* Closes https://github.com/emilk/egui/issues/1918
* Closes https://github.com/emilk/egui/issues/4437
* Closes https://github.com/emilk/egui/issues/4709
* [x] I have followed the instructions in the PR template
Hiya,
I need new winit for a specific fix for a android_native_actvity. There
are already two PRs, but both don't seem to have a lot of movement, or
are entirely complete:
https://github.com/emilk/egui/pull/4466
Seems to have gone stale & is missing some bits.
https://github.com/emilk/egui/pull/4702
Also seems stale (if less so), and is missing a refactor to
run_on_demand. I also *think* the accesskit integration has a mistake
and can't be enabled. I've marked them as a co-author on this as I
started from this branch. (I think! Haven't done that on git before...).
Sorry for the wall of text but just dumping some details / thoughts
here:
- There's an issue with creating child windows in winit 0.30.1 and up on
macOS. The multiple_viewports, "create immediate viewport" example
crashes on anything later 0.30.1, with a stack overflow in unsafe code.
I've create [a winit
issue](https://github.com/rust-windowing/winit/issues/3800), it *might*
already be fixed in 0.31.0 but I can't test as 0.31 will likely require
another refactoring. For now I have just pinned things to 0.30.0 exatly.
- Winit has deprecated run_on_demand, instead requiring the
ApplicationHandler interface. In 0.31.0 run_on_demand is removed. I've
refactored both the integration and the WinitApp trait to follow this
pattern. I've left user_events a bit more opaque, as it seems 0.31.0 is
doing a rework of UserEvents too.
- I've used the new lazy init approach for access kit from this branch
https://github.com/mwcampbell/egui/tree/accesskit-new-lazy-init and
marked Matt as co-author, thanks Matt!
- There was very similair but not quite the same code for run_and_return
and run_and_exit. I've merged them, but looking at the github issues
graveyard it seems vey finnicky. I *hope* this is more robust than
before but it's a bit scary.
- when receiving new_events this also used to check the redraw timing
dictionary. That doesn't seem necesarry so left this out, but that is a
slight behaviour change?
- I have reeneabled serial_windows on macOS. I wondered whether it was
fixed after this PR and does seem to be! However, even before this PR it
seems to work, so maybe winit has sorted things out before that...
Windows also works fine now without the extra hack.
- I've done a very basic test of AccessKit on Windows and screen reader
seems ok but I'm really not knowleadgable enough to say whether it's all
good or not.
- I've tested cargo tests & all examples on Windows & macOS, and ran a
basic Android app. Still, testing native platforms is wel... hard so if
anyone can test linux / iOs / older mac versions / windows 10 would
probably be a good idea!
- For consistencys sake I've made all event like functions in WinitApp
return a `Result<EventResult>`. There's quite a bit of Ok-wrapping now,
maybe too annoying? Not sure.
Thank you for having a look!
# Tested on
* [x] macOS
* [x] Windows
* [x] Wayland (thanks [SiebenCorgie](https://github.com/SiebenCorgie))
* [x] X11 (thanks
[crumblingstatue](https://github.com/crumblingstatue)!,
[SiebenCorgie](https://github.com/SiebenCorgie))
# TODO
* [x] Fix "follow system theme" not working on initial startup (winit
issue, pinning to 0.30.2 for now).
* [x] Fix `request_repaint_after`
---------
Co-authored-by: mwcampbell <mattcampbell@pobox.com>
Co-authored-by: j-axa <josef.axa@gmail.com>
Co-authored-by: DataTriny <datatriny@gmail.com>
Co-authored-by: Emil Ernerfeldt <emil.ernerfeldt@gmail.com>
`directories-next` was created because `directories` was not maintained
at the time. However, `directories` has gotten active maintainership
since, and it has received more updates than `directories-next`.
`directories` also has more recent downloads than its `next`
counterpart, so it might make sense to switch to it, to avoid
unnecessary duplicate dependencies, where a project depends transitively
on both `directories` and `directories-next`.
The main question is whether we depend on any specific behavior from
`directories-next`.
Currently, if the size of the canvas element changes independently of
the size of the browser window (e.g. due to its parent element
shrinking), then no repaints are scheduled.
This PR replaces the `resize` event with a `ResizeObserver`, which
ensures that _any_ resize of the canvas element (including those caused
by browser window resizes) trigger a repaint. The repaint is done
synchronously as part of the resize event, to reduce any potential
flickering.
The result seems to pass the rendering tests on most platform+browser
combinations. We tested:
- Chrome, Firefox, Safari on macOS
- Chrome, Firefox on Linux (ubuntu and arch, both running wayland)
- Chrome, Firefox on Windows
Firefox still has some antialiasing issues on Linux platforms, but this
antialiasing also happens on `master`, so this PR is not a regression
there.
The code setting `canvas.style.width` and `canvas.style.height` at the
start of `AppRunner::logic` was also removed - the canvas _display_ size
is now fully controlled by CSS, e.g. by setting `canvas { width: 100%;
height: 100%; }`.
The approach used here is described in
https://webglfundamentals.org/webgl/lessons/webgl-resizing-the-canvas.html
Note: The only remaining place where egui updates the style of the
canvas it is rendering to is some of the IME/mobile input handling code.
Fixing that is out of scope for this PR, and will be done in a followup
PR.
<!--
Please read the "Making a PR" section of
[`CONTRIBUTING.md`](https://github.com/emilk/egui/blob/master/CONTRIBUTING.md)
before opening a Pull Request!
