- 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().
Before this commit, the pointer focus is cleared when a menu is closed
by clicking its border. This is because get_cursor_context() returns
type=LAB_SSD_NONE when the cursor is on the menu border and
cursor_update_common() clears the pointer focus. This commit fixes this
by replacing cursor_update_common() with cursor_update_focus(), which
calls get_cursor_context() again after the menu scene-node is hidden.
This is reland of 2ade6a1e which was reverted with 21bd5b00, but this
prevents the x-position of the menu from being negative and slides menus
opened with Alt-Space too.
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.
This is a compile-time check, so it will be too conservative if somebody
updates wlroots after building labwc, but that is still better than the
alternatives.
...to avoid a bug that is present in wlroots <0.18.2 and which has the
potential to crash the compositor when performing a drag-and-drop action
at the same time as the XWayland server is shutting down.
This will be reverted when wlroots-0.18.2 can be linked with.
Related-to: #2371
...when options are specified as elements leading to hitting the assert()
in xstrdup().
Reproduce by using this rc.xml:
```
<?xml version="1.0"?>
<labwc_config>
<touch>
<deviceName>foo</deviceName>
<mapToOutput>bar</mapToOutput>
</touch>
</labwc_config>
```
It is likely that <touch> was only tested with options as attributes.
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.
With 2ade6a1, negative x-values can be passed to menu_configure() when the
window menu is opened from top-left button in a maximized window, but
wlr_output_layout_output_at() couldn't find the output for the coordinate
since it's out of the output layout, thus it just opened the menu without
updating its position.
As the next release is close and the visual improvement by 2ade6a1 is not
very impactful for UX, let's revert it for now.
This commit slides menus opened from buttons (with atCursor="no") to the
left by menu.border.width. This makes the location of the window menu
look more natural, especially when it's opened from the window icon in
the left corner of the titlebar.
In the longer term we want this to be user-configurable (and also depend
on the sandbox engine, app_id and instance).
But dropping privileged protocols for sandboxed clients in the meantime
seems like a sensible thing to do and matches user expectations.
Related: #2392
set_squared_corners(false) was always called when titlebar is created.
However, set_squared_corners(false) sets the width of the titlebar
background buffer to (view width) - (corner radius), which causes pixman
errors due to the negative width set for titlebar background buffer when
the view is so small.
...of modifiers which are part of keybinds. This supports clients (for
example blender) that want to see modifier-release-events even if they are
part of a keybinds.
Most clients (including those using Qt and GTK) are setup to not see these
modifier release events and actually misbehave if they do. For example
Firefox shows the menu bar if alt is pressed and then released, whereas if
only pressed (because the release is absorbed) nothing happens. So, if
Firefox saw bound modifier-release-events it would show the menu bar every
time the window-switcher is used with alt-tab.
<windowRules>
<windowRule identifier="blender" wantAbsorbedModifierReleaseEvents="yes"/>
</windowRules>
Issue #1507
lab_data_buffer->logical_{width,height} are not the actual size of the
backing buffer, so wlr_buffer.{width,height} should be used when
duplicating it in get_cairo_surface_from_lab_data_buffer().
This mistake caused segfault when an icon file 1~2 times larger than
window.button.{width,height} is loaded, because the buffer for the raw
icon file is reused by setting lab_data_buffer->logical_{width,height}
with the size for display which is different from
wlr_buffer.{width,height} (see buffer_convert_cairo_surface_for_icon()),
and it is duplicated with get_cairo_surface_from_lab_data_buffer() for
hovered/rounded variants of titlebar button.
It was a cruft from my experiment of adding borders around titles.
It even caused an integer overflow because theme->menu_border_width can be
INT_MIN when menu_header_height is calculated.
This makes the colors of titlebar and window borders different, but will
let menu.border.color (which will be supported soon) inherit
window.active.border.color just like Openbox does, without making the menu
borders around a selected menu item invisible.
scaled_rect_buffer is an implementation of scaled_scene_buffer and shows
an auto-scaling bordered rectangle. This is intended for menu borders,
but can be also useful for other elements like window switcher items.
We will support rounded corners for scaled_rect_buffer in the future.
When the output is in the process of being enabled, but the new state
has not been committed yet, wlr_output->enabled is still false. So it's
not safe to assert that it's true in output_enable_adaptive_sync().
Fixes: 36e099fc93
"view: ensure output is usable before setting adaptive sync"
Implementers can define impl->equal() which compares two
scaled-scene-buffers so that buffers are not allocated for visually
duplicated scaled-scene-buffers.
Currently this mechanism isn't applied for scaled-font-buffers since we
haven't defined impl->equal() for it.
This allows implementations of the interface to share a single
wlr_buffer for multiple wlr_scene_buffer nodes and still make
the scaled-scene-buffer handle the buffer drops.
wlr_scene_buffer holds an implicit lock while using a wlr_buffer
and scaled-scene-buffer.c holds additional locks for as long as
a buffer is within its internal cache.
This should ensure that a shared wlr_buffer never gets actually
dropped even if wlr_buffer_drop() has been called by another
scaled-scene-buffer instance using the same wlr_buffer.
