Re: Review of WSDL 2.0 - RDF Mapping

David,

thanks a lot for the review, I'll try to address all your comments soon.
So far only on the general comments, please see below. 8-)

On Thu, 2005-12-22 at 15:46 -0500, Booth, David (HP Software - Boston)
wrote:
> Document reviewed: "Web Services Description Language (WSDL) Version
> 2.0: RDF Mapping"
> http://www.w3.org/TR/2005/WD-wsdl20-rdf-20051104/
> 
> GENERAL COMMENTS
> The document thus far only describes the ontology that will be used for
> the resulting RDF.  The mapping itself is still marked as a "to do", so
> I cannot comment on that.  I wonder: Will the mapping be defined in
> XSLT?  That would be really convenient if it is feasible.  And if not, I
> am curious to know why not, since the need to map from the XML world to
> the RDF world is likely to be increasingly common.

We intend to have the mapping from the component model to RDF, not
directly from the XML. But at least I am planning to create an XSLT
transformation as one implementation of this, so if it looks correct and
complete, it might also become a part of the spec itself.

> The document notes that the customer base for this work product is
> unclear.  However, even apart from its value for direct use, I view this
> work as a valuable exercise in bridging from XML to RDF.  Lessons
> learned will be helpful to others.

I believe you are talking about the status section - it is intentionally
direct and threatening so that people actually speak up if they would be
the customer base. If the only identified value would be the exercise
itself, putting the result as a Note would be enough, but the charter
currently says it should become a Rec.

> The design of the ontology corresponds almost directly to the design of
> the WSDL 2.0 component model, which makes it easy to understand.  The
> deviations also seem to be straightforward and sensible.  In general,
> the document is written quite clearly.
> 
> The document clearly says that the ontology is less constraining than
> the WSDL 2.0 specfication.  I wonder: would it be feasible to itemize
> the constraints that are present in the specification but not enforced
> by the ontology?  To what extent would this be feasible?  If it is
> feasible I think it would provide greater insight into the ontology. 

I will try to provide such a list. However, I don't think now that it
would be useful to provide a full list; instead I will aim for a list of
examples and major things.

Thanks a lot again, I'll reply to the rest later,

Jacek

> 
> COMMENTS BY SECTION
> 
> [snip]
> 
> David Booth, Ph.D.
> HP Software
> dbooth@hp.com
> Phone: +1 617 629 8881
>  
>

Received on Thursday, 22 December 2005 22:07:37 UTC