wxAUI => wxWidgets => wxPython?

Gavana, Andrea wrote:

Hello Robin,

thank you for the answer.

I've been thinking about it but haven't decided yet. I need to dig
into the code to see if there will be any issues with wrapping it
for wxPython. As far as whether it is better or not you may be
better able to answer that than I. Currently it is basically the
same code that you ported to Python, so if you've added some
functionality along the way then that is an advantage for keeping
the focus on PyAUI.

Well, Chris Mellon has done a nice work to adapt transparency things
(and to fix other issues) on the Mac for PyAUI. As far as I know,
wxAUI has no ability to deal with transparent windows on all
platforms except Windows. This can be a minor point, but in all the
GUI I wrote since PyAUI has been written, I always liked to use the
docking hint as transparent frame. It is much nicer than the ugly
blue rectangle. But it may easily be only a minor point.

Adding transparency support for the other platforms has been mentioned, and I expect that the wxMac code in PyAUI could be ported to C++ if needed. Plus since there is direct access to the API it may be fairly easy to support in wxGTK2 as well.

On the other hand, now that wxAUI has been added to the core wxWidgets and it has some active developers interested in improving
it and maintaining it then it is less likely to suffer from bit-rot
and leave Python programmers high and dry with old and rusty
unmaintained C++ code. So in that case it probably makes sense to
consolidate and have only one version of the AUI functionality so
there is no risk of divergence over time between the two projects.

Yeah, I see what you mean. I still have to understand how it will
exactly work: have the wxAUI developers given up? I mean, did they
put the code inside wxdev's hands and forget about it? If this is not
the case, we may end up having 2 developments: the wxWidgets one and
the indipendent Kirix one, which is not particularly nice.

IIUC the Kirix developers are the ones that requested that it be put into the wx CVS so it would be easier for them to work on it. I think it was something like their current infrastructure does not easily enable them to make open source releases, have a publicly accessible source code repository, etc. They said that they felt that it would be better for the code to be someplace where others could work on it and integrate patches besides just them.

In any case, my advice would be to wait for your final word about
wrapping it with wxPython. I don't know anything about speed issues,
I always thought that C++ things were much much faster than Python
ones, so this one may be a point against PyAUI and in favour of
wxAUI.

There will be some speed differences, but usually for UI things it doesn't matter because normally the bottleneck is the speed that the user is able to do things, not the speed that the event handlers run at.

I still don't know if you have actually tried PyAUI and wxAUI,
as I would have had some suggestion from your part about possible
bugs/misbehaviours and things to improve.

I did run the demos quite a bit, but every time I found something to report somebody else had already reported it and sometimes you had already fixed it. :slight_smile: I haven't taken the time to look at the source code yet though.

ยทยทยท

--
Robin Dunn
Software Craftsman
http://wxPython.org Java give you jitters? Relax with wxPython!