2013-05-03, 20:38
(2013-05-03, 18:03)HeresJohnny Wrote:According to all the documentation you post here these unicode/windows bugs have mainly been fixed by version 3.0 of python. Unfortunately, XBMC still only using version 2.4. And actually using the getfilesystemencoding() function to convert paths/strings into the right file system encoding using by Windows do not work.(2013-05-03, 03:39)Angelscry Wrote: I'm sorry, but I'm not able to fix this problem. It is over my knowledge... and I do not understand why it happen. Paths containing non-ascii characters are stored and send correctly to the python function starting the command line... but the python function do not recognize the path correclty. Encoding paths to the system encoding before send it to the function generate en error. More stangely this problem only occurs under Windows... and all is working correctly under Linux.
So... if someone could bring its help and take a look at the Advanced Launcher code, it will be welcome.
I am by no means a programmer, but from what I gather online, Python uses UTF-16-LE in Windows and UTF-8 in all other OS's, so maybe that's your problem right there. Windows needs a different environment variable "_wenviron" as opposed to all other OS where it's "environ".
subprocess_hack.py, Line 678 could be a possible culprit.
I'm just blabbering without any real programming knowledge, though.
Also found this: http://stackoverflow.com/questions/17374...-in-python