If you send unicode strings out, please notice that:
# a string with \u unicode characters
mystring = '\u00bfC\u00f3mo es usted?'
# or ...
mystring = "¿Cómo es usted?"
bytestring = mystring.encode("utf8")
This gives the same bytestring that can be used with quopri.encodestring(bytestring) in the code I mentioned. Bytestrings (bytearrays) are probably the reason why the unicode() function has been dropped from Python3.