Hopefully, after adapting the code, python 3 could be better at handling the unicode issues among these. Maybe a little project reviewing the code for these issues would be good. We might consider problems like bug 1016380, where string objects need to be made into byte objects, somewhat related.
I don't have much experience in this area, but it seems like a nice thing to improve, and I'll try to look into it if I get time.
After reading bug 492290, I had suggested setting locale variables to 'C', but now I see that bug 616767 is about apport not starting under those conditions. (I haven't dug through to see if these different workarounds succeeded or failed with different versions of apport.)
It seems like apport has a number of internationaliz ation problems. (After typing that, now I see why they write i18n!) Like,
* bug 492290
* bug 616767
* bug 768613
* bug 827090
Hopefully, after adapting the code, python 3 could be better at handling the unicode issues among these. Maybe a little project reviewing the code for these issues would be good. We might consider problems like bug 1016380, where string objects need to be made into byte objects, somewhat related.
I don't have much experience in this area, but it seems like a nice thing to improve, and I'll try to look into it if I get time.
After reading bug 492290, I had suggested setting locale variables to 'C', but now I see that bug 616767 is about apport not starting under those conditions. (I haven't dug through to see if these different workarounds succeeded or failed with different versions of apport.)
Thanks.