RE: test doc update

Jeremy, I had half a holiday and am only finding now your message
I fully agree with your much improved comments and have
immediately put them into the files (however manually)
and have them checked into w3ccvs
I also recompiled the wowg and now got a nice document (*)
however, still without those testcases

-- ,
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/

(*)(See attached file: OWL_Test_Cases.html)


                                                                                                                                       
                      Jeremy Carroll                                                                                                   
                      <jjc@hpl.hp.com>         To:       jos.deroo.jd@belgium.agfa.com, www-archive@w3.org                             
                                               cc:                                                                                     
                      2002-09-27 11:02         Subject:  RE: test doc update                                                           
                      AM                                                                                                               
                                                                                                                                       
                                                                                                                                       




> yes, time doen't permit very much

I'm getting there.
Thanks for the ref suggestions.

I think the descriptions of your test cases can be improved.
At the moment, the description is essentially echoing the test as triples.

How about:


owl:intersectionOf 001

[[
The order of the classes in an <code>intersectionOf</code>
construct is unimportant.
]]

(that understates the amount of time we have spent on that problem!)
I notice that your test uses only the weak semantics - more of that on the
mailing list and at the f2f. We will need to have also a strong semantics
test, and show a decision, which I suspect will be for the entailment.

owl:maxCardinality 001

[[
A <code>prop</code> with <code>maxCardinality</code> of two cannot take
three distinct values on some <code>sb1</code>.
]]


owl:maxCardinality 002

[[
A <code>prop</code> with <code>maxCardinality</code> of two cannot take
three distinct values on some <code>sb1</code>.
In this example, one of the three values is implicit.
]]

owl:TransitiveProperty 001
[[
A simple illustration of transitivity.
]]

owl:unionOf 001
[[
A union is a superclass of its parts.
]]

owl:unionOf 002
[[
<code>owl:unionOf</code> behaves quite like set theoretic union.
]]

Justification, we will need test cases that illustrate any divergence from
this.
e.g. a strong semantics one, where we comment on which semantics are being
used.

Note XHTML can be copy-pasted straight into the editor; the manifest has
xmlns="XHTML".

I expect to have final draft to you for your approval this morning.
Assuming you can OK it at the RDF Core telecon.

Jeremy