Also remember which avatars we didn't find
ClosedPublic

Authored by vkrause on Aug 20 2017, 5:41 PM.

Details

Summary

This massively reduces the amount of network requests, at very little
disk and memory cost (1Mb per 20+k of missing avatars with both backends,
60+k for just Gravatar).

It's even more important when having both backends enabled, in that case
we always did a request for the first backend even if the second one could
provide it from the local cache.

Diff Detail

Repository
R88 PIM: Gravatar Library
Lint
Automatic diff as part of commit; lint not applicable.
Unit
Automatic diff as part of commit; unit tests not applicable.
vkrause created this revision.Aug 20 2017, 5:41 PM
Restricted Application added a project: KDE PIM. · View Herald TranscriptAug 20 2017, 5:41 PM
mlaurent added inline comments.
src/misc/gravatarcache.cpp
55

qCWarning(..)

75

qCWarning(..)

vkrause updated this revision to Diff 18457.Aug 20 2017, 6:34 PM

qWarning -> qCWarning

mlaurent accepted this revision.Aug 20 2017, 7:46 PM

After fixing coding style please commit :)
thanks

src/misc/hash.cpp
33

coding style QByteArray &data

(see other method too)

This revision is now accepted and ready to land.Aug 20 2017, 7:46 PM
This revision was automatically updated to reflect the committed changes.