[kactivitymanagerd] [Bug 365610] Assert in plasmashell sometimes when creating an activity

2016-07-16 Thread Ivan Čukić via KDE Bugzilla
https://bugs.kde.org/show_bug.cgi?id=365610

Ivan Čukić  changed:

   What|Removed |Added

 Resolution|--- |FIXED
 Status|CONFIRMED   |RESOLVED
  Latest Commit||http://commits.kde.org/kact
   ||ivities/b81bbb87f69a7daef85
   ||39a8a2c308d18b3450615

--- Comment #3 from Ivan Čukić  ---
Git commit b81bbb87f69a7daef8539a8a2c308d18b3450615 by Ivan Čukić.
Committed on 16/07/2016 at 17:30.
Pushed by ivan into branch 'master'.

Method for inserting into QFlatSet now returns index along with the iterator

Something strange was happening in Plasma where iterator - begin was
sometimes returning an invalid value.

M  +4-7src/lib/activitiesmodel.cpp
M  +12   -15   src/utils/qflatset.h

http://commits.kde.org/kactivities/b81bbb87f69a7daef8539a8a2c308d18b3450615

-- 
You are receiving this mail because:
You are watching all bug changes.

[kactivitymanagerd] [Bug 365610] Assert in plasmashell sometimes when creating an activity

2016-07-16 Thread Ivan Čukić via KDE Bugzilla
https://bugs.kde.org/show_bug.cgi?id=365610

Ivan Čukić  changed:

   What|Removed |Added

 Status|NEEDSINFO   |CONFIRMED
 Resolution|WAITINGFORINFO  |---
 Ever confirmed|0   |1

--- Comment #2 from Ivan Čukić  ---
This was fun to detect - for some reason, this issue only appeared in plasma
and *never* in the test app. Strange but true. Will push the fix in a moment.

-- 
You are receiving this mail because:
You are watching all bug changes.

[kactivitymanagerd] [Bug 365610] Assert in plasmashell sometimes when creating an activity

2016-07-13 Thread Ivan Čukić via KDE Bugzilla
https://bugs.kde.org/show_bug.cgi?id=365610

Ivan Čukić  changed:

   What|Removed |Added

 Status|UNCONFIRMED |NEEDSINFO
 Resolution|--- |WAITINGFORINFO

--- Comment #1 from Ivan Čukić  ---
Do you have the output message of the assert?

Since I have no idea how this is possible since the index (at least from what I
can tell) can not be < 0, I've changed .begin()/.end() to .cbegin()/.cend() to
avoid potential detachment of the QVector, though it should have not be shared
anywhere.

Can you update and retest?

-- 
You are receiving this mail because:
You are watching all bug changes.