`setMarkedText:selectedRange:replacementRange:` converted the IME's UTF-16
`selectedRange` into UTF-8 byte offsets by taking `substringToIndex:` prefixes
and measuring them with `NSString::len()`
(`lengthOfBytesUsingEncoding:NSUTF8StringEncoding`). When an index falls inside
a surrogate pair, the prefix ends in a lone high surrogate, which UTF-8 cannot
represent, so `lengthOfBytesUsingEncoding:` returns 0 for the entire prefix and
the offset silently collapses to 0.
By using `OwnedDisplayHandle`, which uses reference-counting, instead of
`DisplayHandle<'static>` (where the user has to guarantee that the
Application doesn't outlive the event loop).
Any winit app on Windows freezes when the keyboard layout is switched by
tools such as Punto Switcher. A minidump of the frozen process shows the
main thread blocked in MsgWaitForMultipleObjectsEx (winit's event-loop
wait), re-entered during message dispatch through the switcher's injected
global hook (pshook64.dll, via CallNextHookEx) — a re-entrant Win32
message-loop wait, not a mutex deadlock (no LAYOUT_CACHE frames present).
Handle WM_INPUTLANGCHANGE to refresh the cached keyboard layout, then
defer to DefWindowProc so the message still propagates to first-level
child windows as the Win32 docs require. The cache refresh is the minimal
change that stops the freeze (verified on Windows 11 by isolation
variants); update_modifiers and swallowing the message are not needed.
Remove the matrix in the CI action; EmbarkStudios/cargo-deny#324 hasn't been resolved yet, but since we've split Winit out into multiple crates, there's still value in minimizing dependency conflicts even if they won't be hit by users, since e.g. Rust-Analyzer will by default check the entire workspace (and thus download and compile duplicate dependencies).
Most events in AppKit go through `sendEvent:`, and they contain a lot of
information, so it's nice to surface this when debugging.
We could override `sendEvent:` in UIKit and track this in there too, but
that's much less important, since there the relevant events are fairly
narrowly scoped, see the link below, other events go through CFRunLoop.
https://developer.apple.com/documentation/uikit/uievent/eventtype
Add two run loop observers that:
- Create a TRACE-level span when the run loop enters a new state.
- Drops the span when the run loop exits that state.
These spans attach information to events, such that e.g. resizing a view
produces messages like:
```
TRACE inside runloop{mode=NSEventTrackingRunLoopMode}:timers:
winit_appkit::util: Triggered `drawRect:` target="winit_appkit::view"
TRACE inside runloop{mode=NSEventTrackingRunLoopMode}:timers:
winit_appkit::util: Completed `drawRect:` target="winit_appkit::view"
```
The `text` field on `KeyEvent` was hardcoded to `None` on Android,
making it impossible for custom `NativeActivity` subclasses that
show the IME to receive functional text input using *for example* the
existing `winit-egui` crate which relies on this field being set.
Use `Key::to_text()` on press events to derive `text` from
`logical_key`, matching the convention used by the Windows and macOS
backends.
Supposedly that doesn't include all kinds of special virtual unicode
keys, but at least the basics work this way.
macOS native Pinyin IME can send a selected_range that exceeds the
marked text string length (e.g. index 8 for a 6-character string).
This caused an NSRangeException in substringToIndex:, crashing the
application with SIGABRT.
Clamp both location and end to the string's UTF-16 length before
calling substringToIndex.
Added a common interface that:
- Uses closures instead of static functions. This should allow easier
refactoring in the future.
- Returns a handle which is invalidated on `Drop`. This should avoid
situations where the event loop has exited, but an observer is still
called because the user spawned the application later on.
- Is properly main-thread safe.
This interface is placed in winit-common, to allow using it in both
winit-appkit and winit-uikit.
* Deduplicate Pixel impl for integers using a macro
* Deduplicate [Logical|Physical][Size|Position] From impls using a macro
Co-authored-by: Osspial <osspial@gmail.com>
At the moment, the wayland code uses ahash to perform hashing
in its various hash mas. This was done because ahash was seen as
the best default in the Rust community at the time. However, most
Rust crates (including `hashbrown`) have since moved to using
foldhash instead.
This move is done for two primary reasons:
- This reduces the number of dependencies in the tree for most GUI
projects. As other projects use foldhash now, this removes ahash
(as well as its five dependencies) from the tree.
- In most cases, foldhash is faster than ahash.
Signed-off-by: John Nunley <dev@notgull.net>
The xfixes implementation is not that reliable and rather simple to
replace, so use x11rb to implement the same functionality.
Fixes#4120.
Co-authored-by: avitran0 <holyhades64@gmail.com>
This avoids using JavaScript exceptions to support `EventLoop::run_app`
on the web, which is a huge hack, and doesn't work with the Exception
Handling Proposal for WebAssembly:
https://github.com/WebAssembly/exception-handling
This needs the application handler passed to `run_app` to be `'static`,
but that works better on iOS too anyhow (since you can't accidentally
forget to pass in state that then wouldn't be dropped when terminating).
Unify the values of `MouseButton` and thus remove `Other` variant in
limit possible buttons to 32, which was picked based on platform
capabilities, where 32 is the highest.
For the reference, SDL has identical limit.
When pressing Ctrl+Space, KeyEvent::text_with_all_modifiers would return
\0. That would get included in the text-input text. When an input method
gets activated, the invalid string returning \0 would get sent as
surrounding text, resulting in a Wayland protocol error and crashing the
application.
This can be tested with the `application` example, looking at the events
shown for mouse wheel movement.
`wl_pointer::axis_discrete` isn't sent in version 8 or higher of
`wl_pointer`. And `sctk` doesn't convert the `value120` events, so
on compositors advertising version 8, only pixel scroll events were
being sent.
This sends `MouseScrollDelta::LineDelta` with a fractional value,
without doing any accumulation. Given `LineDelta` contains `f32` values,
this presumably is expected?
Though it might be good to change the definition of `MouseScrollDelta`
to include both discrete and pixel values, when the compositor sends
both. I'm not familiar with how this works on non-Wayland backends
though.
The API is integrated into the `WindowEvent::Pointer*` API and is
present in form of `TabletTool` variant on corresponding data entries.
For now implemented for Web, Windows, and with limitations for Wayland.
Fixes#99.
Co-authored-by: daxpedda <daxpedda@gmail.com>
For earlier Windows 10 builds (pre-22000), a workaround was necessary
to fix dragging window onto a monitor with different DPI. This commit makes
the old DPI workaround to only apply conditionally on affected Windows versions.
Fixes#4041.
On macOS 26+ the window drop was leading to unwrap, since
events were coming after the window was already destroyed,
while it sounds rather strange, guard against such things just
in case.
Fixes#4333.
This is making official what's basically been the
case for the past few months at this point. I no
longer have the capacity to effectively review
and merge PRs for the Windows and X11 backends.
Thanks again for having me.
Signed-off-by: John Nunley <dev@notgull.net>
Allow updating IME state atomically to make it easier for platforms
where it's atomic by its nature, like Wayland. The old API is marked
as deprecated and is routed to the new atomic API.
Co-authored-by: dcz <gilapfco.dcz@porcupinefactory.org>
The casting was doing an incorrect check on the `ref` instead of
actually trying to downcast a ref as `cast_ref` does. So use `cast_ref`
to check whether we can safely `cast` to owned type.
* iOS: Refactor queued_gpu_redraws out from AppStateImpl
To allow AppStateImpl to be Copy, and to move redraws into the window in
the future.
* iOS AppState: Avoid RefCell and static mut
Instead, prefer Cell and Copy types, as those will never have crashes
on re-entrancy / if forgetting to make a state transition.
This is unnecessary nowadays, unwinding in CF observer callbacks is safe
(and is safe in Rust after the introduction of `extern "C-unwind"`).
Panicking elsewhere (such as in NSNotificationCenter callbacks or
delegate methods) _may_ still lead to an abort, if AppKit tries to catch
it with libc++, since Rust panics are not compatible with those.
That's "just" a quality-of-implementation detail of current Rust though,
not an inherent limitation, and should really be solved in rustc.
In Windows versions < 17763, `GetProcAddress("#132")` from `uxtheme.dll`
also returns a non-null pointer. However, the function does not match
the expected `extern "system" fn() -> bool` prototype, which causes a
crash when it is called.
This fix ensures compatibility by adding proper checks to prevent such
crashes on older Windows versions.
Generally, winit-core doesn't know about underlying platforms, though,
some general information which will true for any implementation was
left in place.
The `text` is retrieved as UTF-8 while `attributes` are based on UTF-16,
thus the offset was getting out of sync on some unicode payloads
like surrogate pairs.
Fixes#3967.
Only mark that the grab was applied when it actually got applied.
Previously there was an issue with grab being marked as applied without
a pointer over the window, when in reality it wasn't.
Fixes#4073.
Same as for `CustomCursor`. However, the API uses `dyn` stuff only
because of `Windows` backend at the time of writing, generally, platforms
should just have a separate method that deals with all of that, and e.g.
top-level winit only guarantees `Rgba`.
By removing `NamedKey::Space` and rename "super" key to "meta".
- `NamedKey::Space` is not in the spec, and doesn't make sense to
special-case. We use `Key::Character("")` instead..
I've added an extra check on the Windows backend, to ensure that the code
functionally works the same before and after. Whether that check is
desirable or not can be figured out later.
- "super" is inconsistent with the W3C spec, and while it's arguably not the
best name, it's worse that Winit is diverging and choosing a different name.
List of renamings:
- `KeyCode::SuperLeft` -> `KeyCode::MetaLeft`
- `KeyCode::SuperRight` -> `KeyCode::MetaRight`
- `KeyCode::Meta` -> `KeyCode::Super`
- `NamedKey::Meta` -> `NamedKey::Super`
- `NamedKey::Super` -> `NamedKey::Meta`
- `ModifiersState::SUPER` -> `ModifiersState::META` (deprecated)
- `ModifiersState::super_key` -> `ModifiersState::meta_key`
- `ModifiersKeys::LSUPER` -> `ModifiersKeys::LMETA`
- `ModifiersKeys::RSUPER` -> `ModifiersKeys::RMETA`
Using `Duration::Zero` with `Wait` polling mode was still blocking until
the event was actually delivered. Thus when `pump_events` API is used,
ensure that it's not happening.
Fixes#4130.
Calling the `Drop` impl of the user's `ApplicationHandler` is important on
iOS and Web, since they don't return from `EventLoop::run_app`.
And now that we reliably call `Drop`, the `ApplicationHandler::exited`
event/callback is unnecessary; using `Drop` composes much better (open files
etc. stored in the app state will be automatically flushed), and prevents
weirdness like attempting to create a new window while exiting.
* Prevent panic when calling set_simple_fullscreen(false) on macos
Calling `set_simple_fullscreen(false)` to restore the window after
a previous call to `set_simple_fullscreen(true)` panics with
`view must be installed in a window` in the call to `set_style_mask`
with the old style.
Moving the `set_style_mask` call after the frame has been resized
fixes the issue.
