Review Request 127216: [KStatusNotifierItem] MinimizeRestore does not "run" over the desktop on X11

[Thomas Lübking]

> On March 3, 2016, 10:16 p.m., Thomas Lübking wrote:
> > src/kstatusnotifieritem.cpp, line 934
> > <https://git.reviewboard.kde.org/r/127216/diff/2/?file=446044#file446044line934>
> >
> >     append
> >     associatedWidget->setAttribute(Qt::WA_Moved);
> >     
> >     
> >     
> >     This should tell Qt to demand an explicit position and skip placement by the WM (yes, this generally works in KWin ;-)
> >     
> >     However, this flag should be implicitly set by ::move() unless the point is invalid, so you might want to "qDebug() << associatedWidgetPos;" to check whether it's invalid for the failing clients.
> 
> Anthony Fieroni wrote:
>     Fun facts :) void KStatusNotifierItemPrivate::minimizeRestore(bool) is called only when you click on tray icon i.e. if i close app with big red X it's not called => i can't store position :)
> 
> Anthony Fieroni wrote:
>     More fun facts :) associatedWidget->setGeometry(associatedWidget->frameGeometry()); works in any case ! I will add path if you want, but you will not be happy to set geometry by myself :)
> 
> Anthony Fieroni wrote:
>     ^ No, extend geometry with decoration size :)
> 
> Thomas Lübking wrote:
>     meehhh...
>     
>     That means one will have to filter QEvent::Close of associatedWidget and store the position from there.
> 
> Anthony Fieroni wrote:
>     It's strange, about me, associatedWidget has correct frameGeometry even it's hide.
>     QObject::connect(KWindowSystem::self(), &KWindowSystem::windowRemoved, q, [this](WId winId) {
>             if (associatedWidget && associatedWidget->winId() == winId) {
>                 associatedWidgetPos = associatedWidget->frameGeometry().topLeft();
>             }
>         });
>     Position is correct, but again not work event with Qt::WA_Moved.
> 
> Anthony Fieroni wrote:
>     KWindowInfo info(associatedWidget->winId(), NET::WMDesktop | NET::WMGeometry);
>     info.geometry().topLeft(); <----------------------------- decoration is included again
>     to filed bug?

You may file a bug, but I frankly doubt your findings.
(Just more or less ported my kwindowsystem tool and the values are correct ;-)

=> How exactly did you determine this assumption? (Just from the walking position?)


- Thomas


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/127216/#review93143
-----------------------------------------------------------


On Feb. 29, 2016, 5:18 a.m., Anthony Fieroni wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/127216/
> -----------------------------------------------------------
> 
> (Updated Feb. 29, 2016, 5:18 a.m.)
> 
> 
> Review request for KDE Frameworks, Martin Gräßlin, Thomas Lübking, and Martin Klapetek.
> 
> 
> Bugs: 356523
>     https://bugs.kde.org/show_bug.cgi?id=356523
> 
> 
> Repository: knotifications
> 
> 
> Description
> -------
> 
> Store position of widget before hide it
> 
> 
> Diffs
> -----
> 
>   src/kstatusnotifieritem.cpp cf3e7b5 
>   src/kstatusnotifieritemprivate_p.h 8fdfd4c 
> 
> Diff: https://git.reviewboard.kde.org/r/127216/diff/
> 
> 
> Testing
> -------
> 
> Tested on pixel ration = 1
> 
> 
> Thanks,
> 
> Anthony Fieroni
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-frameworks-devel/attachments/20160305/32a1c0b1/attachment.html>