Re: swrl.owl is OWL Full and Protege does not support OWL/XML

[These are my personal opinions, and not necessarily those of the JC.] 

From: nvdesai@ncsu.edu
Subject: swrl.owl is OWL Full and Protege does not support OWL/XML
Date: Tue, 12 Oct 2004 16:11:10 -0400 (EDT)

> As this thread points out in
> http://lists.w3.org/Archives/Public/www-rdf-rules/2004Aug/0003.html
> swrl.owl is OWL Full because of rdf:Property type of argument2.
> 
> Maybe this has been pointed out earlier,
> but another (and probably more significant) problem is the definition of
> AtomList as a subClassOf rdf:List. And OWL normative spec. says clearly
> that rdf:List is not in OWL DL.
> 
> What are the plans of the JC on this ?

I do not believe that there are any plans in the JC to do anything here.  

I don't see any problem with such documents being in OWL Full for two
reasons: 
1/ Such documents talk about the syntax (yes, this is a bit of abuse of the
   word "syntax", but it gets the point across) almost always have to be in
   OWL Full (or some other formalism that allows messing with syntax). 
2/ I don't see much point in these documents.  They don't have any
   normative impact.  They don't say much about the meaning of the
   formalism in question, be it RDF, RDFS, OWL, SWRL, or whatever.

> Also, the SWRL spec. seems to prefer the OWL XML presentation syntax over
> OWL RDF syntax. A problem for developers and users is that current
> ontology editing tools e.g. Protege *do not* work with XML syntax (and
> probably they will never). So for any serious use/development, RDF syntax
> is a requirement. Hence, swrl.owl is important. What are the plans of JC
> on this issue ? is/will RDF syntax be deprecated ?

I do not believe that the JC has any plans here either.  

The presence or absence or status of swrl.owl has nothing to do with
whether there is an RDF encoding for SWRL, so I don't see why swrl.owl is
important.

I, personally, would be very happy if the RDF syntax was deprecated.
However, this would require a significant change to the W3C Semantic Web
vision, and thus is unlikely to happen quickly.


There are tools (Protege, OWLEd, etc) that can insulate you from ugly
exchange syntaxes even when they use the ugly syntax.  Yes, currently some
of these tools do not use the OWL XML syntax, but it is a simple matter to
fix them to do so.  You could even consider it an opportunity - you could
establish your credentials by providing interfaces (for Protege, perhaps)
that allow a tool to use the OWL XML syntax as an exchange syntax.  This
would help pave the way for an extension of the tool to handle SWRL.

> Thanks,
> 
> --
> Nirmit Desai
> Graduate Student
> Department of Computer Science
> NC State University

Peter F. Patel-Schneider
Bell Labs Research

Received on Thursday, 14 October 2004 11:52:01 UTC