* Hide the doc and menubar on macos when using set_borderless_game
with set_simple_fullscreen
* Fix a pause in the event loop when clicking the title bar on windows
When clicking the title bar on Windows, to drag the window, there is
a noticible pause in continuous redraw requests. This was fixed
in #839 and then regressed in #1852. The cursor blinks in both
cases and is unrelated. The regression made the blink happen after
the pause instead of immediately.
* Update the event loop pause note on the WM_NCLBUTTONDOWN handler
The application example was also updated to optionally animate the fill color
in order to demonstrate continuous redraw without pauses in the event
loop.
Relying on just `as_any` was error prone and will become redundant in
the future, once upcasting will be stable, we also won't to impose a
restriction on to which concrete type we're casting, since casting
to a type that doesn't implement a base trait doesn't make much
sense.
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
Due to casts and use of platform specific crates in those modules
it's not really feasible to build docs for them.
After separating crates, thus should become way easier to navigate,
since backends information would be publicly available.
This also alters `VideoMode` to be a regular object and not reference
the `MonitorHandle`, since it's a static data.
Given that `VideoMode` set may change during runtime keeping the
reference as a some sort of validity may not be idea and propagating
errors when changing video mode could be more reliable.
For EventLoop, EventLoopBuilder, EventLoopProxy and by requiring it as
a supertrait of Window and ActiveEventLoop.
It is especially useful for user to be able to know that Window is Debug.
This is done to avoid order-dependent behavior that you'd otherwise
encounter where `EventLoop::new` had to be called at the beginning of
`fn main` to ensure that Winit's application was the one being
registered as the main application by calling `sharedApplication`.
Fixes https://github.com/rust-windowing/winit/issues/3772.
This should also make it (more) possible to use multiple versions of
Winit in the same application (though that's still untested).
Finally, it should allow the user to override `NSApplication` themselves
if they need to do that for some reason.
This commit does not represent a release and only synchronizes CHANGELOG
from the latest release.
This also includes 0.30.8 sync, since it was forgotten.
The serial was not unique, thus leading to issues and replay being
triggered for normal input. Track modifiers based on they keycodes
instead, since it's more unique.
Links: https://github.com/alacritty/alacritty/issues/8461
Let's just not forward events to the IME once the user requested that
it should be disabled, though, still try to change its state explicitly.
Fixes#4082.
* Add cursor position drag and drop events.
* Reword drag events to match pointer ones.
* appkit: Use `convertPoint_fromView` for coordinate conversion.
* appkit: use ProtocolObject<dyn NSDraggingInfo>.
* x11: store dnd.position as pair of i16
It's what translate_coords takes anyway, so the extra precision is
misleading if we're going to cast it to i16 everywhere it's used.
We can also simplify the "unpacking" from the XdndPosition message--we
can and should use the value of 16 as the shift instead of
size_of::<c_short> * 2 or something like that, because the specification
gives us the constant 16.
* x11: store translated DnD coords.
* x11: don't emit DragLeave without DragEnter.
* windows: only emit DragEnter if valid.
* windows: in DnD, always set pdwEffect.
It appears other apps (like Chromium) set pdwEffect on Drop too:
61a391b86b/ui/base/dragdrop/drop_target_win.cc
* docs: make it clearer that drag events are for dragged *files*.
* examples/dnd: handle RedrawRequested event.
Co-authored-by: amrbashir <amr.bashir2015@gmail.com>
When all we'll be doing is setting a new preedit, the preedit doesn't
have to be explicitly cleared first. This change is perhaps debatable.
The direct reason for this is to make it easier to work around
quirks/bugs: in Masonry we've found IBus appears to resend
the IME preedit in response to `Window::set_ime_cursor_area`
(`zwp_text_input_v3::set_cursor_rectangle`). Because currently the
preedit is first cleared, a new IME cursor area is sent, which again
causes IBus to resend the preedit. This can loop for a while.
The Wayland protocol is mechanically quite prescriptive,
it says for zwp_text_input_v3:event:done.
> 1. Replace existing preedit string with the cursor.
> 2. Delete requested surrounding text.
> 3. Insert commit string with the cursor at its end.
> 4. Calculate surrounding text to send.
> 5. Insert new preedit text in cursor position.
> 6. Place cursor inside preedit text.
Winit currently doesn't do surrounding text, so 2. and 4. can be
ignored. In Winit's IME model, without a commit, sending just the
`Ime::Preedit` event without explicitly clearing is arguably still
equivalent to doing 1., 5., and 6.
This no longer exposes `CGDisplayCreateUUIDFromDisplayID` and instead
uses `CFUUID` to avoid a leak.
Monitor comparisons should also be more stable now.
The video mode is generally a static data and not a reference to some
video mode. This changes the exclusive fullscreen API to match that an
accept a monitor now.
`AnyClass::responds_to` delegates to `class_respondsToSelector`, a
function provided by the Objective-C runtime. However, at some point,
this began to return `false` for selectors referring to undocumented
cursors, despite the cursors remaining accessible via said selectors.
That this check fails prevents the cursors from being used.
We can instead send `respondsToSelector:` to the `NSCursor` class
itself. As an instance method, this is nominally impossible; however,
Apple grants an exemption[1] that permits class objects to perform
instance methods defined in the root class.
Checking for the undocumented cursors in this way gets them working
again, at least on macOS Sequoia 15.1.1.
[1]: https://developer.apple.com/library/archive/documentation/Cocoa/Conceptual/ObjectiveC/Chapters/ocObjectsClasses.html#//apple_ref/doc/uid/TP30001163-CH11-TPXREF120
Events like `resumed`, `new_events`, `about_to_wait`, and so on will
never happen as a result of programmer action, so we'll never need to
queue those. This allows us to remove the need for the old `Event`
struct in the iOS backend.
Furthermore, we can now remove `InUserCallback`, since that state is
already stored inside `EventHandler`.
I've tried to otherwise keep the semantics as close to the original by
calling `handle_nonuser_events(mtm, [])`, which flushes pending events.
Events emitted by `flagsChanged:` cannot access
`charactersIgnoringModifiers`. We were previously doing this because we
were trying to re-use the `create_key_event` function, but that is unsuited
for this purpose, so I have separated the `flagsChanged:` logic out from it.
The `NSViewFrameDidChangeNotification` that we listen to is emitted when
`-[NSWindow setContentView]` is called, since that sets the frame of the
view as well.
So now we register the notification later, so that it's not triggered at
window creation.
This behaviour is well described in the documentation:
https://developer.apple.com/documentation/appkit/nsview/postsframechangednotifications?language=objc
Added `Window::safe_area`, which describes the area of the surface that
is unobstructed by notches, bezels etc. The drawing code in the examples
have been updated to draw a star inside the safe area, and the plain
background outside of it.
Also renamed `Window::inner_position` to `Window::surface_position`, and
changed it to from screen coordinates to window coordinates, to better
align how these coordinate systems work together.
Finally, added some SVG images and documentation to describe how all of
this works.
This is fully implemented on macOS and iOS, and partially on the web.
Co-authored-by: daxpedda <daxpedda@gmail.com>
XIM servers currently do not support preedit area reporting from
clients and there may be no standard way to report it.
Fcitx and iBus both place the candidate window descending descending
from the caret, with the reported X font; but since winit does not
report a font, the height of the line is assumed 0, so when we report
the top left corner of the cursor area they will tend to obscure it.
Taking this into account, the best default option is to report the
bottom right corner of the cursor area, because it will tend not to
obscure the preedit area when using `Window::set_ime_cursor_area` in
the way suggested by documentation.
This is not strictly required, but makes the examples a bit easier to
read understand (especially since the `EventLoop` really _should_ be
created inside `fn main`, and not in some random function potentially
running on a random thread).
The proxy is intended to be Clone, thus use `Arc` for it internally and
don't require backends for it to be `Clone`. Use `EventLoopProxyProvider`
to hide the backend's proxy implementation details.
Co-authored-by: daxpedda <daxpedda@gmail.com>
Based on https://github.com/rust-windowing/winit/pull/3973, which should be merged first.
There's an API to programmatically specify X11 screen id (override what is determined from the `DISPLAY` env variable), but it doesn't work.
Seeting up X Server with 2 screens and calling `DISPLAY=:0 X11_SCREEN_ID=1 cargo run --example window` should be equivalent to calling `DISPLAY=:0.1 cargo run --example window`
The latter works (and places the window on the correct screen), but the former yields
`failed to create initial window: Os(OsError { line: 620, file: "src/platform_impl/linux/x11/window.rs", error: X11Error(X11Error { error_kind: Match, error_code: 8, sequence: 219, bad_value: 1319, minor_opcode: 0, major_opcode: 1, extension_name: None, request_name: Some("CreateWindow") }) })`
_Here `1319` is the root window id for screen 0, which doesn't match the screen 1 that we request._
The problem is that we need to factor in the screen id when determining the parent (root) window when not explicitly set. This patch does that.
---
Also: Extend the window example with X11_{SCREEN,VISUAL}_ID env variables
Whether the pointer event is primary or not generally matters for the
context where all input is done by the same event, so users can
_ignore_ non-primary events since they are likely from users clicking
something else with their other fingers.
Having it only on a FingerId made it useless, since it's usually used
to avoid multi-touch, but if you start mapping on touch event you
already can track things like that yourself.
Fixes#3943.
Co-authored-by: daxpedda <daxpedda@gmail.com>
Previous version used [`SetTimer`] with `GetMessageW` for waiting.
The downside of UI timers like ones created by `SetTimer`,
is that they may be late by up to 15-16 ms.
To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`.
In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment.
I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events.
Unfortunately, high resolution timers are available only since Windows 10 1803. However:
1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it;
2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`.
I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic.
After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events.
I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more).
Fixes https://github.com/rust-windowing/winit/issues/1610
[`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw
[`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex
[`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer
[review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
Adds `WindowExtMacOS::set_unified_titlebar` and
`WindowAttributesExtMacOS::with_unified_titlebar`,
which allow you to use a larger titlebar style on macOS.
Key them off of `onStop()` and `onStart()` which seems to match the
other backends most closely. These [Android Activity lifecycle] events
denote when the application is visible on-screen, and recommend that any
heavy lifting for startup and shutdown happens here, as the application
may be demoted to the background and later shut down entirely unless the
user navigates back to it.
[Android Activity lifecycle]: https://developer.android.com/guide/components/activities/activity-lifecycle
WindowId is a window _identifier_, and as such doesn't store anything
(unlike a _handle_). So we can safely make only be defined once, in the
core crate.
There are a few backends where we still use `into_raw` internally; I
consider these patterns discouraged, we should not be passing around
important state in the window id.
- Rename `CursorMoved` to `PointerMoved`.
- Rename `CursorEntered` to `PointerEntered`.
- Rename `CursorLeft` to `PointerLeft`.
- Rename `MouseInput` to `PointerButton`.
- Add `position` to every `PointerEvent`.
- Remove `Touch`, which is folded into the `Pointer*` events.
- New `PointerType` added to `PointerEntered` and `PointerLeft`,
signifying which pointer type is the source of this event.
