As far as I know, I agree with the latest comment. One of the major point
in using 2.6 rather than 2.5 is its solid 64-bit support (at least on Mac
OS), whereas 2.5 is practically only usable in 32-bit.
With python 2.6 and wxwidgets 64-bits built on Cocoa, I can get a 64-bit
wxpython and thus a full 64-bit build of Grass. It is only experimental,
though: I can't do a thing with it (it is not even possible to open a
single file, the file selector code seems to be broken). I could not do
that with Python 2.5.
It will take a while for wxcocoa to evolve into a fully mature version
but, at least, the possibility of a full, modern 64-bit app is
demonstrated.
Don't get me wrong. I'm in favor of moving in that direction. However, I
just think that it is wise to complete the GUI with the version of Python
we started with before porting to a new one. It's especially important to
get 6.4 out the door as a stable version with as many bugs fixed as
possible. Then we can move on to GRASS 7 development. The other issue of
writing for Python 2.6 is that we need to make sure that the majority of
Mac, Windows, and Linux distros will be able to run the GUI. Mac just
moved to 2.6 (with the majority of users still at 2.5) and I'm not sure
where the other platforms are at currently.
I see no reason why wxpython on 2.6 should behave any differently that
wxpython on 2.5. My concern is more with the cocoa 64-bit port, but, at
this stage, I think that all the bugs I'm seeing are not caused by Grass,
but rather by wxcocoa being at an alpha stage and thus yet buggy. However,
I cannot be 100 % affirmative.