Re: Unresolved issue re media type

Hi Brian.

I'm CCing www-archive ...

On Mon, Jan 05, 2004 at 09:49:13AM +0000, Brian McBride wrote:
> Hi Mark,
> 
> I believe this relates to WG issue:
> 
>    http://www.w3.org/2000/03/rdf-tracking/#mime-types-for-rdf-docs
> 
> You will see that the WG resolved to register a single mime type for RDF.

Yes, I saw that, thanks.  But that decision was made before I raised my
issue.  I also didn't see any discussion in the resolution of that issue
that suggests that the group considered my issue.

> There is also a WEBONT decision not to register separate mime types for 
> Owl or any of its variants:
> 
> "RESOLVED: to close 5.13 without making a new OWL mime type, observing
> that the existing mime types, e.g. app/rdf+xml and app/xml are
> sufficient."
> http://www.w3.org/2001/sw/WebOnt/webont-issues.html#I5.13-Internet-Media-Type-for-OWL

Yes, I think that's much more closely related.

> The RDFCore issue
> 
>    http://www.w3.org/2000/03/rdf-tracking/#rdfms-assertion
> 
> may also be relevant to this.  In part it concerns what entailments a 
> publisher of RDF endorses.  The director has agreed to this issue being 
> postponed, and further discusion is happening in the semantic web 
> meaning forum:
>
>    http://lists.w3.org/Archives/Public/public-sw-meaning/

Yes, that is also very relevant to my issue, though I don't believe that
the decision that has been deferred could yield a future resolution that
addresses my issue.  I could be wrong about that, but I don't think so
based on what I've learned from following public-sw-meaning since its
inception.

> Also of relevance might be the property test:entailmentRules described 
> in the test cases document:
> 
>    http://www.w3.org/TR/rdf-testcases/
>
> This is used to specify what entailments should be applied when running 
> a test case.  Perhaps this or a similar mechanism will meet your needs.

Ah, I wasn't familiar with that.  Very interesting and relevant.

If that feature were part of the RDF model, it may very well have
sufficed (though would still be suboptimal).

> Are the current specs acceptable for this version of RDF?

No, I don't believe so.

From the discussions I've had with folks on this issue, it seems as
though I'm certainly in the minority, so I'm not exactly expecting to
have this resolved in my favour.  But it would be nice to have the WG
respond officially about its position on this subject so that I have
the option of going to the TAG with a URI to the decision.

I don't believe it's a serious enough issue - all things considered -
to hold up going to Rec, though I do believe it (the use
of application/rdf+xml for messages which assume RDF Schema inference)
an architectural flaw that will need to be addressed at some point.
Hence my request for a decision to be rendered.

Thanks for your prompt and thorough(!) investigation into this matter.

Mark.
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca

Received on Monday, 5 January 2004 11:03:44 UTC