Here is the problem:
On the non unicode version of wxpython, encoding *just works*. Non ascii stuff displays correctly, etc.
On the unicode version, doing the wrong thing can lead to a segfault.
While my solution fixes the segfault bit,
it means you now need to know what encoding you use even if you use the non unicode version.
SO:
1. I could add EVERY encoding listed in the python docs as an option in the prefs dialog, and in the file dialog.
2. I could have drpython discretely check wxPython for unicode support, and only check the encoding if it is found.
I think 2 might be a good idea. What thoughts?
(In the event of 2, I would change the documentation to read:
AutoDetect Unicode,
Default Encoding,
Only worry about these if your version of wxPython has unicode support built in. If you have the ansi version, do not worry about it.)