This prevents potential bugs that buffers created by scaled_scene_buffers
before Reconfigure are reused by other newly created scaled_scene_buffers.
This is targeted for scaled_icon_buffer whose buffer creation depends on
server->sfdo.
And make mousebind handlers use that one.
Also remove keyboard_any_modifiers_pressed() and replace its usage
with the new function.
Without this patch we would only request the modifier state of the
keyboard group which makes mousebinds involving keyboard modifiers
break for virtual keyboards like when using wayvnc. Same story for
hiding the workspace overlay or snapping to regions.
Fixes: #2511
This fixes an issue with buffers not updating when an output is configured
for a new scale. It also supports windows being on more than 2 outputs at
once and in general simplifies the code.
This patch splits desktop_entry_icon_lookup() into two separate functions
- desktop_entry_load_icon(): load a icon from the configured icon theme
- desktop_entry_load_icon_from_app_id(): load a icon name from a .desktop
file based on the given app_id and supply it to _load_icon().
The _load_icon() function will be used in a future menu icon
implementation whereas the _load_icon_from_app_id() function is used
within the SSD titlebar window icon lookup routine.
683f67b7 introduced another regression that the modifier state (Ctrl) is
stuck when Ctrl+F is pressed in some applications like Firefox while
Fcitx5 is running. This caused mouse scrolls to zoom in/out the UI.
Let me explain the cause in detail. When Ctrl+F is pressed, an input box
is opened in the application and Fcitx5 creates a new virtual keyboard
(VK), whose initial modifiers is empty. Then prior to 683f67b7, the
key/modifiers events flowed like this:
- The compositor detects F key-release
- Modifiers (Ctrl pressed) are notified via _set_keyboard()
- F key-release is forwarded to IM
- IM sends modifiers (Ctrl) back to the compositor via VK
- **The modifiers on VK is updated (empty->Ctrl)**
- **Modifers (Ctrl) are notified to the app**
- IM sends F key-release back to the compositor via VK
- F key-release is notified to the app
- The compositor detects Ctrl key-release
- Ctrl key-release is forwarded to IM
- Modifiers (empty) are forwarded to IM
- IM sends Ctrl key-release back to the compsitor via VK
- Ctrl key-release is notified to IM
- IM sends modifiers (empty) back to the compositor via VK
- **The modifiers on VK is updated again (Ctrl->empty)**
- **Modifiers (empty) are notified to the app**
Thus, the final modifiers (empty) is notified to the application as
expected. However, after 683f67b7, the key/modifiers events flowed like
this:
- The compositor detects F key-release
- F key-release is directly notified to the app
- The modifiers (Ctrl) is also notified to the app
- The compositor detects Ctrl key-release
- Ctrl key-release is directly notified to the app
- Modifiers (empty) are forwarded to IM
- IM sends modifiers (empty) back to the compositor via VK
- **Modifier on VK is not updated (empty->empty)**
- **The compositor ignores it**
So the final modifier (empty) is never notified to the application, which
causes stuck Ctrl modifier.
This commit fixes this by not forwarding the modifiers when it hasn't been
updated since it was forwarded previously. So after this commit, the
key/modifiers events flow like this:
- The compositor detects F key-release
- F key-release is directly notified to the app
- The modifiers (Ctrl) is also notified to the app
- The compositor detects Ctrl key-release
- Ctrl key-release is directly notified to the app
- The modifiers are directly notified to the app because the modifiers
(empty) are the same as the last forwarded modifier (empty).
as well as 'openbox-3', for example: /usr/share/themes/foo/labwc/themerc
Give 'labwc' higher precedence than 'openbox-3' if both exist.
For several reasons not all themes designed for labwc render well on
Openbox:
1. Labwc support some features like SVG/PNG icons and #rrggbbaa color
definitions which Openbox does not.
