WxPython versus Tkinter.

[rantingrick]

On Jan 24, 12:11 pm, Bryan <bryan.oak... at gmail.com> wrote:

> It is a provable fact that wxPython segfaults. You yourself proved
> that. That is, in and of itself, *not* a reason to pick some other
> toolkit. It's merely a datapoint. It's not a datapoint you can just
> sweep under the rug, however, like you seem to want to do.

Ok, now you are talking with facts and we can have a real debate!

Yes wxPython can be made to segfault by the user in some
circumstances. And YES this is very undesirable and it is the direct
product of a poorly written API. However! I do not blame the wxPython
creators or even Robin Dunn himself for this problem. Robin has stated
many times that he wants to keep wxPython as close to wxWidgets as
possible and he also said there is room for a "top level" API. This is
why wxPython proper has no business in the stdlib! We need an
abstraction API written in Python. So my point is, we need to create
this API independent of wxPython. You can think of it as "wx-inter" if
you like.

> I think if you re-read my post you'll see I don't disqualify it as a
> rich toolkit. wxPython is a fine toolkit. Better than tkinter in some
> ways, worse in others. segfaults is one aspect in which it is
> measurably worse.

Again: good argument! And i stand on my last reply. If you would be
interested i would like to work on this abstraction API with you. Are
you interested? How free are you? We need to work together to solve
this problem.