Re: status of RDF, RDFS, and OWL ``namespace files''

From: Dan Connolly <>
Subject: Re: status of RDF, RDFS, and OWL ``namespace files''
Date: 03 Jan 2003 00:56:22 -0600

> On Thu, 2003-01-02 at 21:30, Peter F. Patel-Schneider wrote:
> > Hi:
> > 
> > What is the status of 
> >
> >
> > Are they normative parts of the RDF specifications?
> I believe the latter is a normative part of the RDFS
> spec; i.e. its contents are part of the tech report.
> That's not the case for the former.
> >   I don't see how, because
> > 
> > 1/ Neither of them are valid in the RDF Model Theory or the RDFS model theory.
> No?
> What leads you to that conclusion?

Well, for one, they contain rdfs:comment properties, which are not valid in
the RDF or RDFS model theories.

> The model theory spec defines validity of inferences; what
> does it mean for a document to be valid? Oh... do you
> mean that they're not entailed by the empty graph?

Sort of.  Valid means true in all interpretations, which is the same as
being entailed by a syntactic structure that places no restrictions on
interpretations (e.g., the empty RDF graph).

> Good point; I think that's a bug, for RDFS; i.e. the
> rdfs:comment's and such need to be consistent in
> the 01/rdf-schema file and the text of the RDFS spec;
> and I guess you're right that the model theory spec
> needs to refer to them.

Well, there should be some consistency.  I would think that it would be
better to change the rdfs:comments to XML comments. 

> For 22-rdf-syntax-ns, I think it could be empty
> (or full of nonsense) without any specific technical
> impact. 

I would agree with this, but the Concepts document seems to indicate
otherwise.  For example, if this document contained something like

<s:Class rdf:ID="Property"
 s:comment="Every RDF statement has the extra meaning that if its subject
 is a person then that person is an idiot">

Then any entity that publishes an RDF (RDFS?) document that contains a
statement whose subject should (?) be considered to refer to a person is
insulting that person.

> But for practial purposes, it's probably
> good for it to contain stuff that all RDF users
> would agree to.

I think that there should be some sharper statement than ``all RDF users
would agree to'' that is true of the document.  

> > 2/
> >    does not have elements that correspond to all the elements of the RDF
> >    namespace. 
> > 
> > If they are not normative, what is their status?
> Umm... really handy web page?
> I'm not sure what sort of answer you're after.

> Actually, I'm not sure what you mean by "normative" either.

Well, perhaps normative is not the right concept here.   However, some sort
of statement concerning their status is needed, I think, particularly as
the RDFS document has been presented in the past as somehow containing the
essence of RDFS.

I think that it would be a good idea to 
1/ Change these two document so that they are valid in the RDFS model
theory, which would mean making quite a few changes.
2/ State that they are RDFS documents that provide part (but not all) of
the meaning of the RDF and RDFS vocabulary in RDFS.

(This would then also be done for the OWL document.)

> > One reason that I ask is that WebOnt has a similar sort of document
> >
> > with some of the same problems, but because of owl:imports the actual
> > contents of the document matters more.
> > 
> > In fact, owl:imports makes the contents of
> >
> >
> > matter more.
> Certainly the latter; 2002/07/owl imports 2001/01/rdf-schema, no?
> I can't think of any way that the contents of 22-rdf-syntax-ns
> would matter much.

Well, OWL users might think that importing it is a good idea, which makes
it matter somewhat more than it did in RDF, which has no notion of

> Dan Connolly, W3C


Received on Friday, 3 January 2003 07:19:18 UTC