• Clean it up and use lesser privileges
• Use _unbind() instead

• Clean it up and use lesser privileges

Please see inline commentary.

This doesn't really have a "pending approval" or "require answers to questions" or somesuch state, so I'm going to have to make this re-appear on your radar by requesting changes.

Please see the inline commentary.

Let's make sure that self.pool = None instead, and test for self.pool is not None right where you made your changes -- A self.pool that exists can still not have a function .close() / .join().

This may be better suited for utils.normalize(), which already iterates over keys and values.

def _bind() may inadvertedly return True.

This issue still exists as per the commit revert.

The state tracking turns out to be invalid.

This may have already been fixed in GIT, but does not progress in Continuous Delivery because the unit tests keep failing.

Done

Pushing to mirror for Wheezy and Jessie

Builds completed, rebuilding for Kolab Enterprise 14 Santiago and Maipo

Builds scheduled.

Fails to build on Twenty Three with the same KDE_KIMAPTEST cmake variable ending up undefined.

Fails to build on Stretch as well, due to an undefined KDE_KIMAPTEST cmake variable. For now, patching out the inclusion of fbdaemon/tests/.

Fixed by importing kcalcore/event.h and kdatetime.h rather than KCalCore/Event and KDateTime.

Does not build in Winterfell on Maipo: