Re: agenda for rdfcore 2003-09-12 telecon

Graham, I'm also broadly sympathetic to your point.
That together with the fact that it's not obvious for
a FOL based reasoner to *pass* a negative entailment
test (theoretically not impossible as FOL is complete
but in practice I guess most implementations will be
incomplete) we then now have either fail or incomplete
for all negative entailment tests, so zero percent pass
(at least in our case if we consider it in that way...)

--
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/


                                                                                                                                        
                      Graham Klyne                                                                                                      
                      <gk@ninebynine.org        To:       Jan Grant <Jan.Grant@bristol.ac.uk>                                           
                      >                         cc:       w3c-rdfcore-wg@w3.org, ext Dan Brickley <danbri@w3.org>                       
                      Sent by:                  Subject:  Re: agenda for rdfcore 2003-09-12 telecon                                     
                      w3c-rdfcore-wg-req                                                                                                
                      uest@w3.org                                                                                                       
                                                                                                                                        
                                                                                                                                        
                      2003-09-12 07:50                                                                                                  
                      PM                                                                                                                
                                                                                                                                        
                                                                                                                                        





I'm broadly sympathetic to Jan's comments, but I do wonder where it leaves
the interoperability report if developers choose to implement something
that goes beyond strict RDF.  My impression was that a test-case failure is

to be regarded as a serious non-conformance.  I suppose we can say
something like:

    _:a eg:prop " 3 "^^xsd:integer .

does not entail

    _:a eg:prop "3"^^xsd:integer .

and I'm fine with that, because if an implementation chooses to infer the
second from the first, it's not violating the RDF specification, just
making an inference not required by it.

In saying this, I have no strong opinion about the handling of whitespace
around XML schema datatypes, and would really prefer to defer to the XML
schema working group on this.

#g
--

At 11:33 12/09/03 +0100, Jan Grant wrote:

>On Fri, 12 Sep 2003, Patrick Stickler wrote:
>
> >
> >
> >
> > 9:  Brian's alternate design writeup: summary of pros/cons
> >
> > See
> http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Sep/0099.html and
> > subsequent discussion.
> >
> >       If I am absent, please record my objection
> >       to Brian's "discussed" design, on the (primary)
> >       basis of backwards incompatibility.
> >
> > 12: " 3 "^^xsd:integer / xml schema and whitespace: can this be wrapped
up?
> >
> > http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Sep/0085.html
> generated
> > a huge amount of traffic. Where are we up to?
> >
> >       I would like to officially propose that the WG's
> >       last decision, to not fudge and to remove the
> >       problemmatic test cases (due to ambiguity
> >       in the XML Schema specs) to be upheld and
> >       that no changes be made to our design nor no
> >       attempt be made by us to interpret/clarify the
> >       XML Schema specs.
>
>While I'm ambivalent about the resolution of xmlsch-02, I'd like to
>propose we keep the test cases involved:
>
>- they represent our considered opinion on the xmlsch-02. I've heard
>PatS argue strongly that the situation in those test cases is _exactly_
>what the XML schema specs say; there certainly appears to be no
>ambiguity in his interpretation of those.
>
>- this appears to be a point where interop between implementations may
>fall down.
>
>On the other hand:
>
>- we've had implementor feedback that existing XML/XSD libraries don't
>do the right thing;
>
>- I'm sorry to say that I have the impression that the withdrawal of
>this test case was motivated by the desire to "smoothe the process".
>
>I expect to be at the meeting; in any case unless there's a better
>technical argument I object to the removal of the test cases in
>question.
>
>--
>jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/
>Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/
>Political talk? / What is said can be unsaid / with good old BS
>   -- ASCII haiku

------------
Graham Klyne
GK@NineByNine.org

Received on Friday, 12 September 2003 16:10:07 UTC