To take a _very_ different example, we are close to PR for RDFa, and this includes over a hundred of test cases (note that RDFa is _way_ simpler than OWL:-). I am one of the implementers for RDFa and those were of a great help to me. More importantly, it also helped to clarify tons of issues in the RDFa spec. The way the test suite came into existence (Kendall, I presume the same happened in the SPARQL case) is that implementations accompanied the discussions on specifications, and implementation themselves provided a bunch of test cases for contentious issues. At least for OWL RL we already have experimental implementations (I guess Alan has one, I have one...), so harvesting test cases may be simplified by that... Ivan Kendall Clark wrote: > On Thu, Aug 14, 2008 at 6:39 PM, Jim Hendler <hendler@cs.rpi.edu> wrote: >> great. Will we have test cases for all the profiles? That could make a lot >> of this discussion moot... > > I have no idea if this has been talked about, but my vote would be +1 > for test cases covering everything that is, in fact, testable. While I > had nothing whatever to do with it, IMO the test suite is the best > thing about SPARQL, for example, and I see no reason why that > shouldn't be the case for OWL2, too. > > Cheers, > Kendall -- Ivan Herman, W3C Semantic Web Activity Lead Home: http://www.w3.org/People/Ivan/ PGP Key: http://www.ivan-herman.net/pgpkey.html FOAF: http://www.ivan-herman.net/foaf.rdf
Received on Friday, 15 August 2008 07:37:28 UTC