Improve multi-monitor handling under wayland #1276
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
I'm down to merge this once someone has tested if the behavior breaks anything under X11, once the conficts are resolved ^^ |
When a monitor gets disconnected, the destroy event of all associated windows gets called, and the window gets removed. This patch changes that behavior: the window is still closed but the configuration is kept using the existing reload mechanism. In addition, a callback is added to listen for new monitors, triggering a reload when a new monitor gets connected. This logic also reloads already running windows, which has a positive and negative effect: - positive: if currently running e.g. on the second monitor specified in the list, the window can get moved to the first monitor - negative: if reloading starts it on the same monitor, it gets reset (e.g. graphs) I also had to work around an issue: the monitor model is not yet available immediately when a new monitor callback triggers. Waiting in the callback does not help (I tried 10 seconds). However, waiting outside, it always became available after 10ms. Tested with a dual monitor setup under KDE through a combinations of: - enabling/disabling individual monitors - switching between monitors - specifying a specific monitor in the yuck config - specifying a list of specific monitors in the yuck config In all these cases the behavior is as expected, and the widget gets loaded on the first available monitor (or stays unloaded until one becomes available). It also works when opening a window without any of the configured monitors being available. There is one remaining error from GTK when closing the window: GLib-GObject-CRITICAL **: 20:06:05.912: ../gobject/gsignal.c:2684: instance '0x55a4ab4be2d0' has no handler with id '136' This comes from the `self.gtk_window.disconnect(handler_id)` call. To prevent that we'd have to reset `destroy_event_handler_id`.
bkueng
force-pushed
the
improve_multi_monitor_wayland
branch
from
39b18a9 to
9aec9e9
Compare
In case 'dirty' was set, we could end up with 'reuse_window = false', but not closing the existing window. This moves the closing check to after 'if reuse_window', which allows to simplify the logic. Fixes regression from 29fa158
bkueng
force-pushed
the
improve_multi_monitor_wayland
branch
from
9aec9e9 to
6b0f42e
Compare
|
I still have an old system with X11 installed, and was able to test there. It works as well with 2 differences in behavior:
I also rebased and resolved the conflicts. There was a regression in a recently added commit (29fa158) which I fixed in 6b0f42e. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
When a monitor gets disconnected, the destroy event of all associated windows gets called, and the window gets removed.
This PR changes that behavior: the window is still closed but the configuration is kept using the existing reload mechanism. In addition, a callback is added to listen for new monitors, triggering a reload when a new monitor gets connected.
This logic also reloads already running windows, which has a positive and negative effect:
I also had to work around an issue: the monitor model is not yet available immediately when a new monitor callback triggers. Waiting in the callback does not help (I tried 10 seconds). However, waiting outside, it always became available after 10ms.
Tested with a dual monitor setup under KDE through a combinations of:
In all these cases the behavior is as expected, and the widget gets loaded on the first available monitor (or stays unloaded until one becomes available).
It also works when opening a window without any of the configured monitors being available.
There is one remaining error from GTK when closing the window:
GLib-GObject-CRITICAL **: 20:06:05.912: ../gobject/gsignal.c:2684: instance '0x55a4ab4be2d0' has no handler with id '136'.This comes from the
self.gtk_window.disconnect(handler_id)call. To prevent that we'd have to resetdestroy_event_handler_id.I have not tested under X11, and I'm not sure if the behavior should be the same there.
Fixes #1158
Checklist
Please make sure you can check all the boxes that apply to this PR.
docs/content/maindirectory has been adjusted to reflect my changes.cargo fmtto automatically format all code before committing