Re: Formation of RDF terms

Jonathan Borden wrote:
> 
> Graham Klyne wrote:
> > Jonathan,
> >
> > I pretty much agree with most of your response.
> >
> > Maybe where we differ is that I would like to see the model straightened
> > out, or clarified, THEN sort out the syntactical oddities with reference
> to
> > a stable, well-understood model.
> 
> I don't see why these need to be done in serial fashion. I'm also not sure
> what so pressingly needs to be straightened out in the model.

Brian McBride's issue page (see
http://www-uk.hpl.hp.com/people/bwm/rdf/issues.htm) has more issues
under the model section than the syntax section. Its tends to be
easier for people to poke holes in "ugly" XML syntax related issues
than the seemingly simple model.

<snip />

> I think it has been conclusively demonstrated that the syntax is
> insufficient for exchanging model data -- unless I'm missing something but
> unless there is a way to serialize every URI as a qname, and back again, how
> are model's to communicate? (since rdf:Statement serializes only a reified
> statement, not the statement itself)

I could flip this and say that the model is insufficient for
exchanging syntax data ;-). All it would take is a single additional
triple for each predicate/typednode that indicated the namespacePart
of the URI. 
 
> But let's assume that a part of the RDF community is solely interested in
> the RDF Model and the purpose of the syntax is solely to exchange data
> between models ... this stands the rest of the community which is concerned
> with either embedding RDF into XML documents, or interpreting XML documents
> as RDF.

I'm not sure how to read this paragraph. Did you mean to say "... this
strands the rest of the community"? 

> Let me throw this out: Compare and contrast the RDF Syntax vs. the XLink
> syntax as a means to exchange models between applications.
> 
> hint: for this specific purpose XLink, which deals solely the URIs not
> qname<->URI, is a perfectly good syntax, XLink similarly allows embedding of
> RDF triples within XML documents. What XLink does not help with is the RDF
> interpretation of 'existing' XML documents.

The ability to round-trip RDF/XML documents is a problem. URI<->qname
is one part of the problem. The issues list identifies many others.
The ability to round trip a model through RDF/XML is a problem. My
guess is that this can be dealt with via some mechanical URI<->qname
algorithm. Either way, I wouldn't want to use XLink to solve the
problem but to address it directly with fixes to both the model and
the syntax.

Gabe


> 
> -Jonathan

-- 
--------------------------- 
http://www.jfinity.com/gabe

Received on Wednesday, 31 January 2001 21:56:21 UTC