W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > June 2002

Re: Outline for new RDF document

From: pat hayes <phayes@ai.uwf.edu>
Date: Tue, 25 Jun 2002 10:12:48 -0500
Message-Id: <p05111b05b93e363819e3@[65.217.30.113]>
To: "Jeremy Carroll" <jjc@hplb.hpl.hp.com>
Cc: w3c-rdfcore-wg@w3.org

>I have been reviewing M&S to see which parts of it are:
>- not currently in any of our new specs
>- have not been decided to be dropped
>
>These sections appear to be candidates for inclusion in the new top-level
>document, fulfilling its dustbin role.
>
>The main content is to do with vocabulary, particularly:
>- containers
>- reification
>- rdf:value
>
>I understand that we are maintaining M&S containers but without formal
>semantic support and hence somewhere we need informal but normative

I do not understand what that is supposed to mean, and would strongly 
suggest that we do not make any informal text normative. If we wish 
to make some aspect of containers normative, then lets put that 
aspect into the model theory, which is the normative account of RDF 
meaning. The reason there is no formal semantic support for the M&S 
text on containers is because that text is simply incoherent: the 
intent that it expresses *cannot* be rendered in RDF, for essentially 
logical reasons. Rdf:Bags *cannot possibly* be bags. It seems to me 
to be a serious error to put into the spec as a normative passage 
something that is provably impossible given the rest of the spec.

>text
>corresponding to para 90.
>http://lists.w3.org/Archives/Public/www-archive/2001Jun/att-0021/00-part#90

OK, provided it is careful to not support, and preferably to 
explicitly repudiate, the content of paragraphs 93/94, which are in 
direct contradiction to paragraph 92.

>I understand that we are maintaining reification but clarifying it as
>supporting the provenance use case (and hence 'stating' rather than
>'statement'). As such text corresponding to M&S section 4 is needed - (all
>subject to complete rewrite).
>http://www.w3.org/TR/REC-rdf-syntax/#higherorder

Please take care to not use the term 'higher-order' in the context, 
which was a disastrous misuse of terminology in M&S.

>
>I do not understand what we are doing with rdf:value, but there is a lot of
>legacy which is based on stuff written in M&S and we should, IMO, carry some
>of that text forward even if we deprecate it.
>e.g. show some of the use cases.
>
>
>While this stuff might belong in the schema specification, the current
>schema draft appears to want to be very conservative and not suggest
>semantics that are not in the model theory.

Which is entirely appropriate, seems to me, in fact the only rational 
choice. I do not mean to suggest that the MT should rule the WG, but 
that if the MT as it stands does not reflect the intent of the WG , 
then it - the MT - should be brought in line with that intent. BUt 
there is no point in even having a precise semantics if the spec 
itself ignores that precision in its normative prose. Under these 
circumstances I would suggest simply abandoning the model theory 
altogether and reverting to the use of informal prose descriptions of 
meaning.

>e.g.
>http://www.w3.org/TR/rdf-schema/#ch_containervocab
>http://www.w3.org/TR/rdf-schema/#ch_value
>
>Assuming that we do want to carry such intent from M&S forward, and that the
>new document is the right home for that intent,

I do not think that any part of the final spec should express as 
normative any aspect of RDF meaning which is not reflected in the 
model theory. To do so makes the model theory worthless.

Pat

-- 
---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes
Received on Tuesday, 25 June 2002 11:12:51 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:49:22 EDT