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

Re: Please update OWL Test results -- today if possible

From: Sandro Hawke <sandro@w3.org>
Date: Fri, 12 Dec 2003 12:14:29 -0500
Message-Id: <200312121714.hBCHET8P017160@roke.hawke.org>
To: Jack Berkowitz <jack.berkowitz@networkinference.com>
Cc: www-webont-wg@w3.org, evren@cs.umd.edu, "Minsu Jang" <minsu@etri.re.kr>, "Sean Bechhofer" <seanb@cs.man.ac.uk>


That makes perfect sense.  Yes, the main page will be continually
updated.  The urgency comes mostly from the fact that we'd like the
Directory to make great claims about OWL implementations ("75% of the
X passed Y") in the announcement to the membership.  Some of the
claims we drafted are no longer quite met since we approved more of
the proposed tests, and it would be nice to get the numbers up again.

      -- sandro


> Sandro,
> 
> we only release results based on our publicly shipping version of our 
> software (3.0.3).  We actually have more results to share, but due to 
> needing to be sensitive to clients looking at that sheet, and then 
> saying "why the heck isn't it working in my release like it says on the 
> test results page?", we will be holding them off until we ship our 3.1 
> version, which will be in Q1 2004.  For various reasons, we can't share 
> what we intend to pass at that time, but it is a much larger 
> proportion.  We however only release engines and subsequent inference 
> results until we have gone through extensive scalability tests with the 
> algorithms.
> 
> My understanding is that the results page can be updated continually.  
> Is it your intention to freeze the results as of Monday or so, and then 
> stop the updating on the page?  If there is no objection, we'd just as 
> soon continually update our RDF feed as the state of the products 
> emerge, and would encourage the others to do the same.
> 
> Jack
> 
> On 12 Dec 2003, at 15:35, Sandro Hawke wrote:
> 
> >
> > If OWL progresses to Proposed Recommendation, as I hope it will do
> > very soon, the member companies will be asked to consider whether OWL
> > should become a W3C Recommendation.  They will be given a link to the
> > OWL Test Results page [1] as evidence of wide implementation, so the
> > more green on the page, the better the prospects for OWL.
> >
> > I just noticed that only one feed (Bossam) has been updated since the
> > changes to the editor's test suite [2] on 4 December.   Those changes
> > were mostly involved Proposed tests, so I don't expect any surprises,
> > but I think it would still be a very good idea to re-run the tests.  I
> > also hope some OWL implementation have improved.   We have a few 100%
> > columns on the reasoning tests; it would be nice to see a lot more.
> >
> > The target date for PR was three days ago, and we obviously didn't
> > make it, but it might still happen very soon.  So there's some urgency
> > here.  Improvements in the weeks to come would still be useful, of
> > course.  If you know someone producing data who I missed on the "To"
> > field, or who might not be reading their e-mail today, please try to 
> > let
> > them know.
> >
> > If there's a possibility that your system has regressed on some tests,
> > feel free to use a different URL for the new results, so you or I can
> > easily see the differences.  Just let me know the new URL, of course.
> >
> >    -- sandro
> >
> > [1] http://www.w3.org/2003/08/owl-systems/test-results-out
> > [2] http://www.w3.org/2002/03owlt/editors-draft/draft/Manifest.rdf
Received on Friday, 12 December 2003 12:19:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:03 GMT