W3C home > Mailing lists > Public > www-webont-wg@w3.org > September 2003

RE: More tests

From: Ian Horrocks <horrocks@cs.man.ac.uk>
Date: Thu, 25 Sep 2003 19:01:19 +0100
Message-ID: <16243.11631.162340.335493@merlin.horrocks.net>
To: Jim Hendler <hendler@cs.umd.edu>
Cc: "Jeremy Carroll" <jjc@hplb.hpl.hp.com>, <www-webont-wg@w3.org>

On September 25, Jim Hendler writes:
> 
> At 3:58 PM +0100 9/25/03, Ian Horrocks wrote:
> >On September 25, Jeremy Carroll writes:
> >>
> >>
> >>  > One type of test that we are missing is small tests that, while
> >>  > potentially easy may prove difficult for some naive implementations.
> >>  > Attached is an example of such a test. The "TEST" class, and hence the
> >>  > ontology, is inconsistent. I would like to add several of these kinds
> >>  > of test to the test suite.
> >>  >
> >>  > I would be interested to hear how the various implementations fare on
> >>  > this test (FaCT can pass it in about 10ms, not including parsing).
> >>  >
> >>  > The test is currently in DL but could easily be converted into Lite.
> >>  >
> >>
> >>  I will add this, and also see if my autoconvert DL=>Lite code still works.
> >
> >Thanks.
> >
> >>  You suggested switching from DatatypeProperties to ObjectProperties or
> >>  vice-versa didn't you?
> >
> >Yes. The use of DatatypeProperties doesn't make tests qualitatively
> >any more difficult, but just means that you have to implement more
> >(i.e., minimal support for datatypes) in order to try them.
> 
> but this is a good thing, right?  We want people to implement all 
> aspects of our design...

If we taking this to its logical conclusion, then every test should
exercise all aspects of the design. In fact the intention of many/most
tests is to exercise some particular feature or combination of
features in isolation.

It is not a big deal though - in the context that datatype properties
are used in these tests, reasoners can simply treat them as object
properties without impacting soundness or completeness.

Ian

> 
> >
> >>
> >>  If you've more tests to donate, let's put them in.
> >
> >Working...
> >
> >Ian
> >
> >>
> >>  (It won't be til tomorrow at the earliest though).
> >>
> >>  Jeremy
> 
> -- 
> Professor James Hendler				  hendler@cs.umd.edu
> Director, Semantic Web and Agent Technologies	  301-405-2696
> Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
> Univ of Maryland, College Park, MD 20742	  *** 240-277-3388 (Cell)
> http://www.cs.umd.edu/users/hendler      *** NOTE CHANGED CELL NUMBER ***
Received on Thursday, 25 September 2003 14:09:17 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:02 GMT