W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > November 2001

Re: Data Types Use Cases

From: Dan Connolly <connolly@w3.org>
Date: Tue, 13 Nov 2001 11:15:05 -0600
Message-ID: <3BF15519.7E0BF31B@w3.org>
To: Brian McBride <bwm@hplb.hpl.hp.com>
CC: rdf core <w3c-rdfcore-wg@w3.org>
Brian McBride wrote:
> 
> Thanks to Graham for his rapid and excellent response on a use case for data
> types in RDF.  I wanted to ask the WG for other complimentary use cases from
> their own experience, including specifically:
> 
>     RSS    Aaron
>     PRISM  Ron
>     DC     Danbri
>     SWAD   DanC

We haven't implemented much in the way of integers, floating
point, that sort of thing. We participated in the design
of DAML+OIL's handling of datatypes, but I'm not happy with
the result. The tricks in the model theory where the
object of <ex:shoeSize>10</ex:shoeSize> denotes an integer
conflict with one of our requirements, which is that
RDF documents be syntacitcally self-evident.
(cf literals must be self-evident Dan Connolly (Wed, Oct 17 2001) 
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Oct/0338.html )

We have a pile of use cases ( http://www.w3.org/2000/10/swap/test/ )
that depend on the interpretation
of the object of
  <dc:title>foo<dc:title>
as a string of three characters.

In particular, we represent dates in YYYY-MM-DD fashion
and sort them using string:lessThan and such. I suppose
string:lessThan could be defined as string-representation-less-than,
but we have other stuff that is more to the point...

Our implementation of instance of
(from http://www.w3.org/TR/2001/WD-rdf-mt-20010925/#skolem ;
it's called log:includes in our implementation)
relies on the assumption that if the characters
in a string literal are the same, then they denote the
same thing; i.e. xml:lang can't impact it.

In fact, that sort of literal matching underlies the
entire cwm rules engine.

We've only started implementing rdf:parseType="Literal";
that's a whole other story. See
parseType="Literal" as syntactic sugar for infoset description
         (#rdfms-literal-is-xml-structure) Dan Connolly (Wed, Oct 10
2001) 
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Oct/0153.html


-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Tuesday, 13 November 2001 12:17:10 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:42:37 EDT