The encoding test test_Issue_1713 (for #1713) tries to load a file encoded in UTF-32BE, and check that the timezone key is set to UTC.

The problem is it does not use_default_template=False, so the default rcparams are loaded, and the default for timezone is UTC already. It does pass fail_on_error=True, but that flag only validates known rcparams. Invalid rcparams print a message, but do not otherwise fail. Thus, nothing is actually verified.

The flags can be modified, and the test rightly fails. However, to actually fix it, I'm not sure what it's trying to do. Does changing LANG actually have an effect on a running interpreter? Is this test actually useful at all? Should we switch to utf-8 only?