- New `PointerSource` added to `PointerMoved`, similar to `PointerType`
but holding additional data.
- New `ButtonSource` added to `PointerButton`, similar to `PointerType`
but holding pointer type specific buttons. Use
`ButtonSource::mouse_button()` to easily normalize any pointer button
type to a generic mouse button.
- In the same spirit rename `DeviceEvent::MouseMotion` to `PointerMotion`.
- Remove `Force::Calibrated::altitude_angle`.
Fixes#3833.
Fixes#883.
Fixes#336.
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
As discussed in today's meeting, this commit creates two tiers of
support. Tier 1, which is what we actively test, and Tier 2, what we
hope compiles.
Signed-off-by: John Nunley <dev@notgull.net>
Co-authored-by: daxpedda <daxpedda@gmail.com>
* Rename `WindowEvent::Resized` to `SurfaceResized`
* Rename `InnerSizeWriter` to `SurfaceSizeWriter`
* Replace `inner_size` with `surface_size`
* Rename `resize_increments` to `surface_resize_increments`
This should allow us to make future split of backends much easier.
The `Box<dyn Window>` is a _temporary_ solution, which will be
removed with the future updates when we decide on how the Window
should be stored.
This makes it so, when X11 fails to initialize due to not loading a
library, it provides more verbose information on what exactly happened.
Fixes#3883.
Signed-off-by: John Nunley <dev@notgull.net>
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
Another one bites the dust.
This replaces the code dependent on libxcursor with equivalent code
written using x11rb, featuring its special "cursor" module.
cc #3198
Signed-off-by: John Nunley <dev@notgull.net>
A panic doesn't really put any of the fields in XConnection into an invalid
state, so there is no real reason to panic when poisoning is detected.
So just ignore the poison.
Closes#3870
Signed-off-by: John Nunley <dev@notgull.net>
This implements basic iOS IME support (typing, backspace, support for emojis
etc but no autocomplete or copy / paste menu).
Co-authored-by: Mads Marquart <mads@marquart.dk>
* Update version docs to link to `rustc`'s supported versions
* Document how to run Winit on Mac Catalyst
* Improve instructions for building iOS applications
The old instructions are outdated, and suggested a workaround that is
unnecessary.
The user-story in the ecosystem is sadly not very clear-cut, so the
instructions here are still woefully incomplete.
* iOS: Clean up notes on main thread safety
These platform-specific notes on `Window` methods were unnecessary, as
it's already discussed in the top-level `Window` docs.
Instead of storing the event handler within the AppState, and extracting
it our every time we need it, we now use the same event handling
implementation as for macOS that ensures we don't re-entrantly call the
event handler, and that we un-register the handler again after we're
done using it (`UIApplicationMain` won't return, but may still unwind,
so this is very important for panic safety).
- Change `OrientationData::natural` type from `bool` to `Orientation`, just telling the user what the natural orientation is.
- Fix and improve some monitor related documentation.
This allows the user to override the application delegate themselves,
which opens several doors for customization that were previously closed.
To do this, we use notifications instead of top-level application delegate
methods.
One effect of not providing an application delegate on iOS is that we no
longer act as-if the application successfully open all URLs there.
This is a breaking change, although unlikely to matter in practice, since the
return value of `application:didFinishLaunchingWithOptions:` is seldom used by
the system (and is likely the preferred behaviour anyhow).
This removes the direct requirement to implement `as_any` and it could
be just derived.
Also implement HasDisplayHandle for dyn ActiveEventLoop + '_.
Suggested-by: daxpedda <daxpedda@gmail.com>
This was implemented for `&dyn ActiveEventLoop` before, but that's less
general than implementing it directly for `dyn ActiveEventLoop`.
The `+ '_` is required to tell the compiler that we want to implement
this for all `dyn ActiveEventLoop`s, not just `'static` ones (even
though they're all going to be `'static`).
Because we don't want to force all methods on `VideoModeHandle` to return `Option`, we decided to remove the already incomplete support in Android for both types.
`VideoModeHandle::refresh_rate_millihertz()` and `bit_depth()` now return a `Option<NonZero*>`.
`MonitorHandle::position()` now returns an `Option`.
On Orbital `MonitorHandle::name()` now returns `None` instead of a dummy name.
This also fixes macOS returning `None` in `Window::theme()` if no theme
override is set, instead it now returns the system theme.
MacOS and Wayland were the only ones working correctly according to the
documentation, which was an oversight. The documentation was "fixed"
now.
Fixes#3837.
We decided to remove them because they contained too little information
for the user to be useful. The assumption is that they were originally
implemented to enable gamepad support, which we already decided we are
not going to add directly to Winit.
This had no real use because we don't expose any information on
`DeviceId` except on Windows. But there we only expose the name. The
assumption is that this was originally added for gamepad support, which
never made it into Winit.
- Improved the documentation to point users into the right direction from all kinds of methods and types.
- De-duplicated some code and added more comments.
- Implement an ID system to correctly and efficiently implement `Eq`, `Hash`, `Ord`, `PartialEq` and `PartialOrd`.
- Fixed screen locking support being cached thread local, ergo calling from a different thread would require to make the check again, not fulfilling its purpose as a fast-path at all.
This moves our JS file to use TS instead, which allows us to use a proper linter to check the code.
All related files where moved out from the root in a dedicated folder to avoid polluting the Rust environment.
Requires getting permission from the user to get "detailed" support.
Also enables users to go fullscreen on specific monitors.
Exposes platform-specific orientation API.
Most functionality depends on browser support, currently only Chromium.
- Internal: Fix dropping `Notifier` without sending a result causing `Future`s to never complete. This should never happen anyway, but now we get a panic instead of nothing if we hit a bug.
- Internal: Remove a bunch of `unwrap()`s that aren't required when correctly using `MainThreadMarker`.
- `Window::canvas()` is now able to return a reference instead of an owned value.
Extracted from #3801.
Without smooth resizing, the window will appear to jitter when it is being
resized. This is because X11 completes the resize before the client gets a
chance to draw a new frame.
This is fixed by using the "sync" extension to ensure that the resize of the
X11 window is synchronized with the server.
Closes#2153
This allows the user more control over how they pass their application state
to Winit, and will hopefully allow `Drop` implementations on the application
handler to work in the future on all platforms.
Nothing wrong will happen if we ignore events when compositor is at
wrong, at least crashing because compositor is just _wrong_ probably is
not a great option.
Links: https://github.com/alacritty/alacritty/issues/8065
Introduced in 3a7350c with unclear motivation.
Nowadays, this feature is incomplete and unsound, and the equivalent
functionality can be trivially achieved outside of `winit` using
`objc2-ui-kit`.
Move iOS and macOS implementations to a shared folder called `apple`, to allow
us to reduce the code-duplication between these platforms in the future.
The folder structure is now:
- `src/platform_impl/apple/`
- `appkit/`
- `uikit/`
- `example_shared_file.rs`
- `mod.rs`
* Add preliminary support for tvOS, watchOS and visionOS
* Reduce duplication in Cargo.toml when specifying dependencies
Let the users wake up the event loop and then they could poll their
user sources.
Co-authored-by: Mads Marquart <mads@marquart.dk>
Co-authored-by: daxpedda <daxpedda@gmail.com>
Rust tooling generally works better this way. This includes
rust-analyzer, but more noticeably the output from `tracing` typically
prints the module path, which did not correspond to the actual file
system before.
Concretely, tracing output from the macOS backend changes from printing:
`winit::platform_impl::platform::util`
To printing:
`winit::platform_impl::macos::util`
Additionally, always queue events in `handle_scale_factor_changed`.
These events were intentionally not queued before, since they are
executed inside `handle_scale_factor_changed`, which is itself queued.
Though that's a bit too subtle, so let's just take the minuscule perf
hit of redundantly checking whether we need to queue again.
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
taiki-e/checkout-action has a few advantages over actions/checkout,
such as:
- It is written in Bash rather than Node.js
- It does not have frequent breaking changes, reducing churn
Signed-off-by: John Nunley <dev@notgull.net>
This new implementation uses:
- The NSAppearanceCustomization protocol for retrieving the appearance
of the window, instead of using the application-wide
`-[NSApplication effectiveAppearance]`.
- Key-Value observing for observing the `effectiveAppearance` to compute
the `ThemeChanged` event, instead of using the undocumented
`AppleInterfaceThemeChangedNotification` notification.
This also fixes `WindowBuilder::with_theme` not having any effect, and
the conversion between `Theme` and `NSAppearance` is made a bit more
robust.
Not using `NSColor::clearColor()` results in Quartz thinking that the
window is not transparent at all, which results in artifacts.
However, not setting the `windowBackgroundColor` in
`Window::set_transparent` results in border not properly rendered.
Fixes: 94664ff687 (Don't set the background color)
I'm preparing to get rid of our application delegate in favour of registering
notification observers, to do so I'm renaming `app_delegate.rs` to
`app_state.rs`, and moving the functionality out of the Objective-C method
into a normal method.
Additionally, `AppState` previously implemented `Default`, but really, this
was a hack done because someone (probably myself) was too lazy to write out
the full initialization in `AppDelegate::new`.
* Use AppKit's internal queuing mechanisms
This allows events to be queued in a more consistent order, they're now
interleaved with events that we handle immediately (like redraw events),
instead of being handled afterwards.
* Only queue events if necessary
This makes the call stack / backtraces easier to understand whenever
possible, and generally improves upon the order in which events are
delivered.
This fixes an issue where the window glitched due to resize
when the user doesn't actually change the window, but macOS
function to update window size was still called.
On Raspberry Pi, using the Rust crate eframe caused the program to crash on
mouse movement. The Backtrace lead to this specific line of code, and the exact
error was a "misaligned pointer dereference: address must be a multiple of 0x8
but is xxxx"
The edit has been tested with the Raspberry Pi, which works now.
When the user decides to use an older version of raw-window-handle,
through the rwh_04 or rwh_05 features, it makes sense to reexport the
crate so they don’t have to depend on it manually and can instead use
winit::raw_window_handle.
- Allow all gestures simultaneously recognized.
- Add PanGestureRecogniser with min/max number of touches.
- Fix sending delta values relative to Update instead to match macOS.
- Fix rotation gesture units from iOS to be in degrees instead of radians.
Co-authored-by: Mads Marquart <mads@marquart.dk>
Setting the background color changes how the window title bar appears,
which is something that the application should customize itself if it
wants this behaviour (and also, it wasn't set when calling
`set_transparent`, so the behaviour wasn't consistent).
On Windows, it is generally unsafe to use the HWND outside of the thread
that it originates from. In reality, the HWND is an index into a
thread-local table, so using it outside of the GUI thread can result in
another window being used instead, following by code unsoundness. This
is why the WindowHandle type is !Send. However, Window is Send and Sync,
which means we have to account for this.
Thus far the best solution seems to be to check if we are not in the GUI
thread. If we aren't, refuse the return the window handle.
For users who want to ensure the safety themselves, the unsafe API
was added.
