W3C home > Mailing lists > Public > public-owl-wg@w3.org > October 2007

Re: Backward compatibility audit

From: Michael Smith <msmith@clarkparsia.com>
Date: Wed, 24 Oct 2007 17:35:26 -0400
To: Bijan Parsia <bparsia@cs.man.ac.uk>
Cc: public-owl-wg@w3.org
Message-Id: <1193261726.11840.97.camel@msmith-laptop>


On Wed, 2007-10-24 at 19:53 +0100, Bijan Parsia wrote:

> Some of the documents make BC claims like:
> 
> """" Every OWL 1.1 ontology can be serialized in RDF, so every OWL  
> 1.1 ontology in RDF is a valid OWL Full ontology. The RDF syntax of  
> OWL 1.1 is backwards-compatible with OWL DL, this is, every OWL DL  
> ontology in RDF is a valid OWL 1.1 ontology."""
> 
> It would be good to verify these.

Some verification based on C&P's implementation experience:

(As it is defined in the current docs) OWL 1.1 can be loaded into Pellet
using the OWLAPI, which does not have different modes for OWL DL and OWL
1.1 operation. When using OWLAPI + Pellet for the DL portions of the OWL
test suite all tests are passed (given a few exceptions due to
reasonable memory constraints and other issues that are not 1.1
related).

I.e, for one deployed OWL 1.1 system (OWLAPI + Pellet) reasoning over
OWL DL gives results consistent with the OWL DL test suite, and does so
without any special action to insure backwards compatibility.

Pellet paired with its Jena loader produces similar results.

Also, OwlSight [1] uses OWLAPI + Pellet and supports 1.1 and maintains
full backwards compatibility without any special code.
-- 
Mike Smith

Clark & Parsia

[1] http://pellet.owldl.com/ontology-browser/
Received on Wednesday, 24 October 2007 21:35:17 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:13:26 GMT