Firstly, welcome to the tracker, and thanks for helping test the early releases of Python 3.6.

For future reference, it's much preferred if you can copy-paste error messages into the text of the message. Attaching large log files is okay, but screenshots should only really be used for things that are not text.

In this case, it would seem to be an issue with the pefile package. It is reading a file with a default encoding as part of its setup.py, but since the file appears to be part of its own package it really should specify the encoding of the file. It's likely that the file "works" under US English encoding, but obviously not for other encodings such as CP949. If the authors of pefile specify the encoding when they open the file, then this problem will go away.

(As it happens, I investigated the possibility of changing Python to handle this situation better, and unfortunately it is not feasible. We've provided the tools for authors to avoid these issues on their own, but we can't reasonably change how they work without breaking a lot of existing code. So it really needs to be fixed by the pefile authors here.)