Signed-off-by: John Nunley <dev@notgull.net>
This replaces a reference to EventLoopBuilderExtUnix in a panic
message with EventLoopBuilderExtX11 or EventLoopBuilderExtWayland.
Closes#3488
Signed-off-by: John Nunley <dev@notgull.net>
In rare cases destroying subsurfaces before the main surface could
result in a frame where the window is still shown, but decorations
got hidden, right before the window itself disappears.
Changes relevant to Winit:
- `icrate` has been deprecated in favour of separate crates per
framework, in our case `objc2-foundation` and `objc2-app-kit` (and in
the future `objc2-ui-kit` on iOS).
- Moved `MainThreadMarker::run_on_main` to free-standing function
`run_on_main`.
- Changed how features work, this should result in less code that we
need to compile.
- Enums are now real structs instead of type-aliases and free constants.
Give better recommendations for contributors on how to propose start the
work, write commit messages, and handle review by explicitly stating
expectations, however make it clear that maintainers will help you if
you have issues.
Write down the guidelines for maintainers on how they should handle
contributions and interact with other maintainers to help maintaining
good code quality level by listing maintainers responsibilities and
providing suggestions on how to handle various situations.
Co-authored-by: John Nunley <dev@notgull.net>
Signed-off-by: John Nunley <dev@notgull.net>
This is a breaking change, because the system versions of the libraries
used by sctk-adwaita were changed. Such changes cascade through all
the deps, so all libraries using winit MUST do a breaking change.
Due to a copy/paste error a comment in the migration in the changelog mentioned
"window events" in a context that was clearly meant for device events. This
commit fixes this issue.
Add a migrations for the big API changes showcasing how to upgrade
some common snippets of old code to new APIs.
Co-authored-by: Mads Marquart <mads@marquart.dk>
There are AV rules out there which cause almost any
program that contains github URLs to become marked as malware.
While those rules are spurious, they're years old, and AV vendors have a
very poor reputation at appropriately following up with these problems.
Remove these strings from the panic data present in the binary
prevents binaries linking the winit from being spuriously marked as
Trojan:Win32/Wacatac.B!ml.
In particular, we don't want to emit those events inside of
`pressureChangeWithEvent:`, since the mouse motion value is sometimes
outdated.
Additionally, we want to ensure the events have been emitted during
other gestures.
Fixes https://github.com/rust-windowing/winit/issues/3516
This fixes issues where a hidden and grabbed cursor could leave the
window and become visible on top of the windows taskbar (and potentially
leave the window altogether if the taskbar is clicked) under at least
two occasions:
- When a window is overlapping the taskbar.
- When a window is maximized and Automatically hide the taskbar has
been enabled.
This approach of confining the cursor to the center of the window is
used in SDL.
Split changelog file to make it more comprehensible when reading and
also make it a part of documentation so it'll be more discoverable
by the users. This change also makes it possible for rust code inside
the changelogs to be tested with `cargo`.
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
The conversion of PhysicalUnit was wrongly computed and the tests were
not running on the CI for the dpi crate, thus it was missed, even though
the test was correctly failing.
Signed-off-by: John Nunley <dev@notgull.net>
Signed-off-by: Kirill Chibisov <contact@kchibisov.com>
As previously discussed in today's meeting, this commit adds a code
of conduct to winit. I've elected to re-use the Rust project's Code
of Conduct, as it is also in use by other popular Rust projects and I
agree with its intentions. The goal is to set forward a set of explicit
expectations for how discourse in the rust-windowing project should be
conducted.
I've deliberately left out a mechanism for enforcing this code of
conduct. My hope is that, by the act of setting these expectations,
contributors will voluntarily follow them. In addition I have no
interest in being an internet janitor. If this becomes necessary in the
future we can look into it later.
rust-windowing collaborators, please read the linked Code of Conduct.
Not only for expectations of behavior, but also for expectations of how
it will be enforced should it come to it.
Signed-off-by: John Nunley <dev@notgull.net>
Add a simple `ApplicationHandler` trait since winit is moving towards
trait based API. Add `run_app` group of APIs to accept `&mut impl
ApplicationHandler` deprecating the old `run` APIs.
Part-of: https://github.com/rust-windowing/winit/issues/3432
Tracing is a modern replacement for the log crate that allows for
annotating log messages with the function that they come from.
Signed-off-by: John Nunley <dev@notgull.net>
Closes: #3482
Usually, if mouse events are equal to (0, 0) we filter them out.
However, if the event is very close to zero it will still be given to
the user. In some cases this can be caused by bad float math on the X11
server side.
Fix it by filtering absolute values smaller than floating point epsilon.
Signed-off-by: John Nunley <dev@notgull.net>
Closes: #3500
* Convert usage of Lazy to OnceLock on macOS and iOS
* Remove a few uses of Lazy that wrapped Mutex or RwLock
The `new` functions on these were made `const` in Rust 1.63.0
* Use AtomicBool instead of RwLock
This makes our use of `unsafe` to make the event handler temporarily 'static be local to a module, in a way that's (hopefully) much easier to reason about.
Removes the once_cell dependency, instead using std::sync::OnceLock and a
minimal polyfill for std::sync::LazyLock, which may be stabilized soon
(see rust-lang/rust#121377).
This should not require a bump in MSRV, as OnceLock was stabilized in 1.70,
which this crate is using.
* Move platform-specific documentation to `winit::platform` module
* Document cargo features in crate docs
* Move version requirements to crate-level docs
Given that `ModifiersChanged` is a window event, it means that clients
may track it for each window individually, thus not sending it between
focus changes may result in modifiers getting desynced on the consumer
side.
Replace the `CustomCursorBuilder` with the `CustomCursorSource` and
perform the loading of the cursor via the
`EventLoop::create_custom_cursor` instead of passing it to the builder
itself.
This follows the `EventLoop::create_window` API.
Creating window when event loop is not running generally doesn't work,
since a bunch of events and sync OS requests can't be processed. This
is also an issue on e.g. Android, since window can't be created outside
event loop easily.
Thus deprecate the window creation when event loop is not running,
as well as other resource creation to running event loop.
Given that all the examples use the bad pattern of creating the window
when event loop is not running and also most example existence is
questionable, since they show single thing and the majority of their
code is window/event loop initialization, they wore merged into
a single example 'window.rs' example that showcases very simple
application using winit.
Fixes#3399.
Mainly fix typos in comments, but also some minor code changes:
* Rename `apply_on_poiner` to `apply_on_pointer`.
* Rename `ImeState::Commited` to `ImeState::Committed`
* Correct `cfg_attr` usage: `wayland_platfrom` -> `wayland_platform`.
While there's a separate event to deliver modifiers for keyboard,
unfortunately, it's not even remotely reflects the modifiers state.
Thus use events along side regular modifier updates to correctly
detect the state. Also, apply the modifiers from the regular
key event by converting their state to xkb modifiers state.
Links: https://github.com/alacritty/alacritty/issues/7549Closes: #3388
We could fail to setup property watcher and fail to start, thus
don't require XSETTINGS to work.
Fixes: df8805c0 (On X11, reload DPI on _XSETTINGS_SETTINGS)
This fixes an issue when setting the position of the window on a new monitor and immediately maximizing it
```rs
window.set_outer_position::<PhysicalPosition<u32>>((2000, 200).into());
window.set_maximized(true);
```
Due to the nature of the event loop, the requested position and maximization state will apply correctly but due to the fact that the new position is a different monitor, a `ScaleFactorChanged` is emitted afterwards to the evenloop and a new size is set while the window is still maximized which results in a window that has `WS_MAXIMZE` window style but doesn't cover the whole monitor.
Given that we merge all the seats, we should consider that window
is not focused once all seats wl_keyboards are no longer present.
We use seats instead of keyboards to track focus to protect against
wl_keyboard::leave not being delivered when removing the seat
(usually it's not the case though).
Fixes: #3376
This also fixes the deadlock when such reload may happen.
Fixes: #3383
Signed-off-by: John Nunley <dev@notgull.net>
Signed-off-by: Kirill Chibisov <contact@kchibisov.com>
Change in state requires a redraw, however drawing when getting
`Occluded` with vsync will block indefinitely, thus the event in
it's current state is rather useless.
To solve this issue winit needs a way to determine whether the user
paused/continued their render loop, so it can commit on their behalf.
This commit also forces redraw when getting configure.
Links: https://github.com/rust-windowing/winit/issues/3442
We use softbuffer as a dev-dependency for rendering into our windows in
examples. However, we do not support a DRM/KMS backend yet, while
softbuffer comes with a DRM/KMS backend by default. This commit removes
the DRM/KMS feature from softbuffer to save some build time during
testing
This let statement swapped the two names, resulting in incorrect
behavior since commit d7ec899d. That commit did not actually introduce
the swap, but the previous code swapped it again before setting the
WM_CLASS property, so no issue was ever observed.
It also brings the documentation in line with the implementation since the
parent commit, and with the ICCCM standard, which states the following
about the WM_CLASS property [1]:
The two strings, respectively, are:
* A string that names the particular instance of the application [...]
* A string that names the general class of applications [...]
[1] https://www.x.org/releases/current/doc/xorg-docs/icccm/icccm.html#WM_CLASS_Property
This was supposed to be rolled out with the rwh v0.6 update, but it
was left behind for some reason. I've added this type back.
Signed-off-by: John Nunley <dev@notgull.net>
Previously we had a sort of artificial split between these, but both were accessing each other's state, since it's really the same state!
It was especially difficult to follow what happens to the fullscreen state.
So instead, we basically merge the window and the delegate files.
This does unfortunately screw a bit with the git history, apologies to whoever reads this in the future!
* make `EventLoopWindowTarget` independent of UserEvent type
the `EventLoopWindowTarget` is needed for window creation. conceptually,
only `EventLoop` and `EventLoopProxy` need to be parameterized, and all
other parts of the backend should be agnostic about the user event type,
parallel to how `Event<T>` is parameterized, but `WindowEvent` is not.
this change removes the dependency on the type of user events from the
`EventLoopWindowTarget` for the Windows backend, but keep a phantom data
to keep the API intact. to achieve this, I moved the `Receiver` end of
the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so
the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`,
all other part of the backend just use unit type as a placeholder for
user events.
it's similar to the macos backend where an erased `EventHandler` trait
object is used so all component except `EventLoop` and `EventLoopProxy`
need to be parameterized. however `EventLoop` of the Windows backend
already use an `Box<dyn FnMut>` to wrap the user provided event handler
callback, so no need for an dedicated trait object, I just modified the
wrapper to replace the placeholder user event with real value pulled
from the channel. I find this is the approach which need minimum change
to be made to existing code. but it does the job and could serve as a
starting point to future Windows backend re-works.
* fix CI clippy failure.
* make UserEventPlaceholder a new type instead of alias
* invariance is maintained by top-level EventLoopWindowTarget<T>
this field is transitional and her to keep API compatibility only.
the correct variance and such is already ensured by the top-level
`EventLoopWindowTarget`, just use `PhantomData<T>` here.
xdotool will update modifiers before Xkb will actually send event
updating them, thus the modifiers will be updating even before the
actual update, which is unfortunate.
