On Fri, Dec 12, 2008 at 1:23 AM, Casey Duncan <[email protected]> wrote: > On Dec 11, 2008, at 8:48 PM, Brian Fisher wrote: > [..] >> >> ...In fact, I think a reasonable thing for pygame to do is say you want >> support for 2.3? then here's an installer for 1.8.1, we kept it around just >> for you. > > I think this is the right way to approach this. You need to use an old > version of python, then use an old version of pygame. I couldn't care less > if folks are lazy, what matters is that the pygame developers lives aren't > complicated by supporting ancient versions of the interpreter. Removing > complexity from the development process is a very good thing.
I agree. If you are limited to py2.3, you are limited to pygame1.8. 1.8 is pretty good as is. But it's not really about lazy, it's about portability. The reason I stick with pygame instead of pyglet, is because pygame runs damn near ANYWHERE. From the oldest system imaginable, to the newest. I don't have any users stuck on python 2.3 fortunately, but I do have users stuck on windows 98. > Opposition to requiring 2.6 has been voiced, and I'd have to agree that at > this time there is little to be gained moving that far. Seems to me that > making the next version of pygame require 2.5 is reasonable. > > -Casey > > Yeah I don't think 2.6 should be required for at least a year or two. As mentioned above, I cut out a significant portion of users with 2.6. It stinks, because I want to upgrade, but there we go. I could just stick with pygame 1.8 though. So even still, I wouldn't argue too loudly if things went in this direction. Not to mention, I could probably use one version of pygame/python for these users specifically, and let everyone else upgrade. PyObjC has been the most difficult issue for my mac users, so I would give a cheer if that dependency were removed.