* Keep your PR:s small and focused.
* The PR title is what ends up in the changelog, so make it descriptive!
* If applicable, add a screenshot or gif.
* If it is a non-trivial addition, consider adding a demo for it to
`egui_demo_lib`, or a new example.
* Do NOT open PR:s from your `master` branch, as that makes it hard for
maintainers to add commits to your PR.
* Remember to run `cargo fmt` and `cargo clippy`.
* Open the PR as a draft until you have self-reviewed it and run
`./scripts/check.sh`.
* When you have addressed a PR comment, mark it as resolved.
Please be patient! I will review your PR, but my time is limited!
-->
Related to #3482
Not sure what the "best practice" is, to me it seems like one should
import from "the original location" if possible, but now it should at
least be possible to not re-export ahash without any breakage in the
egui code base (but possibly in projects using egui, so one should
probably deprecate it if one would like to go that path). It also seems
like epaint re-exports ahash.
<!--
Please read the "Making a PR" section of
[`CONTRIBUTING.md`](https://github.com/emilk/egui/blob/master/CONTRIBUTING.md)
before opening a Pull Request!
* Keep your PR:s small and focused.
* The PR title is what ends up in the changelog, so make it descriptive!
* If applicable, add a screenshot or gif.
* If it is a non-trivial addition, consider adding a demo for it to
`egui_demo_lib`, or a new example.
* Do NOT open PR:s from your `master` branch, as that makes it hard for
maintainers to add commits to your PR.
* Remember to run `cargo fmt` and `cargo clippy`.
* Open the PR as a draft until you have self-reviewed it and run
`./scripts/check.sh`.
* When you have addressed a PR comment, mark it as resolved.
Please be patient! I will review your PR, but my time is limited!
-->
Inspired by:
44d65f41ac/Cargo.toml (L65)
I took the liberty of removing that comment since I *think* that I got
all "relevant" ones (showing up more than once, sort of).
These are a replacement to the `objc` and `cocoa` crates.
This PR prevents:
- An extra copy when creating `NSData`
- A memory leak when creating `NSImage`
- A memory leak when creating `NSString`
And is generally a readability improvement.
Note that we define `NSApp` manually for now, the implementation in
`objc2-app-kit` is currently suboptimal and wouldn't allow you to check
whether the NSApplication has been created or not.
Related: https://github.com/emilk/egui/issues/4219, this should nicely
coincide with the Winit `0.30` release.
---------
Co-authored-by: Emil Ernerfeldt <emil.ernerfeldt@gmail.com>
After merging PR #4036, build errors occurred in eframe-related
applications:
```log
error[E0432]: unresolved import `winapi::um::winuser`
--> crates\eframe\src\native\app_icon.rs:83:9
|
83 | use winapi::um::winuser;
| ^^^^^^^^^^^^^^^^^^^ no `winuser` in `um`
|
note: found an item that was configured out
--> C:\Users\Varphone\.cargo\registry\src\index.crates.io-6f17d22bba15001f\winapi-0.3.9\src\um\mod.rs:290:37
|
290 | #[cfg(feature = "winuser")] pub mod winuser;
| ^^^^^^^
= note: the item is gated behind the `winuser` feature
For more information about this error, try `rustc --explain E0432`.
error: could not compile `eframe` (lib) due to previous error
warning: build failed, waiting for other jobs to finish...
```
* Closes https://github.com/emilk/egui/issues/3941
Workspace dependencies can be annoying.
If you don't set them to `default-features=false`, then you cannot opt
out of their default features anywhere else, and get warnings if you
try.
So you set `default-features=false`, and then you need to manually opt
in to the default features everywhere else.
Or, as in my case, don't.
I don't have the energy to do this tonight, so I'll just revert.
This makes all `wgpu` features opt-in for `egui-wgpu` users.
For `eframe`, I opted to enable some `wgpu` features so users can still
use `eframe` without having to also opt-in to extra ewgpu features
(eframe is batteries-included).
We were using [`tts`](https://github.com/ndarilek/tts-rs) for the
web-only screen reader. This was overkill, to say the least. It is now
replaced with ten lines of `web-sys` calls.
Introduced in the recent multi-viewports work, we accidentally recreated
the wgpu surfaces every frame. This is now fixed.
I found this while improving the profiling of `eframe`
* Silence a few clippy warnings
* Use named threads
* Remove some deprecated functions
* Document Context and Ui fully
* Use `parking_lot::Mutex` in `eframe`
* Expand clippy.toml files
* build fix
* Bump `wgpu` to 0.17.0
This required bumping wasm-bindgen to 0.2.87
* cargo deny exception for `foreign-types`
* sort deny.toml
* Add fragile-send-sync-non-atomic-wasm feature to wgpu
* cargo deny: ignore children of foreign-types
---------
Co-authored-by: Andreas Reich <r_andreas2@web.de>
Co-authored-by: Emil Ernerfeldt <emil.ernerfeldt@gmail.com>
* Expose raw window and display handles in eframe
* Ensure that no one implements `Clone` in the future
* Cleanup
---------
Co-authored-by: Matti Virkkunen <mvirkkunen@gmail.com>
Emil Ernerfeldt
Arthur Brussee
Salman Abuhaimed
crumblingstatue
Jan Procházka
Oscar Gustafsson
Mads Marquart
Cameron Taggart
Varphone Wong
YgorSouza
Fredrik Fornwall
Aaron Hill
τ