Links: https://github.com/alacritty/alacritty/issues/7502
The user may change the size during the on-going resize, meaning that
the size will desync with winit's internal loop which breaks viewporter
setup with fractional scaling.
Links: https://github.com/alacritty/alacritty/issues/7474
This appears to be the solution for the elusive #3335 issue. Previously,
in the Xlib backend, we used the "XRRQueryVersion" function to query for
the Xrandr version, and used that to determine whether we should use the
"GetScreenResources" call or the "GetScreenResourcesCurrent" call.
However, we passed the version "0, 0" into "XRRQueryVersion".
Previously with Xlib this wasn't a problem, as Xlib ignores the version
you pass in and substitutes it with the version of RandR it expects.
https://gitlab.freedesktop.org/xorg/lib/libxrandr/-/blob/master/src/Xrandr.c?ref_type=heads#L386-387
The way that "XRRQueryVersion" is implemented on the server end, it
compares the version passed into the request with the version supported
by the server. If the server's version is greater than the client
version, it just returns the client version. If the client's version is
greater, it passes the server's version. Since we were passing in "0, 0"
this means that the server returned "0, 0".
https://gitlab.freedesktop.org/xorg/xserver/-/blob/master/randr/rrdispatch.c?ref_type=heads#L50-59
To determine whether we use "GetScreenResources" or
"GetScreenResourcesCurrent", we compare the version returned by the
server against "1, 3". Since we got "0, 0"- a version of XRandR so old
it doesn't even exist- we use "GetScreenResources".
The problem manifests in that "GetScreenResources" can take several
seconds to query the screen state based on the current hardware
configuration. On the other hand, "GetScreenResourcesCurrent" is fast;
it uses the server's hardware cache if it is available.
This problem is visible in XTrace. On the latest `master`:
```
000:<:00c2: 12: RANDR-Request(140,0): QueryVersion major-version=0 minor-version=0
000:>:00c2:32: Reply to QueryVersion: major-version=0 minor-version=0
000:<:00c3: 8: RANDR-Request(140,8): GetScreenResources window=0x0000076e
000:>:00c3:1600: Reply to GetScreenResources:
```
On the `v0.28.0` tag:
```
000:<:0019: 12: RANDR-Request(140,0): QueryVersion major-version=1 minor-version=6
000:>:0019:32: Reply to QueryVersion: major-version=1 minor-version=6
...later
000:<:002d: 8: RANDR-Request(140,25): GetScreenResourcesCurrent window=0x0000076e
000:>:002d:1600: Reply to GetScreenResourcesCurrent
```
This commit fixes this issue by requesting "1, 3" instead. This returns
the version we expect, where we can now use "GetScreenResourcesCurrent"
properly.
Fixes#3335
Signed-off-by: John Nunley <dev@notgull.net>
This code confused me. I tried to understand it. I tried to simplify it
while keeping the functional style. But in the end, this just seems too
complicated for its own good. Just doing the exact same thing with a
match statement and the question mark operator makes it sooo much more
obvious what is happening.
Signed-off-by: Uli Schlachter <psychon@znc.in>
* Clean up macOS and iOS monitor code a bit
* Clean up window size methods
Use `setContentSize`, `setContentMinSize`, `setContentMaxSize` and `contentRectForFrameRect` to let the windowing system figure out the required scaling, instead of us doing it manually.
* Use a flipped NSView coordinate system
* Clean up window position methods
* Use icrate's window structs and enums
* Properly implement protocols
* Use icrate's NSWindow
We were previously using undocumented methods on `NSWindowTabGroup`
* Use icrate's NSApplication
And clean up some doc comments regarding NSApplication
* Refactor winit-specific cursor logic out of appkit module
* Add relevant AppKit features that we depend on
* Use icrate's NSImageRep and NSBitmapImageRep
* Use icrate's NSImage
* Use icrate's NSCursor
* Use icrate's NSAppearance
* Use icrate's NSScreen
* Use icrate's NSButton
* Use icrate's NSAppKitVersionNumber
* Use icrate's NSTextInputContext
* Use icrate's NSColor
* Use icrate's NSEvent
* Use icrate's NSMenu and NSMenuItem
* Use icrate's NSPasteboard
* Use icrate's NSResponder
* Use icrate's NSTextInputClient
* Use icrate's NSView
There seems to be many PRs relating to this issue, but they don't include all
platforms and for some reason lost steam. This PR again tries to make this
feature happen, and does it for all desktop platforms (x11, wayland, macos,
windows, web).
I think the best user of this feature and the reason I'm doing this is Bevy and
game engines in general. There non laggy hardware cursors with custom images are
very important. Game devs also like their PNGs so supporting platform native
cursor files is not that important, but I guess could be added too.
Co-authored-by: daxpedda <daxpedda@gmail.com>
Co-authored-by: Mads Marquart <mads@marquart.dk>
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
This protocol is only used for (optional) Client Side Decorations
(where) the compositor still takes the burden of compositing various
window parts together, via subsurfaces that all belong to a single
window.
If this core protocol is not available, as is the case on gamescope,
disable CSD.
The only breaking change is that x11rb no longer reports an error when
querying the WmSizeHints of a window that does not have this property
set. For this reason, the return type of WmSizeHintsCookie::Reply()
changed from Result<WmSizeHints, SomeError> to
Result<Option<WmSizeHints>, SomeError>.
In update_normal_hints(), previously a cryptic error would be reported
to the caller. Instead, this now uses unwrap_or_default() to get a
WmSizeHints. All fields of WmSizeHints are Options, so this produces an
empty object.
resize_increments() queries a value from the window and returns an
Option. Previously, the error for "missing property" was turned into
None via .ok(). This commit adds a call to flatten() to also turn
"property not set" into None.
Finally, request_user_attention() queries a window's WmHints property
and updates one field of it. The code already uses unwrap_or_default()
to deal with missing properties, so just a call to flatten() is needed
to merge "missing property" and "error while querying" into one.
Other changes in x11rb do not seem to affect this crate.
x11rb's MSRV increased from 1.56 to 1.63, which is still below the MSRV
of this crate, which is 1.65.
Signed-off-by: Uli Schlachter <psychon@znc.in>
Closes#3245
notgull forgot to properly interpret float data from the X server,
making him tonight's biggest loser.
Signed-off-by: John Nunley <dev@notgull.net>
The surface was automatically dropped due to new RAII type in SCTK
when dropping the Window, which was not the case at some point with
SCTK.
Thus destroying objects associated with it where causing issues
with some window managers.
Links: https://github.com/neovide/neovide/issues/2109
We decided to add `rwh_06` to the `default` list of features in the
`ndk` to [nudge users to upgrade], but this forces `winit` to always
(transitively) include `raw-window-handle 0.6` even if the user has
set a different `rwh_xx` feature on the `winit` crate. `winit` already
forwards the respective `rwh_xx` feaure to the `ndk` crate anyway, so
this default should just be turned off.
At the time of writing this is the only `default` feature of the `ndk`.
Links: https://github.com/rust-mobile/ndk/pull/434#issuecomment-1752089087
Keep the user provided size in the original values and convert only
when we're getting a `configure` event. On some compositors will
have a scale available, so it'll work, however with some we'll
still have old 'pick 1` as default.
Also configure_bounds when compositor tells the user to pick the size,
that will ensure that initial `with_inner_size` won't grow beyond the
working area.
Fixes#3187.
The cursor hittest was not reloaded on window size changes, only
when `Window::request_inner_size` was called leading to regions
of the window being not clickable.
Also, don't try to apply hittest logic when user never requested a
hittest.
Links: https://github.com/alacritty/alacritty/pull/7220
While working with device events, I noticed that there was an inconsistency in the mouse button device events between Windows/X11 and for example web, because web uses the same ids/order as the MouseButton enum, and Windows/X11 are using the X11 ids, and hwheel device event was ignored on Windows.
Mouse button device events are now using the same order as the MouseButton enum, and I also added hwheel device events for Windows.
When calling `Window::request_redraw` from the `RedrawRequested`
handler the `RedrawWindow` won't result in `WM_PAINT` being delivered
due since user callback is run before `DefWindowProcW` is called.
Track whether the user called `Window::request_redraw` and ask for
`RedrawWindow` after running the said function during `WM_PAINT`
handling.
Fixes#3150.
Split `Key` into clear categories, like `Named`, `Dead`, Character`, `Unidentified`
removing the `#[non_exhaustive]` from the `Key` itself.
Similar action was done for the `KeyCode`.
Fixes: #2995
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
Sometimes FlagsChanged events don't carry any KeyCode information, thus
we can't create a synthetic presses events for them.
However in such cases, modifiers information is still accurate, thus
propagate it.
Fixes#2872.
Window builder is always accessed by winit on the thread event loop
is on, thus it's safe to mark the data it gets as `Send + Sync`.
Each unsafe object is marked individually as `Send + Sync` instead
of just implementing `Send` and `Sync` for the whole builder.
Fixing this could require a massive rework to how redraw is handled
on windows to the point of removing `WM_PAINT`, since it's not reliable
by any means for our use case.
For now at least document that the API is broken. It was broken like
that for a long while.
The update is pretty minor, however we support now
`WindowEvent::Occluded` when xdg-shell v6 is available.
It also adds support for `Window::show_window_menu`.
Fixes#2927.
While the changelog entries for beta releases doesn't really matter. The
change wasn't marked as breaking, while it is.
Fixes: 93f1000a0 (Add Occluded and MemoryWarning events for iOS/Android)
Hook `Occluded` event to foreground/background evens on iOS.
This commit also enabled the `MemoryWarning` event, since it's
emitted from the windowing system.
Co-authored-by: Dusty DeWeese <dustin.deweese@gmail.com>
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
Add a method to request a system menu. The implementation
is provided only on Windows for now.
Co-authored-by: daxpedda <daxpedda@gmail.com>
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
The new implementation of select_xkb_events apparently misconfigures
the server. This commit does a temporary fix by just reverting it to its
previous implementation.
This is temporary until I can figure out what Xlib is doing behind the
scenes or until I read xkbproto.pdf.
Fixes: #3079
Signed-off-by: John Nunley <dev@notgull.net>
They are safe, since they use the rust `std::env` stuff. Making them
safe lets downstream to determine that `std::env` is used and not the
`libc` env manipulation routines, which are unsafe.
* Make Linux platforms less dependent on the root monitor handle
* Add various functions to the Wayland platform to reduce cfgs
* Don't use a cfg in listen_device_events
* Don't use a cfg in set_content_protected
* Fix instance of a target_os cfg
Removes Xlib code by replacing it with the x11rb equivalent,
the commit handles xrandr, xinput, xinput2, and xkb.
Signed-off-by: John Nunley <dev@notgull.net>
* macOS & iOS: Refactor EventWrapper
* macOS & iOS: Make EventLoopWindowTarget independent of the user event
* iOS: Use MainThreadMarker instead of marking functions unsafe
* Make iOS thread safe
We use raw-window-handle extensive in the public API as well as we
force the users to use it to get some essential data for interop, thus
reexport it.
