> On June 4, 2013, 3:22 p.m., Martin Gräßlin wrote:
> > I think the fix is applied at the wrong place. The actual problem is that
> > window not on the current activity are included in the layouting and so on
> > which results in the empty spots where a window is. So the fix should be to
> > not include such windows in the first place, then it doesn't matter where
> > the user clicks.
>
> Stefanos Harhalakis wrote:
> This would mean changing one of the following:
>
> * Workspace::xStackingOrder()
> * EffectsHandlerImpl::stackingOrder()
>
> I assume you don't mean Workspace::xStackingOrder() so that would be
> EffectsHandlerImpl::stackingOrder(). Right?
>
> In any case, if isOnCurrentActivity() is expensive as Thomas said, would
> it be ok to proactively run this against all available windows instead of
> doing it at the very last moment?
>
> Changing EffectsHandlerImpl for example, would mean this:
>
> foreach (Toplevel *w, list)
> if (!w->isOnCurrentActivity())
> ret.append(effectWindow(w));
>
Correction:
foreach (Toplevel *w, list)
if (w->isOnCurrentActivity())
ret.append(effectWindow(w));
- Stefanos
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://git.reviewboard.kde.org/r/110824/#review33751
-----------------------------------------------------------
On June 4, 2013, 3:02 p.m., Stefanos Harhalakis wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://git.reviewboard.kde.org/r/110824/
> -----------------------------------------------------------
>
> (Updated June 4, 2013, 3:02 p.m.)
>
>
> Review request for kde-workspace.
>
>
> Description
> -------
>
> Fix windowAt function to only return windows from current activity.
>
>
> This addresses bug 301447.
> http://bugs.kde.org/show_bug.cgi?id=301447
>
>
> Diffs
> -----
>
> kwin/effects/desktopgrid/desktopgrid.cpp dc3d82b
>
> Diff: http://git.reviewboard.kde.org/r/110824/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Stefanos Harhalakis
>
>
