On Wed, Feb 4, 2009 at 11:07 AM, Brian Fisher <[email protected]> wrote: > I know some computers will be a lot better with the directx backend - but > don't you think the actual expected breakdown matters? I mean, how many > older computers that break that way our there? how many older computers are > better with windib? Do we have a sense of those numbers? > > I'd be happy to test on a bunch of computers I have access to to see - what > is a game that reports fps that I can download that uses HWSURFACE's well? > What is a game that reports fps and doesn't use HWSURFACE that I can check > with? >
hi again, regarding testing the timings... Including the unittests with pygame should make testing this sort of thing much easier. Since now games should be able to get stats much more easily. You can see the timing of results like this: python run_tests.py -T 5 -d Also the examples.testsprite could be used to test various rendering routines with a little modification - you could get it to return the fps -- not just print it. That is included now with pygame too. We could probably easily make some code so that test results are recorded and returned to a central repository... well Nicholas already started on that project. If we made it easy for games to ask their users if it is ok to do... then run some tests and submit the results we could quickly get many 1000s of results. cu,
