's important to have
the history of these issues explained, but this is more of an
application-orientated list, so it's a bit distracting to get sucked
into discussions about the underlying technology (especially when these
discussions have happened many times before elsewhere).

cheers

Libby

On Fri, 22 Aug 2003, Victor Lindesay wrote:

>
> > -----Original Message-----
> > From: Dave Beckett [mailto:dave.beckett at bristol.ac.uk]
>
> > And syntax is just an endless discussion that I'm not going
> > into myself.
>
> I would have thought that one of the principle authors of RDF / XML
> syntax (and great respect to you for that Dave) could give some reasons
> why RDF / XML is complicated and why a simpler alternative has not been
> pursued.
>
> What ever happened to that RPV idea that Tim Bray proposed?
> http://www.xml.com/pub/a/2002/11/20/rpv.html
>
>
> _______________________________________________
> rdfweb-dev mailing list
> rdfweb-dev at vapours.rdfweb.org
> wiki: http://rdfweb.org/topic/FoafProject
> http://rdfweb.org/mailman/listinfo/rdfweb-dev
>
>




More information about the foaf-dev mailing list