Fixes: #2913.
Inner panics could make it hard to trouble shoot the issues and for some
users it's not desirable.
The inner panics were left only when they are used to `assert!` during
development.
This reverts commit 9f91bc413fe20618bd7090829832bb074aab15c3 which
reverted the original patch which was merged without a proper review.
Fixes: #500.
Up until now the Android backend has been directly mapping key codes
which essentially just represent the "physical" cap of the key (quoted
since this also related to virtual keyboards).
Since we didn't account for any meta keys either it meant the backend
only supported a 1:1 mapping from key codes, which only covers a tiny
subset of characters. For example you couldn't type a colon since
there's no keycode for that and we didn't try and map Shift+Semicolon
into a colon character.
This has been tricky to support because the `NativeActivity` class doesn't
have direct access to the Java `KeyEvent` object which exposes a more
convenient `getUnicodeChar` API.
It is now possible to query a `KeyCharcterMap` for the device associated
with a `KeyEvent` via the `AndroidApp::device_key_character_map` API
which provides a binding to the SDK `KeyCharacterMap` API in Java:
https://developer.android.com/reference/android/view/KeyCharacterMap
This is effectively what `getUnicodeChar` is implemented based on and is
a bit more general purpose.
`KeyCharacterMap` lets us map a key_code + meta_state from a `KeyEvent`
into either a unicode character or dead key accent that can be combined
with the following key. This mapping is done based on the user's chosen
layout for the keyboard.
To enable support for key character maps the
`AndroidApp::input_events()` API was replaced by
`AndroidApp::input_events_iter()` which returns a (lending) iterator for
events. This was changed because the previous design made it difficult
to allow other AndroidApp APIs to be used while iterating events (mainly
because AndroidApp held a lock over the backend during iteration)
Inner panics could make it hard to trouble shoot the issues and for some
users ints not desirable.
The inner panics were left only when they are used to `assert!` during
development.
At the moment, the with_x11_visual function takes a pointer and
immediately dereferences it to get the visual info inside. As it is safe
to pass a null pointer to this function, it is unsound. This commit
replaces the pointer parameter with a visual ID, and then uses that ID
to look up the actual visual under
the X11 setup. As this is what was already practically happening before,
this change shouldn't cause any performance downgrades.
This is a breaking change, but it's done in the name of soundness so it
should be okay. It should be trivial for end users to accommodate it,
as it's just a matter of getting the visual ID from the pointer to the
visual before passing it in.
Signed-off-by: John Nunley <dev@notgull.net>
There's no need to force the static on the users, given that internally
some backends were not using static in the first place.
Co-authored-by: daxpedda <daxpedda@gmail.com>
Mark it as breaking, since some clients relied on that behavior, simply
because dispatching clients queue always woke up a winit, meaning that
they won't be able to use user events for this sake.
Nothing changed from the user point of view, other than they should
use the `raw-window-handle`, which is objectively better, given that
it reduces the amount of `cfg` guards in downstream code.
#2662 renamed `VirtualKeyCode` to `Key` yet references to the former
type still exist in `src/platform_impl/linux/x11/events.rs`. As it
turns out the `mod events;` in `x11/mod.rs` was removed in the same PR,
but the file accidentally stuck around without being referenced anywhere
else.
Lifetimes don't work nicely when dealing with multithreaded environments
in the current design of the existing winit's event handling model, so
remove it in favor of `InnerSizeWriter` fences passed to client, so they
could try to update the size.
Fixes#1387.
The idea that redraw events are dispatched with a specific ordering
that makes it possible to specifically report when we have finished
dispatching redraw events isn't portable and the way in which we
dispatched RedrawEventsCleared was inconsistent across backends.
More generally speaking, there is no inherent relationship between
redrawing and event loop iterations. An event loop may wake up at any
frequency depending on what sources of input events are being listened
to but redrawing is generally throttled and in some way synchronized
with the display frequency.
Similarly there's no inherent relationship between a single event loop
iteration and the dispatching of any specific kind of "main" event.
An event loop wakes up when there are events to read (e.g. input
events or responses from a display server / compositor) and goes back
to waiting when there's nothing else to read.
There isn't really a special kind of "main" event that is dispatched
in order with respect to other events.
What we can do more portably is emit an event when the event loop
is about to block and wait for new events.
In practice this is very similar to how MainEventsCleared was
implemented except it wasn't the very last event previously since
redraw events could be dispatched afterwards.
The main backend where we don't strictly know when we're going to
wait for events is Web (since the real event loop is internal to
the browser). For now we emulate AboutToWait on Web similar to how
MainEventsCleared was dispatched.
In practice most applications almost certainly shouldn't care about
AboutToWait because the frequency of event loop iterations is
essentially arbitrary and usually irrelevant.
Considering the possibility of re-running an event loop via run_ondemand
then it's more correct to say that the loop is about to exit without
assuming it's going to be destroyed.
This renames all internal implementations of pump_events_with_timeout
to pump_events and makes them public.
Since all platforms that support pump_events support timeouts there's
no need to have a separate API.
This layers pump_events on a pump_events_with_timeout API, like we have
for Linux and Android.
This is just an internal implementation detail for now but we could
consider making pump_events_with_timeout public, or just making it so
that pump_events() takes the timeout argument.
Considering the strict requirement that applications can't keep windows
across run_ondemand calls, this tries to make the window_ondemand example
explicitly wait for its Window to be destroyed before exiting each
run_ondemand iteration.
This updates the example to only `.set_exit()` after it gets a
`Destroyed` event after the Window has been dropped.
On Windows this works to ensure the Window is destroyed before the
example waits for 5 seconds.
Unfortunately though:
1. The Wayland backend doesn't emit `Destroyed` events for windows
2. The macOS backend emits `Destroyed` events before the window is
really destroyed.
and so the example isn't currently portable.
Although we document that applications can't keep windows between
separate run_ondemand calls it's possible that the application has only
just dropped their windows and we need to flush these requests to the
server/compositor.
This fixes the window_ondemand example - by ensuring the window from
the first loop really is destroyed before waiting for 5 seconds
and starting the second loop.
A minimal example that shows an application running the event loop more
than once via `run_ondemand`
There is a 5 second delay between each run to help highlight problems
with destroying the window from the first loop.
This re-works the portable `run()` API that consumes the `EventLoop` and
runs the loop on the calling thread until the app exits.
This can be supported across _all_ platforms and compared to the
previous `run() -> !` API is now able to return a `Result` status on all
platforms except iOS and Web. Fixes: #2709
By moving away from `run() -> !` we stop calling `std::process::exit()`
internally as a means to kill the process without returning which means
it's possible to return an exit status and applications can return from
their `main()` function normally.
This also fixes Android support where an Activity runs in a thread but
we can't assume to have full ownership of the process (other services
could be running in separate threads).
Additionally all examples have generally been updated so that `main()`
returns a `Result` from `run()`
Fixes: #2709
Wayland:
I found the calloop abstraction a little awkward to work with while I was
trying to understand why there was surprising workaround code in the wayland
backend for manually dispatching pending events.
Investigating this further it looks like there may currently be several issues
with the calloop WaylandSource (with how prepare_read is used and with (not)
flushing writes before polling)
Considering the current minimal needs for polling in all winit backends I do
personally tend to think it would be simpler to just own the responsibility for
polling more directly, so the logic for wayland-client `prepare_read` wouldn't
be in a separate crate (and in this current situation would also be easier to fix)
I've tried to maintain the status quo with calloop + workarounds.
X11:
I found that the recent changes (4ac2006cbc) to port the X11 backend
from mio to calloop lost the ability to check for pending events before
needing to poll/dispatch. (The `has_pending` state being queried
before dispatching() was based on state that was filled in during
dispatching)
As part of the rebase this re-introduces the PeekableReceiver and
WakeSender which are small utilities on top of
`std::sync::mpsc::channel()`. This adds a calloop `PingSource`
so we can use a `Ping` as a generic event loop waker.
For taking into account false positive wake ups the X11 source now
tracks when the file descriptor is readable so after we poll via
calloop we can then specifically check if there are new X11 events
or pending redraw/user events when deciding whether to skip the
event loop iteration.
The implementation of `pump_events` essentially works by hooking into the
`RunLoopObserver` and requesting that the app should be stopped the next time
that the `RunLoop` prepares to wait for new events.
Originally I had thought I would poke the `CFRunLoop` for the app directly and
I was originally going to implement `pump_events` based on a timeout which I'd
seen SDL doing.
I found that `[NSApp run]` wasn't actually being stopped by asking the RunLoop
to stop directly and inferred that `NSApp run` will actually catch this and
re-start the loop.
Hooking into the observer and calling `[NSApp stop]` actually seems like a
better solution that doesn't need a hacky constant timeout.
The end result is quite similar to what happens with existing apps that
call `run_return` inside an external loop and cause the loop to exit for
each iteration (that also results in the `NSApp` stopping each
iteration).
A surprising amount of work was required to enable these extensions
on Windows.
I had originally assumed that pump_events was going to be very similar
to run except would use PeekMessageW instead of GetMessageW to avoid
blocking the external loop but I found the Windows backend broke
several assumptions I had.
Overall I think these changes can hopefully be considered a quite a
significant simplification (I think it's a net deletion of a fair amount
of code) and I think it also helps bring it into slightly closer alignment
with other backends too
Key changes:
- I have removed the `wait_thread` that was a fairly fiddly way of handling
`ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works
with the same messages picked up by `GetMessage` and `PeekMessage`.
- I have removed the ordering guarantees between `MainEventsCleared`,
`RedrawRequested` and `RedrawEventsCleared` events due to the complexity in
maintaining this artificial ordering, which is already not supported
consistently across backends anyway (in particular this ordering already
isn't compatible with how MacOS / iOS work).
- `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages
- comparable to how `RedrawRequested` is dispatched via `drawRect` in the
MacOS backend.
- I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared`
get dispatched to be more in line with the MacOS backend and also more in line
with how we have recently discussed defining them for all platforms.
`NewEvents` is conceptually delivered when the event loop "wakes up" and
`MainEventsCleared` gets dispatched when the event loop is about to ask the
OS to wait for new events.
This is a more portable model, and is already how these events work in the
MacOS backend.
`RedrawEventsCleared` are just delivered after `MainEventsCleared` but this
event no longer has a useful meaning.
Probably the most controversial thing here is that this "breaks" the ordering
rules for redraw event handling, but since my changes interacted with how the
order is maintained I was very reluctant to figure out how to continue
maintaining something that we have recently been discussing changing:
https://github.com/rust-windowing/winit/issues/2640.
Additionally, since the MacOS backend already doesn't strictly maintain this
order it's somewhat academic to see this as a breakage if Winit applications
can't really rely on it already.
This updates the documentation for `request_redraw()` to reflect that we
no longer guarantee that `RedrawRequested` events must be dispatched
after `MainEventsCleared`.
This adds two new extensions for running a Winit event loop which will
replace `EventLoopExtRunReturn`
The `run_return` API is trying to solve multiple problems and address
multiple, unrelated, use cases but in doing so it is not succeeding
at addressing any of them fully.
