W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > April 2003

Re: pfps-08 test cases

From: pat hayes <phayes@ai.uwf.edu>
Date: Sat, 26 Apr 2003 17:05:29 -0500
Message-Id: <p05210611bad0ae3304f6@[]>
To: Jeremy Carroll <jjc@hpl.hp.com>
Cc: w3c-rdfcore-wg@w3.org

>I was surprised to see in the IRC log your claim that Peter is happier with
>the treatment of XMLLiteral ...
>I had a look at:
>to try and work out what the document currently says.
>As I see it, it says
>a) langauge tags are irrelevant for all other datatypes
><eg:a> <eg:p> "foo"@en^^<eg:d> .
>  entails
><eg:a> <eg:p> "foo"@fr^^<eg:d> .
>b) language tags are not irrelevant for rdf:XMLLiteral i.e.
><eg:a> <eg:p> "foo"@en^^rdf:XMLLiteral .
>does not entail
><eg:a> <eg:p> "foo"@fr^^rdf:XMLLiteral.
>Hence I conclude that:
><eg:a> <eg:p> "foo"@en^^<eg:d> .
><eg:d> owl:sameAs rdf:XMLLiteral .
>cannot entail in OWL
><eg:a> <eg:p> "foo"@en^^rdf:XMLLiteral .
>which I took to be the heart of Peter's comment.

You are right, and I hadn't noticed this.  Maybe Peter didn't either. 
But this wasnt the heart of the comment, seems to me.  What was 
bothering Peter was that in the older version, the semantic 
conditions on XML Literal were stated without referring to the 
denotation of XMLLiteral, only in terms of literals typed with a URI 
of that exact syntactic form, so that (never mind the lang tags) it 
wouldn't have been correct to infer XML datatyping from a literal of 
the form "foo"^^<eg:d> in any case. That has now been fixed for 
datatyped interpretations (but not, deliberately, for RDFS 
interpretations, so that RDFS engines are not obliged to implement 
datatype reasoning just for the built-in Dtype: they can treat XML 
literals as just another syntax category.}

Dos this help?

I agree that we need to fix the above issue over lang tags, if they 
are still going to be around.

Seems to me there are two ways out of this mess. One is to forget the 
idea of XMLLiteral being a datatype, and just have these literals as 
a special syntactic class in RDF. (There could be a comment in 
semantics 3.4 about how one can 'look at' XML typed literals as being 
datatyped by a kind of imaginary datatype, like the comment there now 
that plain literals can be seen that way.) The other is to get rid of 
the  lang tags from the literal syntax.  I guess I would prefer the 
latter, myself.  I gather than doing this without also purging the 
XML syntax is likely to mislead some readers; but tough, we have 
already confused more readers than we can shake a stick at; everyone 
brings some intellectual baggage to RDF which we need to teach them 
to put down first, seems to me, so why should XML folk be an 

Do you think we have a hope of getting rid of these tags altogether? 
They cause a LOT of trouble.

IHMC					(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola              			(850)202 4440   fax
FL 32501           				(850)291 0667    cell
phayes@ai.uwf.edu	          http://www.coginst.uwf.edu/~phayes
s.pam@ai.uwf.edu   for spam
Received on Saturday, 26 April 2003 18:05:33 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:24:22 UTC