WxPython versus Tkinter.

[CM]

On Jan 25, 10:13 am, Nicholas Devenish <misno... at gmail.com> wrote:

Nicholas,

> I think even more damaging to any python newcomers than choosing the
> 'wrong' gui toolkit would be stumbling across this thread whilst looking
> for a toolkit; and thinking some of the behaviour here was
> representative of the python (or wx) community as a whole, which
> couldn't be further from the truth. I know that if I had found this
> thread when looking around I would certainly have been put off of wx
> (which is the toolkit I decided on when looking around).

I don't know--you sound too reasonable to extrapolate from this goofy
thread to a huge toolkit project that has been around for years and is
used in project such as Audacity (that's the wxWidgets version, but
close enough).  But yes, it almost at times seemed like--from what I
could manage to read--this thread was a "psy-ops" (psychological
operations) trick to turn off wxPython adopters by associating it with
juvenile nonsense, and yes, on a quick scan it could turn people
off.

Which would be a shame, because, as you, Andrea, and others have
noted, wxPython is a nice toolkit.  For those interested, download it
and make sure to download the Demo, that shows what can be done with
it.  (Very early in this discussion the screenshots on the website
came up; they are horrifically out of date and wxPython deserves
better and looks great on, say, Windows 7 or Ubuntu....well, it looks
native, and that's the point).

> Perhaps there is room for a balanced, adult discussion on the future of
> GUI toolkits in python; But I don't believe that this can happen here
> without substantial changes to a certain persons attitudes (or other
> peoples kill files).

It would be more likely that wxPython would be in the stdlib than
those attitudes will change.  :D

But what I would enjoy is a discussion about GUIs in terms of "develop
once, deploy many".  For example, pyjamas, since I think being able to
develop one GUI that works as desktop or web-based is kind of
exciting.  Unfortunately, it seems it is far off from anything easily
usable at this point.  Part of that might be it doesn't have a big
enough community of developers yet.  It's also just really difficult,
I'm sure.

Another interesting issue in this is mobile phone app development.  It
is frustrating to devote a lot of time to learning a desktop widget
toolkit and Python and while that is occurring the culture moves more
and more toward app use in which that is not too applicable.  Some of
that cannot be helped if Apple, e.g., restricts what can be used on
their phones.  I guess for Android one can already develop with PyQt
and it will run on desktop or phone?