Once that is done I'd like to have it run the tests after every commit and report any failures to the dev list. It would then be the offending committer's responsibility (well primarily at least) to fix the problem as soon as possible, much like a build failure. If we can't get everyone to agree to take on that responsibility then I might as well stop now because I'll be damned if I'm going to spend any more time fixing tests that I didn't break :-)
Any thoughts?After that I'll starting looking at what can be done to improve the test tools and make tests easier to write, start looking at selenium and start writing tests for open bug reports to be committed with the fix so that they never occur again.
All of this should help us increase the stability of the trunk and our confidence when taking a checkout that we won't half to spend our development time fixing things that used to work. It'll hopefully also encourage the community to contribute more tests with the knowledge that doing so will increase the stability of the functionality they depend on. Fix a bug and you're good for a day, write a test and you're good a lifetime :-)
Regards Scott
smime.p7s
Description: S/MIME cryptographic signature
