This doesn't change much, since the Smithay implementation is based on
the `cosmic-comp` version, but made more generic. We provide our own
implementation for our workspace capture protocol, but otherwise Smithay
handles the boilerplate now.
This should not cause any change in behavior.
Replacement for https://github.com/pop-os/cosmic-comp/pull/1687, that
works correctly with multiple outputs.
We don't want another window to show a focus indicator while a window is
being dragged, so keep the window in the focus stack. If a window is
being moved out of a stack, change the focus from the stack to the
window.
`refresh_focus_stack()` doesn't seem to be called here, but for good
measure, make sure that calling that function also won't remove a
`CosmicMapped` from the focus stack if it is currently part of a move
grab for the seat.
`Action::Close` already used the keyboard focus target, but some other
bindings didn't. Presumably it's most intuitive if all "current window"
key bindings affect the window with keyboard focus.
These used the focus stack on the `focused_output()` (the one with
keyboard focus), so I guess the main impact is when the keyboard target
is a window being dragged? Then the binding will operate on that window,
or have no effect.
This seems related to some of the behaviors discussed in
https://github.com/pop-os/cosmic-comp/issues/453.
This seems like the correct way to use an `IndexSet`. It shouldn't be
possible to have multiple entries that match, since it's a "set".
We can't define `Borrow<CosmicMapped> for FocusTarget`, so the blanket
impl of `indexmap::Equivalent` won't work, but implementing seems fine.
Tested on `SDL_VIDEO_MINIMIZE_ON_FOCUS_LOSS=1 supertux2`, using
XWayland. If the game starts with fullscreen configured, it minimizes
immediately on start. But after this change, this no longer happens.
This hopefully will improve behaviors with various X11/wine fullscreen
clients. It may fix
https://github.com/pop-os/cosmic-comp/issues/1510, though looking at the
video the behavior there may be a bit different.
I hoped to split this up into multiple commits, but the api
changes to `shell/workspace.rs` were to invasive to feasibly do this.
Here is a rough list of changes:
- Fullscreen windows aren't mapped to other layers anymore
- This they need their own logic for:
- Sending frames
- Dmabuf Feedback
- Primary outputs
- On commit handlers
- cursor tests
- They get their own unmap/remap logic
- They get a new restore state similar to minimized windows
- Refactored the minimized window state to reuse as much as possible
here
- They need to be part of focus stacks, which means adjusting them
to a new type `FocusTarget` as they previously only handled
`CosmicMapped`.
- Various shell handlers (minimize, move, menu) now have dedicated
logic for fullscreen surfaces
- This was partially necessary due to relying on CosmicSurface now,
partially because they should've had their own logic from the
start. E.g. the context menu is now reflecting the fullscreen
state
- Fullscreen windows may be rendered behind other windows now, when they
loose focus.
- This needed changes to input handling / rendering
A solution for https://github.com/Smithay/smithay/issues/1714.
With this, the lock screen is able to get keyboard focus normally, but
focus then reverts to the XWayland grab surface. This can be tested with
the example client from the issue.
Adding a new variant of `KeyboardFocusTarget` is annoying. Maybe it
could map to a different variant of the enum. But it presumably needs to
handle any `wl_surface` XWayland uses. (Override redirect surfaces?
Subsurfaces?) This seems as good as anything for now.
Previously removing the last output could have left seats with an
invalid active output. We already have logic to check this in
`refresh_focus` but failed to apply it before `update_pointer_focus`.
Let's fix that.
If the surface under the pointer has changed, we should send an `enter`
event immediately, instead of waiting for the next motion event. This
seems to fix that, without producing unnecessary events.
Instead of `time: 0`, this and other synthesized events should probably
use `CLOCK_MONOTONIC`. It seems libinput does document that it uses
that.
