----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/21765/#review43608 -----------------------------------------------------------
/trunk/qpid/python/qpid/messaging/driver.py <https://reviews.apache.org/r/21765/#comment77891> I think the distinction between null and a zero length string is important. So it may be the saslwrapper (or interface to that) that needs to be changed to return null where appropriate and a zero length string where appropriate. - Gordon Sim On May 21, 2014, 11:45 a.m., Pavel Moravec wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/21765/ > ----------------------------------------------------------- > > (Updated May 21, 2014, 11:45 a.m.) > > > Review request for qpid, Gordon Sim and Rafael Schloming. > > > Bugs: QPID-5778 > https://issues.apache.org/jira/browse/QPID-5778 > > > Repository: qpid > > > Description > ------- > > CRAM-MD5 requires an empty response string to be treated as null in > sasl_client_start - the patch handles this right. But I dont want to > introduce a regression like https://issues.apache.org/jira/browse/QPID-3522 > where this kind of fix broker EXTERNAL mechanism that interprets zero string > as a string and not null. > > I have no experience with EXTERNAL / SSL in Python client hence raising > review request. > > > Diffs > ----- > > /trunk/qpid/python/qpid/messaging/driver.py 1596542 > > Diff: https://reviews.apache.org/r/21765/diff/ > > > Testing > ------- > > CRAM-MD5 now works, haven't tested EXTERNAL. > > > Thanks, > > Pavel Moravec > >
