W3C home > Mailing lists > Public > www-ws-desc@w3.org > February 2006

RE: Review of WSDL 2.0 - RDF Mapping

From: Jacek Kopecky <jacek.kopecky@deri.org>
Date: Mon, 06 Feb 2006 22:36:58 +0100
To: "Booth, David (HP Software - Boston)" <dbooth@hp.com>
Cc: WS-Description WG <www-ws-desc@w3.org>
Message-Id: <1139261818.3786.22.camel@localhost>

David, thanks for the reply, and see below on your question.

On Mon, 2006-02-06 at 15:42 -0500, Booth, David (HP Software - Boston)
wrote:
> Yes, this is what I had in mind.  Thanks.  One further question though
> . . .
> 
> >       * In WSDL, an interface is always a different thing 
> > from a binding
> >         (even if they have the same name), and they will stay 
> > different
> >         when mapped from a WSDL file to RDF. In the ontology, however,
> >         one can assert that one resource is both an interface and a
> >         binding and this introduces no RDF-level inconsistency.
> 
> Doesn't Arthur's QName --> URI mapping algorithm prevent this?

When a valid WSDL file is mapped to RDF, Arthur's URI fragments make
sure this case doesn't happen, that's true. 

I see I was again confusing in the second half, though. It tries to say
that when somebody crafts an RDF file that follows the ontology, they
can easily use identifiers for the interfaces, bindings, services etc.
that do not follow Arthur's scheme, and indeed they can say about the
same thing that it's an interface and a binding at the same time and
there shouldn't even be much confusion, actually.

In effect, this is a very similar point to what's being said in the
ending of section 3.1.

I'll try to reformulate the text to be cleaner,

Jacek
Received on Monday, 6 February 2006 21:37:38 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:38 GMT