Le 03/05/2016 01:45, Scott Kostyshak a écrit :
On Sun, May 01, 2016 at 03:17:31PM +0100, Guillaume Munch wrote:
Le 01/05/2016 00:26, Scott Kostyshak a écrit :
On Sat, Apr 30, 2016 at 03:44:48PM +0100, Guillaume Munch wrote:
This is candidate for 2.2.0 and 2.1.5.
How noticeable and bad is the regression?
On the one hand I did not notice any particular slowness, on the other
hand it nullified the fix to #7138 which was a performance complaint.
OK, I suggest not for 2.2.0 then at this point. No one has noticed the
regression for 8 months.
In the end I do not care personally if it is for 2.2.0 or 2.2.2. But I
must warn you against this reasoning. The commit has not yet found its
way into a release despite being made 9 months ago. In particular, that
nobody complained means little. Also, I suspect that some kind of bad
interaction may be behind #7138, like there has been between ibus and
the source panel for instance, explaining why it is hard to reproduce
(actually a bad interaction with ibus could very well explain #7138).
Guillaume