2. Openbox generally defaults to black/white colors and will not render
themes accurate without certain options specified in themerc. For
example, in Openbox the following will just render as black:
window.active.title.bg.color: #589bda
...because Openbox defaults the texture to "Gradient Vertical" and
therefore either needs:
window.active.title.bg.colorTo: #3c7cb7
...or remove 'Gradient' from the bg definition and just use something
like:
window.active.title.bg: Flat Border
Whilst none of this is a problem when using labwc, it causes problems for
setups with both Openbox and labwc because themes designed for labwc only
will now show up in obconf, lxappearance, etc causing confusion for users.
16dbdc64 changed the padding around the app icon in the titlebar to be
applied both vertically and horizontally rather than only horizontally
because it was more natural from a developer's perspective, but some users
complained about the smaller icons in certain configurations.
So let's undo the change in 16dbdc64 and apply the icon padding only
horizontally for now.
We can add configurations for the icon padding (or icon size independent
from window.button.{width,height}?) later.
Before this commit, there was a use-after-free bug on Reconfigure:
- theme_finish() destroys lab_imgs for titlebar icons
- For some reason, undecorate() calls _create_buffer() in
scaled-img-buffer.c, which calls img_render() on a destroyed lab_img.
So in this commit, the lifetime of lab_img is expanded to when the
scaled_img_buffers referencing it are all destroyed. This is achieved by
calling lab_img_copy() when setting a lab_img to scaled_img_buffer and
calling lab_img_destroy() when clearing a lab_img.
Now that scaled_img_buffer.img are always different, lab_img_equal() is
added to compare the content of scaled_img_buffer.img.
lab_img.theme member was referenced by "modifier" functions when drawing a
hover effect on button buffers or rounding the button buffer for corner
buttons, but this can be prone to use-after-free and was not very clean
because theme and lab_img were referencing each other.
Instead, let's just reference rc.theme in the modifier functions and
remove theme from lab_img.
Before this commit, keystrokes were interpreted based on following
hard-coded rules while the window switcher is active:
1. Up/Left arrow keys cycle the window forward.
2. Down/Right arrow keys cycle the window backward.
3. Other keystrokes cycle the window in the initial direction specified
by NextWindow/PreviousWindow actions. But while Shift key is pressed,
the direction is inverted.
...and keybind actions were never executed.
However, this lead to a counter-intuitive behavior for new, especially
pre-Openbox users. For example, in the following keybinds, after the user
activates the window switcher with Super+n, Super+p cycles the window
_forward_:
<keybind key="W-n">
<action name="NextWindow" />
</keybind>
<keybind key="W-p">
<action name="PreviousWindow" />
</keybind>
This is because the key 'n' is recognized just as a normal key in the
third hard-coded rule.
So this commit changes the rules to be more Openbox-like:
1. Up/Left arrow keys cycles the window forward.
2. Down/Right arrow keys cycles the window backward.
3. Other keystrokes are matched against keybinds and execute their
actions. If they include NextWindow/PreviousWindow action, it cycles
the selected window forward/backward even while the window switcher
is active.
Currently, labwc automatically enables outputs at startup and when new
outputs are connected. Make this behavior optional (but still enabled by
default).
With autoEnableOutputs disabled, tools such as kanshi can be used to
give finer-grained control of which outputs are enabled and when.
Currently we may end up in an endless loop of Reconfigure requests
if the Reconfigure action was called by a keybind. If the reconfigure
takes too long (which may happen on slow systems with libsfdo full
debug logging for example) the reconfigure might be triggered again
and again.
To prevent that, simply cancel all keybind_repeat timers on reconfigure.
The previous revert fixed the problem of stuck modifier keys with
keybinds in Blender, but made Firefox show its menu bar with Alt-*
keybinds. This is fundamentally inevitable due to the limitation of
wayland protocol, but at least for the default Alt-Tab keybind for
window switcher, we can mitigate this problem by clearing the keyboard
focus when the window switcher is activated. This is what KWin does, and
we decided to follow that.
