D10512: try to preload certain applets in a smart way
Marco Martin
noreply at phabricator.kde.org
Thu Feb 15 12:51:58 UTC 2018
mart added a comment.
sooo, benchmarks:
i tracked the time that occurs between mouse click and the return of setExpanded, which is reasonably near to the popup actually appearing (numbers consistent doing things more than once)
this is done on a fairly fast machine, would be interesting to do the same on a really slow one, like the pinebook
Kickoff: no preload, before incubator removal: 279 msecs
Kickoff: no preload, after incubator removal: 211 msec
Kickoff: with preload: 15 msecs (the first time still takes some tme to build all the nodes and texture, second time 0 msecs)
digital clock (with akonadi integration which makes the graphics a bit more complex)
digital clock: no preload, before incubator removal: 739 msecs
digital clock, no preload , after incubator removal: 234 msecs
digital clock, preload: 46 msecs (graphically quite complex, takes a more to create all textures, second time is 0 as usual)
media player: no preload, before incubator removal: 148 msecs
media player, no preload , after incubator removal: 89 msec
media player, preload: 0 msecs (graphics simple enough for texture creation to not register?)
Clipboard, no preload , before incubator removal: 191 msec
Clipboard, no preload , after incubator removal: 79 msec
Clipboard, preload: 0 msec
kicker: no preload, after incubator removal: 79 msecs
kicker: preload: 9 msec
REPOSITORY
R242 Plasma Framework (Library)
REVISION DETAIL
https://phabricator.kde.org/D10512
To: mart, #plasma
Cc: davidedmundson, broulik, apol, ngraham, plasma-devel, #frameworks, michaelh, ZrenBot, progwolff, lesliezhai, ali-mohamed, jensreuterberg, abetts, sebas, mart
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/plasma-devel/attachments/20180215/19804e76/attachment.html>
More information about the Plasma-devel
mailing list