Closed devurandom closed 11 years ago
This should definitely be included. Works in Python 2 and 3.
Since the header needs to be encoded eventually, the original method stored the header as bytes rather than a string. This regression came from 2736b3d9ce1b67cec9daf6e0c9a20ed56b3ff2e7. It might be ideal to just do the same as it used to do, in case for any reason someone needs to use a non-UTF8 user/pass.
Oh, sorry! I thought the .encode
would not have such an impact. It's probably better to do it how it was.
Closed, obsolete
Currently, the authpair is encoded as b'...' within the HTTP request, which is obviously wrong. This commit fixes it.