The notable use cases we have are:
1. Applications want to be able to implement their own external
event loop and call some Winit API to poll / pump events, once
per iteration of their own loop, without blocking the outer,
external loop. Addressing #2706
2. Applications want to be able to re-run separate instantiations
of some Winit-based GUI and want to allow the event loop to exit with
a status, and then later be able to run the loop again for a new
instantiation of their GUI. Addressing #2431
It's very notable that these use cases can't be supported across
all platforms and so they are extensions, similar to
`EventLoopExtRunReturn`
The intention is to support these extensions on:
- Windows
- Linux (X11 + Wayland)
- macOS
- Android
These extensions aren't compatible with Web or iOS though.
Each method of running the loop will behave consistently in terms of how
`NewEvents(Init)`, `Resumed` and `LoopDestroyed` events are dispatched
(so portable application code wouldn't necessarily need to have any awareness
of which method of running the loop was being used)
Once all backends have support for these extensions then we can
remove `EventLoopExtRunReturn`
For simplicity, the extensions are documented with the assumption that
the above platforms will be supported.
This patch makes no functional change, it only introduces these new
extensions so we can then handle adding platform-specific backends
in separate pull requests, so the work can be landed in stages.
Unfortunately this isn't a total removal, for two reasons:
- We still need "libc" for the Xlib XIM implementation, for locales.
- BSD requires libc to check for main-threadedness.
First one we can likely resolve in the near future, not so sure about
the second one without using some weird pthreads trick.
The utils in this module should help the users to activate the windows
they create, as well as manage activation tokens environment variables.
The API is essential for Wayland in the first place, since some
compositors may decide initial focus of the window based on whether
the activation token was during the window creation.
Fixes#2279.
Co-authored-by: John Nunley <jtnunley01@gmail.com>
It was discovered that on GNOME the click sometimes being swallowed
by the mutter's `wl_pointer::enter/leave` sequences. This was happening
due to `xdg_toplevel::move` making the pointer to leave the surface.
To make handling of that more robust, we could start the
`xdg_toplevel::move` when the actual pointer motion is being performed.
Links: https://github.com/alacritty/alacritty/issues/7011
Links: https://gitlab.gnome.org/GNOME/mutter/-/issues/2669#note_1790825
This should provide a way to iterate all the tabs and select the last
tab. The tab indicies are now zero based as any other sane index.
Follow-up-to: c5941d105f (add tabbing API)
Winit now supports tabbing identifiers, thus set that we prefer tabbing,
in particular it'll make windows tab when using the same tabbing identifiers,
which is desirable for the end users.
Some systems could resize the window immediately and we'd rather
inform the users right away if that was the case, so they could
create e.g. EGLSurface without waiting for resize, which is really
important for Wayland.
Fixes#2868.
Instead of a single `bool` indicating that a key press has occured and
no key has been released since then, we store the scancode of the last
pressed key (if it is a key that repeats when held). This fixes a bug
where pressing a new key while one is already held down will be flagged
as a repeat even though it is obviously not a repeat.
* Make iOS declared classes not use &mut
* Prepare `init` methods for not having access to &mut self
* Prepare WinitWindow methods for not having access to &mut self
* Convert a bit of WinitView's to use interior mutability
* Convert a bit more of WinitView's to use interior mutability
* Convert the rest of WinitView to use interior mutability
* Use interior mutability instead of a Mutex for the CursorState
* Use interior mutability in WinitWindowDelegate
The change to xinput2 completely disabled IME support, thus we've got
a dead keys reporting, because nothing was eating the key events
anymore, however that's not what we really need, given that not
working IME makes it impossible for some users to type.
The proper solution is to not use Xlib at all for that and rely on
xcb and its tooling around the XIM and text compose stuff, so
we'll have full control over what is getting sent to the XIM/IC or not.
Fixes#2888.
Fixes issue on Wayland due to drop order, since TLS is being dropped
after the event loop, while it shouldn't. In particular it fixes the
crash in the window_run_return example.
During the migration some logic wrt `none` decorations was lost along
the way, however we also now try to ask for client side decorations if
the user wants to disable server side decorations.
Fixes#2902.
The recent overhaul of the keyboard API broke keyboard input on Android.
The recent keyboard changes also broke building against the
game-activity backend of android-activity because it was assumed that
the backend is based on the NDK input API which isn't the case with
with game-activity since it doesn't use the InputQueue API from the NDK.
Any alphanumeric keycodes were being mapped to `Unidentified` Keys
which meant even crude keyboard input support was broken.
We do need to expose `getUnicodeChar` (or the ability to look
up characters based on the current character map and modifiers) but
for now we should at least map alphanumeric keycodes to `Key::Character`
for basic interim support of virtual keyboards.
This moves all the keycode mapping into a separate `keycodes.rs` file
to reduce clutter.
This adds back the mapping from Android key codes to Winit key codes
that we had before the keyboard API overhaul.
Android activity does expose scan codes but key codes currently seem
like the more appropriate mapping to Winit physical key codes.
This removes the gnarly, unsafe cfg() guarded digging into
'native-activity' and 'game-activity' specific implementation details. I
never intended to expose these details in the public API and really
hope to avoid there being a release of Winit that would depend on this.
I'm also hoping/considering if I can get away with sealing this without
necessarily requiring a semver breaking release of android_activity
since this absolutely should never have been possible, and can probably
safely assume this was the only code in the wild that has briefly done
this.
I'm also a bit unclear as to what led to doing this. There is a
`.key_code()` and `.scan_code()` getter and we even already accessed the
keycode in the Android backend so I'm not sure how those APIs were missed.
The correct handling of this setting requires to change the events
we're getting from the macOS on the fly and call `interpretKeyEvents`,
which could affect handling of the next events, meaning that we can't
provide them on `KeyEvent`.
Add named variants for physical back and forward keys which could
be found on some mice. The macOS bits may not work on all the
hardware given that apple doesn't directly support such a thing.
Co-authored-by: daxpedda <daxpedda@gmail.com>
The use of `Filter` was confusing so it was removed inverting the
behavior of the enum and methods using it.
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
Overhaul the keyboard API in winit to mimic the W3C specification
to achieve better crossplatform parity. The `KeyboardInput` event
is now uses `KeyEvent` which consists of:
- `physical_key` - a cross platform way to refer to scancodes;
- `logical_key` - keysym value, which shows your key respecting the
layout;
- `text` - the text produced by this keypress;
- `location` - the location of the key on the keyboard;
- `repeat` - whether the key was produced by the repeat.
And also a `platform_specific` field which encapsulates extra
information on desktop platforms, like key without modifiers
and text with all modifiers.
The `Modifiers` were also slightly reworked as in, the information
whether the left or right modifier is pressed is now also exposed
on platforms where it could be queried reliably. The support was
also added for the web and orbital platforms finishing the API
change.
This change made the `OptionAsAlt` API on macOS redundant thus it
was removed all together.
Co-authored-by: Artúr Kovács <kovacs.artur.barnabas@gmail.com>
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
Co-authored-by: daxpedda <daxpedda@gmail.com>
Fixes: #2631.
Fixes: #2055.
Fixes: #2032.
Fixes: #1904.
Fixes: #1810.
Fixes: #1700.
Fixes: #1443.
Fixes: #1343.
Fixes: #1208.
Fixes: #1151.
Fixes: #812.
Fixes: #600.
Fixes: #361.
Fixes: #343.
This update rewrites the winit's Wayland backend using new wayland-rs
0.30 API. This fixes long standing issue with the forward compatibility
of the wayland backend, meaning that future updates to the wayland
protocol won't break rust code anymore. like it was before when adding
new shm/enum variants into the protocol.
Fixes#2560.
Fixes#2164.
Fixes#2128.
Fixes#1760.
Fixes#725.
This adds an ability to control left and right `Option` keys to be
treated as `Alt`, thus not producing diacritical marks.
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
The docs state that it accepts `PATH`, but not like the env variable.
So to make it work each `PATH` should be emmited from each `println!`.
Fixes#2657.
Makes WindowAttributes public and adds window_attributes() getter to
WindowBuilder.
In version 0.27, the WindowAttributes struct was made private, but this
removed the ability to introspect the default WindowBuilder values.
* Use a bit less `unsafe` on iOS
I did test this in XCode 11.3's "Debug View Heirarchy", the NSStringRust problem is no longer applicable (likely because Rust got better at emitting correct debug info).
* Avoid using `id` on iOS
This commit fixes it, by not updating the `latest_error` when
any of the hooks handled the error, otherwise it'd interfere
with the winit's error checking.
When other application minimized the winit window the
minimize state was going out of sync. This commit fixes
it by polling the state in `set_minimized`.
* Add Redox OS support
* Simplify control flow usage
* Apply more recommendations
* Update naming to indicate that Orbital is a platform
* Adjust import order
* feat: add pointer events to web
* feat: remove PointerType for touch events
* Remove duplicate
* Changelog and features
* Remove PointerType
* feat: renamed events, added touch type guard
* Rename
* Flip the y axis
* Fix physical position and add force
* Update comment
* Update features
* Use normalized force
* Remove unnecessary todos
* Update comment
* Refactor add touch_handler
* Rephrase by Liamolucko
* Update CHANGELOG.md
* Fix duplicate mouse and touch events
* Removed workaround for scale factor
* Flip the y axis
* Fix
* Fmt
* Replace `match` with a single pattern with `if let`
* Update documentation
* Have one callback per event
* Remove a comment
* Fix
* Remove y-axis flip
* Update src/event.rs
Co-authored-by: Mads Marquart <mads@marquart.dk>
* Fix platform specific comment
* Fix extra argument to `touch_position` function
Co-authored-by: Dany Sluijk <me@dany.dev>
Co-authored-by: Johan Klokkhammer Helsing <johanhelsing@gmail.com>
Co-authored-by: oscrim <oscar@widefind.se>
Co-authored-by: Mads Marquart <mads@marquart.dk>
* fix clippy lints on Windows
* fix lints on other platforms
* a couple more
* again
* don't know what's goging on anymore
* fix examples
* comon
* how about now?
* this is getting annoying
* hmmm
* explicitly set a type
* 😢
* don't cast on x64 targets
* apply code review requests
* fix attributes on expressions
* fix ios
* On macOS, add `WindowBuilderExtMacOS::with_parent_window`
* Replace Parent with Option<Id<NSWindow, Shared>>
* Add addChildWindow method on NSWindow instead
* Update with_parent_window to be unsafe fn
* Add unified `with_parent_window`
* Remove `WindowBuilderExtUnix::with_parent`
* Remove `WindowBuilderExtWindows::with_parent_window`
* Clean up CI warnings
* Update CHANGELOG.md
It's `WindowBuilderExtX11` rather than `WindowBuilderExtUnix`
* Rename parent to owner
* Make with_parent_window unsafe and update its doc
* Add another way to get window on mac
* Add more documentations
* Add match arm and panic on invalid varients
* Add Xcb arm
* Update child_window example to make it safer and work in i686
* Remove duplicate entry in CHANGELOG.md
* Propogate error instead of expect
* Replace unreachable to panic
* Add platform note to X11
Co-authored-by: Wu Yu Wei <wusyong9104@gmail.com>
* Close windows synchronously on main thread
* Set style mask synchronously on main thread
* Set title synchronously on main thread
* Set visibility and focus synchronously on main thread
* Set window level synchronously on main thread
* Set position and size synchronously on main thread
* Set cursor hittest synchronously on main thread
* Add changelog entry
This fixes a crash on macOS when trying to get the monitor
refresh rate from the disabled monitor.
