Thanks for the report, converting the unicode part to a byte string sounds like the right idea, but unfortunately just using str() is probably not enough, some kind of encoding probably needs to happen. Follow up in the existing bug if you have more thoughts.
Thanks for the report, converting the unicode part to a byte string sounds like the right idea, but unfortunately just using str() is probably not enough, some kind of encoding probably needs to happen. Follow up in the existing bug if you have more thoughts.