Comment 31 for bug 1855736

Tim, your issue in comment 26 is confusing to me. It looks like it's caused by my patch in comment 6?

Specifically, it looks like subprocess.Popen returns a unicode string as stdout for you. I can't reproduce locally, in either python2.7 or python3.7:

>>> import subprocess
>>> subprocess.Popen([u'ls'], stdout=subprocess.PIPE).communicate()[0].decode('utf8')

The documentation for Popen says:
"If encoding or errors are specified, or text is true, the file objects stdin, stdout and stderr are opened in text mode with the specified encoding and errors, as described above in Frequently Used Arguments. The universal_newlines argument is equivalent to text and is provided for backwards compatibility. By default, file objects are opened in binary mode."

None of the arguments that would cause Popen to return a unicode string seem to be passed to it by duplicity...?