+++ This bug was initially created as a clone of Bug #278697 +++ Version: SVN (using Devel) OS: Linux Atm. I have to screens, the setup is as following: Screen 0 | Screen 1 When I move an application that it sticks to the left screen edge (e.g. it is either fullscreen, halfscreen or one fourth of the screen) of screen 1 its shadows are displayed on screen 0 which is to the left of screen 1. Reproducible: Always Steps to Reproduce: 1. Two screens (left/right) 2. Move application to the left edge of the right screen 3. Make sure that it is automatically resized to either the half or one fourth of the screen Expected Results: In that specific case no shadows should be displayed on the other screen.
@Hugo: I split the bug into two - one for Oxygen and one for Aurorae. I want to try doing a fix in Aurorae some day, so I give you this one in case you want to do something in Oxygen.
@Martin, I'm not actually sure that I consider this as a bug. If I move a window (with no quick maximization involved, on the edge of the screen, its shadow expends on the next screen. That's the "normal" behavior. Right ? Now I'm not convinced that quick-tiling, (not fullscreen) should get a different treatment. After all, the shadows is also still displayed on the other edges. Why not this specific edge ? Also: 1/ How can I fix that in oxygen ? right now, I don't know how to detect quick tiling, other than full maximize). 2/ this would also require fix on the other edges (including top and bottom) right ? (in case of fancy screen settings). right ? And including the dealing with the plasma panel ? (should the shadow also be removed below plasma panel ?) My personal feeling would be to close as "wont fix". To me QuickTiled window are just "normal" window with special size and dimensions (as opposed to maximized windows), at least that's how they are deal with inside oxygen. And thus the shadow should be handled accordingly. ...
Maybe it is possible to have a more generic solution (ie. for all windows rather than just for quick tiling windows) that shadows never leak to another screen at all.
@Kai but I do want it to leak ! :-) it will (and should) leak anyway as soon as I have one pixel of the window on other screen, no ? if you consider two screens as a contiguous space (which is the case, I think, as long as you can put a window across these screens), then I do not understand why it should apply on window contents but not on their shadows. Comments ? What do I miss ?
I was actually thinking of what Kai suggests. And that's what I try to achieve. Why? Because I have never window overlapping screen borders. The overlapping shadow is distracting on the windows. Looking at my current setup I seem to workaround it by aligning the windows in a way that the shadows do not overlap. Also there is the KWin rendering perspective (which of course influences my thinking): having a window on screen A not cause repaints on screen B can be quite an improvement.
@Martin isn't the easiest solution to make kwin pass the right mask (QRegion) to the client paint event called ? Mask can be set (by kwin) using screen rect, client rect, and its OuterPadding settings. This way there'd be no duplicated code, and I wouldn't need any change inside oxygen ... (though I'm still not sure that I would like the result, to say the least)
I wouldn't want to do it in KWin core as I'm afraid of bug results due to cut-off shadows. Let's assume we would give enough information to the decos that they know that a window is screen edge aligned, I could imagine that one would want to have custom corners, too, or maybe no border at all at that side. Just thinking aloud what makes me think that manipulating in KWin core is not the best idea.
Thank you for the bug report. As this report hasn't seen any changes in 5 years or more, we ask if you can please confirm that the issue still persists. If this bug is no longer persisting or relevant please change the status to resolved.