Review Request 112727: Shrink (fancy) unhide trigger when entered while FS window is active

[Thomas Lübking]

> On Sept. 24, 2013, 10:31 a.m., Martin Gräßlin wrote:
> > > no idea whether that's acceptable
> > shouldn't be much of an issue. KWindowSystem is emitting a signal, isn't it? activeWindowChanged()
> 
> Thomas Lübking wrote:
>     There's been concerns about waking up when windows were moved and KWindowSystem has no disconnectNotify implementation, so once you hooked onto KWindowSystem signals, you're afair not easily getting rid of it's X11 event filtering.
> 
> Christoph Feck wrote:
>     Thomas, bug 217560 comment #19 has a simpler patch, but we are unsure which route to go. Could you have a look?

It appears to me that patch would only apply when the dock is autohidden (and the trigger thus re/created) while there's an active fullscreen window, while that should not (or rather hardly) be possible, since ::hintOrUnhide() exits early when there's a fullscreen window.
Ie. you'd need the panel visible, then get an active fullscreen window, then have the panel hidden. If the active fs window enters while the panel is hidden, the patch would not have any effect?!?

-> Has anybody checked whether that patch actually works?

Iff it does, the implications should be
- the panel is shown/hidden at least once while there's a fullscreen window
- if the panel by this event claims the focus, this will break the "active fullscreen window" condition (not sure, but i think some plasmoid with a lineedit does?)

All approaches might for the latter also have touble with F(S)UM focus policy - not sure whether that's an itch to scratch, but I also doubt that I tested it.


- Thomas


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/112727/#review40664
-----------------------------------------------------------


On Sept. 14, 2013, 2:36 p.m., Thomas Lübking wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/112727/
> -----------------------------------------------------------
> 
> (Updated Sept. 14, 2013, 2:36 p.m.)
> 
> 
> Review request for Plasma, Aaron J. Seigo and Martin Gräßlin.
> 
> 
> Bugs: 217560
>     http://bugs.kde.org/show_bug.cgi?id=217560
> 
> 
> Repository: kde-workspace
> 
> 
> Description
> -------
> 
> Drawback: the next trigger will go unfancied (would require listening to active window changes, no idea whether that's acceptable) but that's still much better than occluding 30 outer px of a window (see bug and http://forum.kde.org/viewtopic.php?f=111&t=112163)
> 
> Notice that
> a) nor auto unhiding neither hinting happens at all while there's a(n active) fullscreen window - this is unchanged (see early exit in patched branch)
> b) the patch does nothing if you don't attempt to enter the occluded area of a fullscreen window (so not even the next hinting will be affected)
> 
> 
> Diffs
> -----
> 
>   plasma/desktop/shell/panelview.cpp dcd051a 
> 
> Diff: https://git.reviewboard.kde.org/r/112727/diff/
> 
> 
> Testing
> -------
> 
> m_unhideTrigger shrinks and the next unhide is not indicated (but the panel appears), subsequent indication remains unaffected.
> 
> 
> Thanks,
> 
> Thomas Lübking
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/plasma-devel/attachments/20140209/7623c15f/attachment-0001.html>