This is great - I mean the report, not what it says :-)
I've been considering an extension of the abstract Keel unit test class that is specifically for testing models - this might be a good foundation for getting model unit tests out there, as that's clearly a big gap according to the report.
I think that will definitely help. Anything we can do to get rid of the tedious, repetitive tasks will make it more likely that we write tests :-)
In that vein, I also have a first pass of JUnitDoclet hooked up and it does a great job of generating empty tests, thereby eliminating the tedious parts. One of the things it warns about is not to get hung up on the *number* of tests, since it generates oodles of tests (one for each public method), so your count goes way up. This is where Clover will come in handy in evaluating the true coverage, and Jester to get us to make our tests better! I have added a per-module junittest capability in build-commons.xml, so it is easy to just concentrate on the tests for the particular module one maybe working on. Still need to get the whole container started up and configured, eliminating that may be the last thing we need to do to get rid of most of the reasons to not have unit tests :-)
Shash
http://keelframework.org/documentation Keelgroup mailing list [EMAIL PROTECTED] http://lists.keelframework.com/listinfo.cgi/keelgroup-keelframework.com
