Re: Proposal to resolve ISSUE-106 and ISSUE-109: namespaces for owl2 rdf/xml and xml/owl

Peter F. Patel-Schneider wrote:
> I propose to resolve ISSUE-106 and ISSUE-109 by making the owl2 and
> owl2xml namespaces be the same as the owl namespace.
> 
> Instead of the following namespaces
> owl 	      http://www.w3.org/2002/07/owl#
> owl2 	      http://www.w3.org/2006/12/owl2#
> owl2xml       http://www.w3.org/2006/12/owl2-xml#
> the situation would be
> owl 	      http://www.w3.org/2002/07/owl#
> owl2 	      http://www.w3.org/2002/07/owl#
> owl2xml       http://www.w3.org/2002/07/owl#
> Actually there would then be no need for separate namespace prefixes.
> 


I've discussed these with a couple of colleagues. Here are some thoughts:

1) owl & owl2 the same is good.

2) we should state a policy for management of owl namespace in our 
documents; and as an errata in the old document too.

e.g. the meaning of terms in the owl namespace will not change, but new 
terms may be added. (This would mean that if the meaning of say 
owl:onProperty is changed in a new design then we would change the 
spelling of onProperty and use a new name for the new meaning).

3) it would be a mistake to combine the owl namespace and the owl2xml 
namespace.

Rationale for 3:
    a) potential confusion between owl/xml and rdf/xml
    b) an RDF/XML doc using the single node construction might appear to 
be in the owl/xml namespace, which would license the GRDDL transform for 
owl/xml
    c) the triple view and the functional syntax view of an ontology are 
sufficiently different to merit different namespaces

Jeremy

Received on Thursday, 17 April 2008 13:53:53 UTC