Re: document status update

>On Mon, 18 Aug 2003, Eric Miller wrote:
>
>>  > I'm waiting on one thing. As I understand it, the value spaces of
>>  > xsd:integer and xsd:decimal no longer intersect. If that is correct
>>  > then
>>  > one datatype test case needs changing, since it currently suggests
>>  > otherwise.
>>  >
>>  > The test case is
>>  > http://www.w3.org/2000/10/rdf-tests/rdfcore/datatypes/
>>  > Manifest.rdf#semantic-equivalence-between-datatypes
>>  >
>>  > which says that
>>  >
>>  > [[
>>  >	eg:foo eg:bar "10"^^<xsd:integer> .
>>  > ]]
>>  >  rdfs+dt(xsd:integer, xsd:decimal)-entails
>>  > [[
>>  >	eg:foo eg:bar "10.0"^^<xsd:decimal> .
>>  > ]]
>>  >
>>  > I'm waiting on a confirmation that this is, or is not, actually the
>>  > case, in light of recent xsd clarifications.
>>
>>  Are you waiting on a particular answer from the RDFCore group to
>>  repond? The XML Schema group? A particular individual?
>
>Hopefully for Pat to just confirm that I didn't imagine it when I
>thought I heard him say that this is now the treatment of XSD types: ie,
>that their denotation is a pair of (typename, value).

What you probably heard me say was a report of an email discussion 
with members of the XSD WG a while ago, trying to reconcile 
apparently inconsistent assertions in the XSD specs, which seem to 
define some value spaces in the same way, but also state 
categorically that all simple XSD datatypes have disjoint value 
spaces.  The upshot of this discussion was (1) the current state of 
the XSD specs is confused (2) the XML schema WG is aware of this and 
plans to fix it (in 1.1) in the near future, probably (3) by formally 
defining value spaces to be pairs of (type, value). However, that (3) 
was just one person's report, and from more recent email discussions 
with other members of the WG I have the sense that they will be using 
a different technique altogether, one which relies on making a 
distinction between values being 'equal' and values being 
'identical'.  I have sent them some rather critical remarks on this 
approach, but it seems clear that they are about as tired of this 
issue as we are of the datatyped literal issue.

I would strongly suggest that what we do at this stage is simply to 
not get into this tar-pit. Questions of what XSD values are 
identical/equal to what others should be left to XSD to sort out. If 
we have ANY test cases which presume ANY issues of identity or 
non-identity between xsd: datatypes, please let us simply remove them 
without further comment.  We have agreed that RDF plain literals are 
the same as xsd:string literals: that is all we need to do; all the 
other questions are up to other people to decide. The semantics has 
been carefully worded so as to refer all such issues to the defining 
authority of the datatype, which in this case is XML schema WG, not 
us.  If we try to second-guess what they are really saying, we will 
still be arguing about this stuff in 2010.

Pat

-- 
---------------------------------------------------------------------
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@ihmc.us       http://www.ihmc.us/users/phayes

Received on Friday, 22 August 2003 17:45:05 UTC