Not typing 'encode' or 'decode' when dealing with strings. I know it sounds stupid, but it's really distracting compared to everything else in python that just flows easily.
Then you're doing it wrong. Sorry but that's definitely better solved in Python3 than in 2. Always decode/encode when opening a file/stream/whatever and only use unicode internally.
Sounds like you are shooting yourself in the foot. Your code may work with ASCII but will immediately break in non-naïve cases.