- From: Jeremy Carroll <jjc@hpl.hp.com>
- Date: Tue, 30 Oct 2007 12:13:46 +0000
- To: public-owl-wg@w3.org
Three possible approaches to publishing tests:
A) as an incremental revision of OWL 1.0 Test Cases
B) as a separate rec track doc that lists only new OWL 1.1 tests, and
quite possibly uses different conventions etc.
C) off the rec track
Sandro:
> (Test Cases is a whole other question, since it's more of a database
> dump than a text document [1].)
here are some options:
A)
resurrect code that ran the database and db dump ....
add new features to that code to support:
- additional OWL 1.1 tests
- additional annotations of old OWL 1.0 tests to reflect any changes
in OWL 1.1 to those old tests (e.g. an OWL 1.1 system would classify
this file as DL, even though an OWL 1.0 system would classify it as Full).
hence move towards a new OWL 1.1 test cases rec that is clearly an
incremental revision of OWL 1.0 test cases
B) ignore OWL 1.0 tests and have separate approach to OWL 1.1 tests
- could be an OWL 1.1 test rec
C)
- could simply be a set of OWL 1.1 tests, without a rec track document
====
Option (A) would minimally require some effort by me specifically to
kick it off - the old code was more abandoned on publication of the rec
than mothballed.
Option (A) comes with a process, the process used by the previous OWL
WG, for test approval etc. In any case, we may wish to either reuse or
revisit this process.
Jeremy
Received on Tuesday, 30 October 2007 12:14:33 UTC