A new test system

I've started working on a test system, because I think testing is one
of the chores in working on Anygui, and one of the things we need to
do if we're ever going to get to a release... ;)

It's basically a wizard-like program that lets you press "pass/fail"
for each test, and the tests are simple modules with a specified API
(very minimalistic for now) that are automatically loaded by the
test-program. I still have some modifications I want to do to it
before I start porting our existing tests (or those we actually wish
to keep), but if you want to check the program out, it's in

  /nondist/sandbox/magnus/tester

For some odd reason, cvs picked up on my including CVS in the list of
directories to add (because I used a glob), and now complains about
that every time I update etc. in this directory -- it doesn't seem to
do any harm, though; it just likes to whine about it :P

Anyway, I hope that investing some effort in the test suite will make
things easier -- I'll look into standardizing the test descriptions
too, so that checking whether a test works or not can be a no-brainer,
even for those tests we cannot fully automate.

And, oh yeah -- I used PyQt/qtgui for more or less the first time
while writing this. Yay!

-- 
Magnus Lie Hetland               "Nothing shocks me. I'm a scientist." 
http://hetland.org                                   -- Indiana Jones


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com