W3C home > Mailing lists > Public > public-rdf-wg@w3.org > March 2013

Re: RDF semantics draft revision

From: Pat Hayes <phayes@ihmc.us>
Date: Tue, 5 Mar 2013 19:57:25 -0600
Cc: public-rdf-wg@w3.org
Message-Id: <C10C37D2-0DC2-4DDE-B2AB-2A5AA3365410@ihmc.us>
To: Antoine Zimmermann <antoine.zimmermann@emse.fr>

On Mar 5, 2013, at 11:34 AM, Antoine Zimmermann wrote:

> Le 01/03/2013 20:49, Pat Hayes a écrit :
>> The latest draft (pushed a few minutes ago, dated 1 March) of the RDF
>> semantics addresses, I believe, all the issues raised by Peter
>> recently. These suggestions greatly improved the document, thanks. I
>> hope I have eliminated all the colorful language, but Peter, feel
>> free to edit if I missed some.
>> 
>> FWIW, treating rdf:langString as a datatype is a royal PITA.
>> 
>> This draft has one other significant change, which I mentioned some
>> time ago. Interpretations are defined over *all* IRIs and literals,
>> rather than over a particular vocabulary V, and so all mention of
>> vocabularies has been eliminated. I see this as part of the KISS
>> process.
> 
> This contradicts our decision not to do this on the 7th November 2012.
> 
> http://www.w3.org/2011/rdf-wg/meeting/2012-11-07#resolution_4
> 
> to close ISSUE-97: "Should the semantics of RDF graphs be dependent on a vocabulary?", and you yourself said on the 6th November that:
> 
> """
> On Nov 6, 2012, at 12:42 PM, Antoine Zimmermann wrote:
> 
> > Re. ISSUE-97: Should the semantics of RDF graphs be dependent on a vocabulary?
> >
> >
> > I suggest we close it and do nothing.
> 
> I agree. I was keen on this idea for a while, until I started to go into the details of the effects it would have. Thanks for hewing a similar path.
> 
> Pat
> """
> 
> http://lists.w3.org/Archives/Public/public-rdf-wg/2012Nov/0067.html
> 
> 
> Semantics is rather late compared to other documents.  It could have been drafted while the concepts were being discussed.  

I was impossible to draft while decisions that would have affected the overall design of the document were still being discussed. 

> Thanks to these discussions, concepts have gone a long way and the document is fairly advanced.

I don't believe that this vocabulary issue affects Concepts in any way. 

>  Now we are in the final rush, we have to cover the semantics of the things we agreed on, without further additions, without reopening old issues, and so on.
> 
> 
> I find the semantics without a dependency to a vocabulary better and simpler, but if we do this, it must be a collective decision, with awareness of the consequences on entailment (it has significant consequences)

Does it? Can you sketch them for me? I couldn't find any, which is why I thought it would amount to an editorial decision.  AFAIKS, the only effect it has is to require interpretations to be infinite, which matters for description logicians (who will be mollified by the appendix mentioned) and irrelevant for everyone else. 

> . It means, we have to reopen the issue, take the time to discuss it by email, have a little talk about it using conference time and finally agree on a resolution.

I had indeed forgotten that this had been the topic of a WG decision. If the WG feel that this cannot be revized, then I will go through and insert all the vocabulary language in a final edit. Nothing else will change. RDF interpretations will still be infinite, because of the RDF container vocabulary. 

> 
> It seems to me hard to argue in favour of it, due to our time constraints.

Well, how about that it will take less time to do it this way? And that the semantics will be simpler, the document easier to read and easier to understand, and there will be no changes to any entailments.

Pat

> 
> 
> 
> AZ.
> 
> 
>> 
>> This makes all interpretations infinite, but an appendix will point
>> out that it is only necessary to interpret the part of this infinite
>> vocabulary which actually occurs in the graphs under consideration or
>> which have applicable semantic constraints (the rdfs: vocabulary for
>> RDFS-entailment, etc.), thereby allowing for finite interpretation
>> structures. (Some care is needed to handle the RDF container
>> properties. All the theoretical work behind this has been done by
>> terHorst, who will be duly cited.)
>> 
>> This keeps the main body of the semantics simpler and easier to read,
>> while providing the necessary information for logicians (who can be
>> expected to read appendices.) This appendix will be normative, and
>> will define the older terminology ("interpretation *of a vocabulary*
>> ") as before, maintaining conformity with spec documents which rely
>> on this aspect of the 2004 semantics. I hope y'all approve.
>> 
>> Pat
>> 
>> PS. If you are looking at a version dated 28 February, get the more
>> recent version before proceeding.
>> 
>> ------------------------------------------------------------ IHMC
>> (850)434 8903 or (650)494 3973 40 South Alcaniz St.
>> (850)202 4416   office Pensacola                            (850)202
>> 4440   fax FL 32502                              (850)291 0667
>> mobile phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
>> 
>> 
>> 
>> 
>> 
>> 
>> 
> 
> -- 
> Antoine Zimmermann
> ISCOD / LSTI - Institut Henri Fayol
> École Nationale Supérieure des Mines de Saint-Étienne
> 158 cours Fauriel
> 42023 Saint-Étienne Cedex 2
> France
> Tél:+33(0)4 77 42 66 03
> Fax:+33(0)4 77 42 66 66
> http://zimmer.aprilfoolsreview.com/
> 
> 

------------------------------------------------------------
IHMC                                     (850)434 8903 or (650)494 3973   
40 South Alcaniz St.           (850)202 4416   office
Pensacola                            (850)202 4440   fax
FL 32502                              (850)291 0667   mobile
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
Received on Wednesday, 6 March 2013 01:57:54 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:25:54 GMT