So in this commit, keyboard and pointer focus are temporarily cleared
while Move/Resize, window switcher and menu interactions and restored
after them. We slightly deviate from KWin as KWin doesn't clear the
keyboard focus while Move/Resize, but it solves our existing problem
that Firefox shows its menu bar after dragging it with default Alt-Drag
mousebind, and this is what Mutter does.
We considered other solutions, but they don't work well:
1. Send wl_keyboard.{leave,enter} every time keybinds/mousebinds are
triggered. This solves the Firefox's menu bar problem, but that
sounds like a workaround and sending unnecessary events every time is
not desirable.
2. Send release events for both modifiers and keys even when they are
bound to keybinds. This is what Mutter is doing, but it looks like an
implementation issue and violates wayland protocol.
- fix that icons for normal/hovered/rounded buttons are not placed
exactly the same position
- fix blurry window button icons in scaled outputs
This commit introduces lab_img and scaled_img_buffer and uses them for
rendering icons in the window titlebar. Now the process of rendering
button icons are split into 2 phases: loading with lab_img_load() and
creating scene-nodes for them with scaled_img_buffer_create(). This
might incur some additional overhead since we no longer preload icon
textures, but the rendering of icon only happens for the first window
as backing buffers are shared and the overhead won't be noticeable.
This commit also simplifies the process of centering icon buffer in the
button, by creating icon buffers in a fixed geometry via
lab_img_render().
Prior to this commit, a backing buffer with scale 1 was always created for
a scaled_scene_buffer before showing it, and backing buffers for specific
scales were created on output_enter events.
This commit removes this redundant re-renderings by calling
wlr_scene_buffer_set_dest_size() upon scaled_scene_buffer creation just to
receive output_enter events and delaying the first rendering to the first
output_enter event.
I needed to add font_get_buffer_size() to obtain the size of a font buffer
without actually creating it.
After commit e2189903 in wlroots, when ctrl-f is pressed in firefox with
a IME client running, the following key-release event for "f" is not
sent, thus "f" is repeated like "ffffffffff..." in the input box of
firefox. This is because the key-release event for "f" is firstly
forwarded to the IME client and then sent via the virtual keyboard created
by the IME client while the key-press event is sent via physical
keyboard, and with e2189903, key-release events without a corresponding
key-press event on the same keyboard is not emitted to the compositor.
So this commit fixes this problem by not forwarding the key-release event
to the IME client unless the corresponding key-press event was also
forwarded.
This allows to use it for a future ext-workspace implementation.
It is also more generalized so can be used for other protocol
implementation in the future in case the protocols require some
kind of transaction management.
This commit delegates the calculation for menu position into wlroots
utilities for xdg_positioner.
Notable functional changes are:
- Slide the menu to fit in the output when it's opened out of the output
(e.g. top-left window menu is opened when the window is overflowing to
the left), rather than not updating the menu at all.
- The horizontal alignment of menus is now determined based on the size of
each (sub)menu alone rather than the total width of entire menu tree.
This means submenus can now overlap with is parents, but this is no
longer a problem since we recently added support for menu borders.
- Fixed that pipemenus always follow the alignment of its parent even when
it overflows from the output.
Arrow signs are specific to submenu items, so they would be more natural
to be handled in menu.c rather than accepting "arrow" in
font_buffer_create().
Also I allowed non-positive numbers for max_width in font_buffer_create(),
in which case the natural font width is used as the buffer width.
- Don't store font/background buffers in menuitem struct since we no
longer dynamically update buffers of existing menuitems.
- Factor out the duplicated codes for creating menu item scenes for each
unselected/selected states.
Prior to this commit, output-relative coordinate was passed to
menu_open_root() as the menu position when it's designated via "x"/"y"
arguments in ShowMenu action, so menu can be misplaced to the output other
than the one in which the cursor is in.
