W3C home > Mailing lists > Public > www-webont-wg@w3.org > August 2002

Re: WOWG: agenda Aug 15 telecon

From: Frank van Harmelen <Frank.van.Harmelen@cs.vu.nl>
Date: Thu, 15 Aug 2002 16:40:16 +0200
To: www-webont-wg@w3.org
Message-ID: <1471007466.1029429616@swpc243.cs.vu.nl>

Frank wrote:

> > Message [1] points out a bug in the DAML+OIL spec that we are about to be
> > bitten by as well if we don't take care. (Essentially the current
> > DAML+OIL  spec of sameClassAs forces its arguments to be of type
> > daml:Class, even if without that statement they would only be of type
> > rdfs:Class).

Peter replied:

> My understanding is that this was a feature, not a bug.  The use of
> sameClassAs in importing terms from RDF/RDFS is thus the bug.

That's certainly one way of looking at it, but realise that >*we*< committed 
this error, and that it would be a very natural thing to do for many people 
(including us, apparently), to apply sameClassAs to rdfs:Class's, with 
unexpected consequences.

Do you agree that there must be a way to import terms from RDF(S) ?
(serious question, because I'm not sure you do)

The authors propose to fix the bug by using equivalentTo, since its domain 
and range are so loose that they don't entail any unexpected consequences for 
their arguments.

I think if we just fix the bug in the DAML+OIL spec in this way and move on, 
then we've missed the point of the message: the use of sameClassAs has rather 
unexpected consequences, caused by (among others) the relation between 
daml:Class and rdfs:Class. I would hate to design OWL with the same gaping 
opportunity for bugs.

Frank.
   ----

[1] 
http://lists.w3.org/Archives/Public/public-webont-comments/2002Aug/0022.html
Received on Thursday, 15 August 2002 10:40:23 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:57:51 GMT