The Fade effect fades unminimized windows. It should not do that because
it alters behavior of effects that animate the uniminimizing of windows,
e.g. the Magic lamp effect.
Details
Details
- Reviewers
- None
- Group Reviewers
KWin
(a) Set animation time factor to 20x
(b) Enabled the Magic lamp effect
(c) Unminimized the System Settings window (it no longer fades in)
Diff Detail
Diff Detail
- Repository
- R108 KWin
- Branch
- effects-fade-dont-fadein-unminimized-windows
- Lint
No Linters Available - Unit
No Unit Test Coverage - Build Status
Buildable 1170 Build 1183: arc lint + arc unit
Comment Actions
Also, it matters whether a window has been grabbed previously(for example, by the Glide effect), but we shouldn't rely on that.
| effects/fade/package/contents/code/main.js | ||
|---|---|---|
| 97 | One could have fade but no other minimise effects. Can we not test for a WindowUnminimizedGrabRole ? | |
| effects/fade/package/contents/code/main.js | ||
|---|---|---|
| 97 | Does it have sense to introduce grab roles for minimization effects? We allow to have only one minimization effect. Yeah, I believe we could, but I think it would be better to have a separate effect for this(the one that fades minimzed windows) because now it's not clear what the Fade effect is doing. | |
Comment Actions
Also, if we are about to allow fading of unminimized windows, we should delete !w.minimized check. Because otherwise behavior is not "symmetrical".
Comment Actions
+1 on the design intent.
but I'm not convinced on the code
IMHO 451bbb54ddf864f80142f36a9d5d54f57e434f1c which introduced this is wrong and it is wrong because ShellClient is wrong. unmapping a window should be considered "deleting" an effectWindow. From an effect POV whether the client is re-using the shellsurface object (and since a Qt change years ago no-one actually does) or not shouldn't behave differently - and now we're on the wrong path.
If I don't get round to it for 5.15 feel free to merge this.
Comment Actions
FYI, I had a go at a lower level fix
branch davidedmundson/waylandWindowClosedOnUnamp Needs a tiny bit of work still.
which would then mean we could simply not handle windowHide here.