Re: model theory for RDF/S

From: Ora Lassila <daml@lassila.org>
Subject: Re: model theory for RDF/S
Date: Mon, 01 Oct 2001 08:18:29 -0400

> Peter,
> 
> Peter F. Patel-Schneider wrote:
> 
> > There are other schemes that could make RDF syntax context free and
> > similarly leave RDF as weakly typed.  I would be very careful in adopting a
> > particular scheme for RDF literals without considering the alternatives.
> 
> We should consider all the alternatives. The idea of literal types was 
> once discussed (in the M+S WG I recall) and then conveniently forgotten, 
> leaving this "we have string literals and then some magic happens, 
> giving us righ data types...".
> 
> Did I mention I don't necessarily like the XML Schema datatypes?
> 
> 	- Ora

I don't think you did.  I, similarly, find some stuff in XML Schema,
including some in XML Schema datatypes, that is not completely laudable.
As an academic it would be totally appropriate for me, or some group of
like-minded people, to propose a completely different mechanism for
providing typed data for XML.

However, as a group that is working within the W3C, I think that the RDF
Core working group should be extremely cautious of providing something
counter to XML Schema datatypes.  I know that the people who have been
putting together DAML+OIL have tried extremely hard to maintain
compatability with other W3C efforts, even though we are not part of W3C,
even when many of us have felt that there were technically better ways to
proceed, and even during this time when I feel that our efforts are not
being supported very well by the W3C process.  Should not the members of
the RDF Core WG, who are much more closely tied to W3C, expend even more
effort to obtain compatability between RDF and other W3C efforts?

In closing let me mention one potential ramification of divergance between
RDF and other W3C efforts.   If RDF does not follow other W3C efforts, then
why should other groups, such as the DAML+OIL people, follow RDF?

Peter F. Patel-Schneider

Received on Tuesday, 2 October 2001 17:31:21 UTC