On 11/30/2010 04:41 PM, Alexandre Julliard wrote:
Jacek Cabanjacek@codeweavers.com writes:
I can see that these tests may be useful sometimes, eg. if someone is interested in old apps that don't run on new Windows. But the honest true is that it's not what happens with our tests. All we usually do with old Windows or old IEs is blindly (well, not always, you do better than that) marking them as broken. The result is that our tests have more complicated code and are less strict. Thus my personal strategy is different: leave win9x (or old IEs for that matter) alone as long as they don't cause troubles. As soon as there is a trouble with a test on a platform that I don't care about, I just disable the whole file. This way I don't waste my time on uninteresting platforms, the code stays cleaner and tests remain stricter, giving win9x tests a chance to prove itself. I've already sent quite a few patches applying this strategy to different tests.
The value of running tests on Win9x these days is certainly questionable. We don't try to emulate the Win9x behavior anyway, except in a very few cases (which most likely don't have tests...) so it only serves to document historical behavior that nobody cares about any longer.
I wouldn't be opposed to switching off win9x test runs and getting rid of the corresponding broken().
So the first thing is a patch to winetest.exe and the winetest website and at the same time tell Greg to exclude it from 'patchwatcher' and the batch running of the (almost) daily winetests.
Cleaning up the tests will come after that and will certainly be a nice long janitorial task.