Re: [gtk-list] Re: Python (was Re: Proposal for a new project)
- From: Andreas Kostyrka <andreas ag or at>
- To: gtk-list redhat com
- Subject: Re: [gtk-list] Re: Python (was Re: Proposal for a new project)
- Date: Wed, 14 May 1997 12:20:34 +0000 (GMT)
On Tue, 13 May 1997, Michael K. Johnson wrote:
> Having seen sample code for PerlGtk, I can say that as it stands, I'm
> not sure I would want a direct translation. There's nothing intrinsically
> wrong with the PerlGtk bindings -- I'd just want a more completely
> object-oriented set of bindings. I'm only in the first stages of thinking
> about what I'd want in such a set of bindings, and I don't have any real
> ideas to offer here yet -- I just have a vague idea that I want the
> binding I use to be truly object-oriented and well tied into Python (better
> than PythonTk, definitely), and I want to be able to use them in code
Yep. But keep in mind, that speed is also important (I've got here a
15000+ line mini app in Python/Tkinter and speed bothers me.), so keep
the interface more on the C side.
> that looks compact; extra verbiage bothers me. I think that means
> exporting lots of symbols rather than using dictionaries['with', 'lots',
> 'of', 'text', 'strings'] and so on.
self['width']=avalue is not nice, but it's understandable if you
see that Tk is quite dynamic, so it's not feasable the override
__getattr__, because it could conflict with the actual python members
somewhere in the future :( But with gtk I've got the impression the stuff
is more managable, so a attribute based solution may be doable :)
Andreas
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]