Test status and overview

Having updated the Test editors draft, and noting Sandro's

 "During PR, we should only make changes
based on new information. "
http://lists.w3.org/Archives/Public/www-webont-wg/2003Nov/0095.

I have been reviewing the proposed tests, (after the approvals which we are 
currently voting on)

0 passes
AnnotationProperty-004
 (late addition - syntactic test really; semantically it's a no-op - two 
systems pass the syntactic test)

Once passed proposed
I5.3-014  
I5.5-007  
** Thing-005
cardinality-005
** description-logic-209
** miscellaneous-010
** miscellaneous-011
description-logic-905
description-logic-906
description-logic-909
description-logic-910
description-logic-666
description-logic-668

Twice passed proposed
** Thing-003  (and Jena makes two, see recent msg Dave Reynolds)
description-logic-208
Thing--004
imports-014
Restriction-006
someValuesFrom-001

Further changes needed
1) obsolete Thing-002 (now duplicates Thing-003)
2) change oneOf-004


Of these there are many that I think we could and should approve.
Of these there are some which I feel that some previous decisions sort of 
presupposed that these tests would be approved, so it seems, at least to me, 
somewhat self contradictory to go to PR without approving them.

Particularly crucial tests from the point of view of earlier decisions are:

** description-logic-209
** miscellaneous-010
** miscellaneous-011
** description-logic-208
 made as a result of feedback on other tests


** Thing-003
** Thing-004
** Thing-005
  illustrating changes made during CR

oneOf-004 approved test needing modification see message I forgot to send ... 
(oh dear)

I5.3-014 reflects Herman's review of RDF Semantics on our behalf.

***************************************************

I think we can dispose of *all* the remaining tests as follows:

EASY CASES:  Twice passed proposed:

approve the following:

** Thing-003  (and Jena makes two)

description-logic-208
Thing--004
imports-014
Restriction-006
someValuesFrom-001

===

harder: approve AnnotationProperty-004 becuase semantically there is nothing 
to test, and we have enough feedback from syntax developers

====

real hard - the following tests were made as a result of WG discussion, but 
have so far only been passed by one system:

** description-logic-209
** miscellaneous-010
** miscellaneous-011

I am not clear that it is coherent to move forward without approving these, 
but we don't really have enough evidence. I think we could approve them and 
revisit if we get bad feedback during the PR period.

====

subtle but small:
The following once passed tests should be approved. They are small enough to 
be verified by hand. Moreover we have implementor feedback (supportive of 
these tests:
http://lists.w3.org/Archives/Public/public-webont-comments/2003Nov/0026.html
)

I5.3-014  
I5.5-007  
** Thing-005

===
three of these tests are not particularly crucial so let's obsolete them

cardinality-005
description-logic-666
description-logic-668

====

The following tests are from an area of the test document that no one (except 
Jos) seems to have implemented. I resisted making them extra credit at the 
test subgroup meeting, but if we are clearing the decks that is probably the 
best course. i.e. make these extra credit:

description-logic-905
description-logic-906
description-logic-909
description-logic-910

====

Two further changes

1) obsolete Thing-002 (now duplicates Thing-003)
2) change oneOf-004 to not use xsd:short but use xsd:integer instead (to avoid 
making it extra credit as under the proposal we are currently voting on)


Given the timescales etc. I am not sure procedurally what we should do. I will 
try and phone Guus today.

Jeremy

Received on Thursday, 27 November 2003 06:46:03 UTC