I'm tempted to just dump the trunk's version of httplib.py onto the release branch. Poring over logs reveals that nearly every checkin is marked as a bugfix candidate (occasionally with a question mark). Certainly it seems ot would be easier to work the checkins that aren't bugfixes out of the trunk than work those that are into the branch, if you see what I mean. But I've never used httplib, so I thought I should ask here first (and Cc: the people responsible for most of the changes). Comments? Cheers, M. -- I've even been known to get Marmite *near* my mouth -- but never actually in it yet. Vegamite is right out. UnicodeError: ASCII unpalatable error: vegamite found, ham expected -- Tim Peters, comp.lang.python
"MH" == Michael Hudson
writes:
MH> I'm tempted to just dump the trunk's version of httplib.py onto MH> the release branch. Poring over logs reveals that nearly every MH> checkin is marked as a bugfix candidate (occasionally with a MH> question mark). MH> Certainly it seems ot would be easier to work the checkins that MH> aren't bugfixes out of the trunk than work those that are into MH> the branch, if you see what I mean. MH> But I've never used httplib, so I thought I should ask here MH> first (and Cc: the people responsible for most of the changes). MH> Comments? I think it makes sense to make httplib identical. The changes to httplib have all been intended to make it more robust. My one worry is that a set of changes I made may have broken pipelined https requests in order to fix a different set of bugs. I had intended to check whether pipelined https requests actually worked in 2.2.1. Jeremy
mwh> I'm tempted to just dump the trunk's version of httplib.py onto the mwh> release branch. Poring over logs reveals that nearly every checkin mwh> is marked as a bugfix candidate (occasionally with a question mwh> mark). ... mwh> But I've never used httplib, so I thought I should ask here first mwh> (and Cc: the people responsible for most of the changes). The one change I applied to httplib since the 2.2 release was to fix a problem with invalid urls. If a colon follows the server name but is followed by a non-numeric string or no string at all before the start of the path, an InvalidURL exception is raised. This is definitely a bugfix candidate. Jeremy's hand has been on that module much more heavily. I think it should probably be his call. Skip
participants (3)
-
Jeremy Hylton
-
Michael Hudson
-
Skip Montanaro