from preamble:

"Change Set:		SUnit-expected-failures-jf
Date:			30 July 2003
Author:			Julian Fitzell

Add the notion of expected failures to SUnit TestCases (this is against
SUnit 3.0 but may also work against 3.1 - just not actually tested
there).

You can define an #expectedFailures method on a test case that should
return a collection of selectors that are expected not to pass. 
Failures or errors in these tests will not cause the test case to report
failure.  If a test that is unexpectedly passes, this is also reported
as a failure.

The idea of expected failures in general is that you can write tests in
advance to describe what an interface should look.  You can then mark
them as expected failures before committing code.  This avoids the
problem of other developers not knowing which tests they have broken and
which are just expected to be failing currently.

This is quite a preliminary hack to get it working.  The UIs need to be
updated so that unexpected passes are identified differently in the list
of failures, for example.

I'm throwing this out there just because I may not have time to do
anymore in the next few weeks and I figured someone might have some
comments.  I'm almost tempted to write my own testing framework so this
feels like slightly less of a hack, but we'll see if there's ever time
for that. :)"!
-------------- next part --------------
A non-text attachment was scrubbed...
Name: SUnit-expected-failures-jf.cs.gz
Type: application/octet-stream
Size: 1522 bytes
Desc: not available
Url : http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20030730/14f2212e/SUnit-expected-failures-jf.cs.obj