Re: Semantic Review

> Given the scope of my review activity
> http://lists.w3.org/Archives/Public/www-webont-wg/2002Dec/0320.html
> 
> this review is somewhat light ...

Unfortunate but understandable.

> I reviewed the copy dated 20th [D]ecemeber, plus the mapping rules of 30th 
> [D]ecember.
> 
> Miscellaneous:
> Seems to be missing discussion on the new ontology versioning stuff; in 
> particular it seems to be illegal to use any of it, since it uses the OWL 
> namespace and the only things allowed in OWL DL from the OWL namespace are 
> owl:Thing and owl:Nothing.

In the abstract syntax these are just ontology annotations, so they work
there.  The only problem is that the translation needs to be fixed to allow
them, which I just did by changing text near the end of Section 4.1 to
        the abstract syntax form does not mention
	any of the URI references from the RDF, RDFS, or OWL namespaces
	that are given special meaning in RDF, RDFS, or OWL
	except owl:Thing and owl:Nothing.

> Section 2.
> 
> Para 6:
> "These are roughly ...OWL/DL as defined in Section 5.3.1"
> Hmm is that really where the definition of OWL/DL is, I would take 4.1 to be 
> the definition.

"as defined in Section 5.3.1" removed.

> Para 7:
> "Named XML Schema non-list simple types ... name of the type."
> Did we decide this?
> I thought we decided not to do this.
> (I support this text though).

Unfortunately, it is now out.  

I had just recently reluctantly made the change. :-)

> 2.1
> 
> Para 3 (excluding examples)
> "If a URI reference .... as well as the ID of an individual."
> 
> While the abstract syntax and semantics may permit this, it seems disingenuous 
> to discuss it given that we explicitly forbid it in the exchange syntax (see 
> section 4.1).
> 
> Suggest change paragraph to apply the same restrictions as applied in 4.1

This section includes ``... although the ontology cannot then be translated
into an OWL/DL RDF graph.''  I'm not sure when I added this caveat.

> 2.2 
> end of para 1:
> Suggest replace:
> "The syntax  here is set up to mirror the normal RDF/XML syntax."
> with
> "The syntax  here is set up to mirror striped RDF/XML syntax without the use 
> of rdf:nodeID".
> 
> (I believe that in the last call candidate for RDF/XML syntax there is an 
> appropriate link for striped RDF/XML syntax).

OK, except that the LCC RDF Core documents are not appropriate for
linking.  It would be better to point to old versions than to LCC versions
that were changed under us, or never got published.

> 2.3.13
> (near end of subsection)
> [[Individual-valued properties that are transitive or that have transitive 
> sub-properties, ...]]
> i..e. all individual-valued properties, since the empty property is transitive 
> and is a subproperty of any other property.
> No suggested rewording though ...

There is no such thing as the empty property in OWL.

> 3.
> Para 1.
> Delete "much"
> (it may be true, but it is probably better to let the reader judge).

OK, ok.  Grumble, grumble.

> 4.
> Mapping rules
> Question:
> can triples of the form
> <datatypeID> rdf:type rdfs:Datatype .
> be deleted from the transformation.

Some can.  I removed the ones that arise from typed literals.

> 5.1
> Para 3:
> Suggest delete "Further, OWL/DL can be processed by efficient DL reasoners" 
> given Ian's recent comments.

Already been done.  :-)

> 5.2
> typo?
> "not adverse" => "no adverse"

Corrrect.

> Appendix B.
> Is this in need of updating?

Yes, yes.  Even the section headings need work.  :-(
This will be done soon. 

> References
> How about referencing the newer RDF documents?

I've upgraded to the latest published ones.

This brings up an important issue.  How can we go to last call without
reviewing our documents against the last call RDF Core documents?   Does
W3C process even allow this?

> Jeremy

peter

Received on Tuesday, 31 December 2002 13:28:45 UTC