In order for GLX support to work, we need to create the X11 window with
the same visual as in the GLX configuration. That requires delaying
the creation of the window.
This is done in four parts:
- The window builder is re-introduced in the winit backend create
internal renderer interface.
- The glcontext helper code for wasm was moved into a separate function
(less indentation).
- The loop over different display preferences was replaced with a
preference to the "window system native GL interface" over EGL,
which should achieve the same as the fix for #2162. Upstream glutin
defaults to this and so do various downstream projects.
- Use glutin-winit for the actual display selection.
This covers the FemtoVG bits of #2269
We cast the result of the body of functions or binding expressions to
the target type via `as _`. When the code contains a return expression,
that works fine, because `return`'s type is the never type, which can be
coerced to a value of any type. However when the return statement is
inside a sub-block, the type becomes `()`, for which the `as _` cast
fails.
Work around this by attempting to detect the situation (return produces
Type::Invalid) and omit the trailing cast.
The break_text algorithm was O(N²) in the number of lines because
break_text needs to layout the whole text dspite it would in theory
only need the first line.
Trim the text we pass to break_text so we don't need to re-layout the
end for the text for every call.
This is still O(N²) if there are very long lines.
Not sure we can fix that with the current femtovg break_text api
This makes TextEdit with many lines much faster. Still a bit slow, but
at least usable
Otherwise, the cache is trashed with already moderate text size causing
TextEdit to become quickly very slow (unusable) when the text becomes a
bit big.
Big text are horribly still slow but at least now it can be used without
freezing the app for minutes
- This allows creating the Skia GL interface without string allocations
- Port the item renderer to use SkImageFilters::Shader instead of
SkImageFilter::Paint
Fix the code to always return a direct child widget or nullptr. The
documentation says to return a direct or indirect child, but the
implementation expects a direct child.
Fixes: #2195
This disabled about 70 clippy warnings like this one:
```
warning: redundant clone
--> target\debug\build\energy-monitor-7b0492feb1633656\out\desktop_window.rs:44279:37
|
44279 | (if (((r#state . clone ()) . r#previous_state as f64) == (1f64 as f64)) {
| ^^^^^^^^^^^ help: remove this
|
note: cloned value is neither consumed nor mutated
--> target\debug\build\energy-monitor-7b0492feb1633656\out\desktop_window.rs:44279:28
|
44279 | (if (((r#state . clone ()) . r#previous_state as f64) == (1f64 as f64)) {
| ^^^^^^^^^^^^^^^^^^^^
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#redundant_clone
```
I get the following clippy errors in the energy monitor:
error: equal expressions as operands to `-`
--> C:\src\slint\target\debug\build\energy-monitor-7b0492feb1633656\out\desktop_window.rs:47254:94
|
47254 | ... (slint :: private_unstable_api :: re_exports :: LogicalLength :: new (((((0f64 as f64) - (0f64 as f64)) as f64) * (((1f64 as f64) *...
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#eq_op
error: equal expressions as operands to `==`
--> C:\src\slint\target\debug\build\energy-monitor-7b0492feb1633656\out\desktop_window.rs:45382:18
|
45382 | (((0f64 as f64) == (0f64 as f64))) as bool }
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#eq_op
I am not surprised that the generated code sometimes produces code like
this (even though the other demos do not!), so I went ahead and silenced
these errors.
When scrolling the printer demo, I often end up closing or opening
element i don't want to because we reached the end and therefore it is
not scrolling and forwarding the event
We should only forward keys that are a single character. Other ones are
just virtual keys that should't be processed.
This regressed in commit 32d2ba7Fix#2327
... as well as opacity
When an item change its geometry or clipping, one really must also
refresh all the children, even the ones outside of the item.
And if the position of the parent changes, the old offset must still be
used to compute the geometry
This change also add the clip to the dirty region so we don't need to
draw so much for clipped items
This would cause some of the image to be drawn outside of the clip
region, which may even lead to panics
(Fix changing pages in the energy-monitor demo)
This will be needed for a future experiment. Unlike the Skia renderer,
which operates on raw window handles, the FemtoVG renderer exposes a
different interface where it assumes that the caller takes care of the
OpenGL context state. This means more boilerplate remains in the winit
backend, including the glutin dependency. The upside is that it will
allow using the FemtoVG renderer in environments without glutin.
In order to work in an environment without fontconfig or memmap, the
crate has two features:
- fontconfig (set when we anticipate fontconfig to be available at
run-time and libloading being available at compile time).
- diskfonts (set when we want to be able to load fonts from disk)
The winit crate enables fontconfig on "Linux" and diskfonts on !wasm.
This is reproducible in the energy monitor when scrolling down in the
settings page.
In the helper function to colorize an image, we did not expect
brush_to_paint() to fail. The condition of a transparent colorize color
was handled earlier, but it turns out that brush_to_paint() also checked
the global state as an optimization.
This optimization is wrong, as the cached colorized image should not
depend on the untracked global transparency.
So this patch adds a graceful handling of brush_to_paint returning None
for defensive measures, but also moves the early return on an global
transparent alpha to the call sites where it makes sense.
Don't try to install a callback handler that captures an invalid
expression. The llr pass has an invalid expression guard, and so does
the rest of this handle_property_bindings_init callback.
Fixes#2319
States are never constant because the generated state binding depend on
the time as it register the time it was set for the pottential animation
Also adjust the assert so it assert correctly which the right message
Fixes: #2274
The logic was inverted: Setting to true should embed everything, while
false should be the same as if it's not set: Only embed built-in
resources (as always).
