From: Kevin A. <al...@se...> - 2002-01-11 22:11:10
|
> From: Dan Shafer > > > Ditto for Python one would assume. Actually, I suggest using > Python 2.2 as > > the initial target - the unification of the class hierarchy is > just too good > > to pass up. (Frankly, I think this was a big enough change in > the language > > to warrant a full bump to 3.0, but so far Guido hasn't sought > my counsel ;). > > One wonders what's in store for Python 3.0.) > > I completely concur. The class stuff in 2.2 has moved the language a > LOT closer to a "pure" object model. I thought it should also be > called 3.0 and I too wonder at what Guido must have in mind for > _that_ release! > > But for PythonCard, I think a 2.2 target is a solid idea. I hesitate to require Python 2.2 for the current prototype, 2.2 is just too new. I've been waiting for a solid ActiveState release of 2.2, but perhaps it is time I just went with the standard distribution? I have no problem with requiring 2.2 for the first real alpha which is not too far off, maybe a month or so before that's started? The changes in 2.2 definitely will help in cleaning up the current design. I'm just speculating at this point. Anyone writing code should make note of 2.2 features they use, so we at least know where it would be required. You can add asserts at the top of your modules for the features you require. ka |