Co-authored-by: Jet Spark <lixiaopeng.jetspark@bytedance.com>
Co-authored-by: Mads Marquart <mads@marquart.dk>
* On Windows and macOS, add API to enable/disable window controls
* fix build
* missing import
* use `WindowButtons` flags
* rename to `[set_]enabled_buttons`
* add example, fix windows impl for minimize
* macOS: Fix button enabling close/minimize while disabling maximized
* Update src/platform_impl/windows/window.rs
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
* compose the flags on a sep line, use `bool::then`
Co-authored-by: Mads Marquart <mads@marquart.dk>
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
This adds `Window::set_window_level` to control the preferred
z level of the window.
Co-authored-by: Markus Siglreithmaier <m.siglreith@gmail.com>
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
Co-authored-by: Mads Marquart <mads@marquart.dk>
* ci: Don't use `$CMD` for Android doc building
Since migrating `cargo-apk` to `clap` [it is now annoying] to pass
unknown arguments to an underlying `cargo` command (like `cargo doc`):
fortunately generating docs doesn't need to go through `cargo apk` to
set up cross-compiler/linker environment variables at all.
[it is now annoying]: https://github.com/rust-windowing/android-ndk-rs/pull/363
* ci: Simplify
* ci: Explicitly build just the `winit` package on Android
Since https://github.com/dvc94ch/cargo-subcommand/pull/23 `cargo-apk`
now strictly searches for workspaces first before committing to finding
the right package _within said workspace_, and bails when no package was
selected since we don't support selecting (building, packaging, running)
>1 target currently.
Perhaps it's a bit hash to enforce this on free-form `cargo apk --`
invocations, but it is what it is.
* Only build, but don't run tests in MSRV CI
Since the MSRV of development dependencies can easily be bumped without it affecting the MSRV of the published version of `winit`
* Run clippy on stable Rust instead of MSRV Rust
clippy inspects the `rust-version` field, and only suggests changes that conform to that.
This updates the Android backend to use the android-activity crate instead
of ndk-glue. This solves a few issues:
1. The backend is agnostic of the application's choice of Activity base
class
2. Winit is no longer responsible for handling any Java synchronization
details, since these are encapsulated by the design of
android_activity
3. The backend no longer depends on global / static getters for state
such as the native_window() which puts it in a better position to
support running multiple activities within a single Android process.
4. Redraw requests are flagged, not queued, in a way that avoids taking
priority over user events (resolves#2299)
To make it possible for application crates to avoid explicitly
depending on the `android-activity` crate (and avoid version conflicts)
this re-exports the android-activity crate under:
`winit::platform::android::activity::*`
This also adds `android-native-activity` and `android-game-activity`
features that set the corresponding android-activity features.
Addresses: PR https://github.com/rust-windowing/winit/pull/1892
Addresses: PR https://github.com/rust-windowing/winit/pull/2307
Addresses: PR https://github.com/rust-windowing/winit/pull/2343
Addresses: #2293Resolves: #2299
Co-authored-by: Markus Siglreithmaier <m.siglreith@gmail.com>
Co-authored-by: Markus Siglreithmaier <m.siglreith@gmail.com>
Even when the protocol explicitly tells to send proper UTF-8
boundaries for cursor, some IMEs don't do that, so sanity check
them before sending downstream.
* MacOS: set value for `accepts_first_mouse`
* Update CHANGELOG and FEATURES
* Field doesn't need to be public
* Convert `bool` to `BOOL`
* Fix formatting
* Move flag from window state to view instance
* Feedback from PR
* Fix changelog location
Introduced in https://github.com/rust-windowing/winit/pull/2479; turns out the definitions were not entirely equal, the `kCFRunLoopEntry` that we were using previously was defined as `0` while the correct value is `1` (which meant the `unimplemented!()` branch suddenly started triggering)
Use the definitions that `core_foundation` exposes (almost the same, except `CFRunLoopSourceContext::perform` is not nullable, so we account for that as well).
* Remove UnownedWindow::inner_rect
* Refactor custom view to use much less `unsafe`
The compiler fence is safe to get rid of now since `interpretKeyEvents` takes `&mut self`
* Refactor Window to use much less unsafe
* Refactor NSApplication usage to have much less unsafe
* Remove cocoa dependency
* Enable `deny(unsafe_op_in_unsafe_fn)` on macOS
Also re-enable clippy `let_unit_value` lint
* Remove #[macro_use] on macOS
* Refactor window delegate to use much less unsafe
* Disable scrolling on web by default but provide method in builder to enable it
* rename enable_web_scroll -> enable_web_page_scroll
* move enable_web_page_scroll into prevent_default option
* final approach
* Mark prevent_default change as breaking
Co-authored-by: Mads Marquart <mads@marquart.dk>
* Begin abstraction over AppKit
* Clean up NSApplication delegate declaration
* Clean up NSApplication override declaration
* Clean up NSWindow delegate declaration
* Clean up NSWindow override declaration
* Clean up NSView delegate declaration
- Pass WM_SYSKEYDOWN to DefWindowProc
- Avoid intercepting WM_SYSCHAR to allow ALT+Space to work: removes ReceivedCharacter events for alt+keypress
- Intercept WM_MENUCHAR to disable bell sound
On Waylnad when asking for redraw before `MainEventsCleared`
would result for redraw being send on the next event loop tick,
which is not expectable given that it must be delivered on the same
event loop tick.
This makes Winit 0.27 compatible with crates like Wgpu 0.13 that are
using the raw_window_handle v0.4 crate and aren't able to upgrade to 0.5
until they do a new release (since it requires a semver change).
The change is intended to be self-contained (instead of pushing
the details into all the platform_impl backends) since this is only
intended to be a temporary trait implementation for backwards
compatibility that will likely be removed before the next Winit release.
Fixes#2415.
* Add CODEOWNERS file
This makes it very clear when you're stepping down from the post as a maintainer, and makes it clear for users who is expected to review their PR
* Fix grammar
* Make @kchibisov receive pings for the X11 platform
* Fix typo
Co-authored-by: Kirill Chibisov <contact@kchibisov.com>
To be more consistent with mobile platforms this updates the Windows,
macOS, Wayland, X11 and Web backends to all emit a Resumed event
immediately after the initial `NewEvents(StartCause::Init)` event.
The documentation for Suspended and Resumed has also been updated
to provide general recommendations for how to handle Suspended and
Resumed events in portable applications as well as providing
Android and iOS specific details.
This consistency makes it possible to write applications that lazily
initialize their graphics state when the application resumes without
any platform-specific knowledge. Previously, applications that wanted
to run on Android and other systems would have to maintain two,
mutually-exclusive, initialization paths.
Note: This patch does nothing to guarantee that Suspended events will
be delivered. It's still reasonable to say that most OSs without a
formal lifecycle for applications will simply never "suspend" your
application. There are currently no known portability issues caused
by not delivering `Suspended` events consistently and technically
it's not possible to guarantee the delivery of `Suspended` events if
the OS doesn't define an application lifecycle. (app can always be
terminated without any kind of clean up notification on most
non-mobile OSs)
Fixes#2185.
Co-authored-by: Marijn Suijten <marijns95@gmail.com>
Co-authored-by: Markus Røyset <maroider@protonmail.com>
When confirming input in e.g. Korean IME or using characters like
`+` winit was sending those twice, once via `Ime::Commit` and the
other one via `ReceivedCharacter`, since those events weren't generating
any `Ime::Preedit` and were forwarded due to `do_command_by_selector`.
This should force the use of system libraries for Fontconfig
and freetype instead of building them with cmake if missing.
This also fixes compilation failures on nightly.
Fixes#2373.
winit's notion of "focus" is very simple; you're either focused or not.
However, Windows has both notions of focused window and active window
and paying attention only to WM_SETFOCUS/WM_KILLFOCUS can cause a window
to believe the user is interacting with it when they're not. (this
manifests when a user switches to another application between when a
winit application starts and it creates its first window)
* web: add `with_prevent_default`, `with_focusable`
`with_prevent_default` controls whether `event.preventDefault` is called
`with_focusable` controls whether `tabindex` is added
Fixes#1768
* Remove extra space from CHANGELOG
This applies https://github.com/rust-windowing/android-ndk-rs/issues/117
on the `winit` side: Android destroys its window/surface as soon as the
user returns from [`onNativeWindowDestroyed`], and we "fixed" this on
the `ndk-glue` side by sending the `WindowDestroyed` event before
locking the window and removing it: this lock has to wait for any user
of `ndk-glue` - ie. `winit` - to give up its readlock on the window,
which is what we utilize here to give users of `winit` "time" to destroy
any resource created on top of a `RawWindowHandle`.
since we can't pass the user a `RawWindowHandle` through the
`HasRawWindowHandle` trait we have to document this case explicitly and
keep the lock alive on the `winit` side instead.
[`onNativeWindowDestroyed`]: https://developer.android.com/ndk/reference/struct/a-native-activity-callbacks#onnativewindowdestroyed
* web: Add `EventLoop::spawn`
This is the same as `EventLoop::run`, but doesn't throw an exception in order to return `!`.
I decided to name it `spawn` rather than `run_web` because I think that's more descriptive, but I'm happy to change it to `run_web`.
Resolves#1714
* Update src/platform/web.rs
Co-authored-by: Markus Røyset <maroider@protonmail.com>
* Fix outdated names
Co-authored-by: Markus Røyset <maroider@protonmail.com>
Make sure `cargo doc` runs cleanly without any warnings in the CI - some
recently introduced but still allowing a PR to get merged.
In case someone wishes to add docs on private items, make sure those
adhere to the same standards.
Android also supports `EventLoopExtRunReturn`. The user will still have
to follow the README to turn this example into a `cdylib` and add the
`ndk_glue::main()` initialization attribute, though.
This commit renames `Window::set_cursor_grab` to
`Window::set_cursor_grab_mode`. The new API now accepts enumeration
to control the way cursor grab is performed. The value could be: `lock`,
`confine`, or `none`.
This commit also implements `Window::set_cursor_position` for Wayland,
since it's tied to locked cursor.
Implements API from #1677.
This is required to help hardware accelerated libraries like glutin
that accept WindowBuilder instead of RawWindowHandle, since the api
to access builder properties directly was removed.
Follow up to 44288f6.
* Remove redundant documentation links
* Add note to README about windows not showing up on Wayland
* Fix documentation links
* Small documentation fixes
* Add note about doing stuff after StartCause::Init on macOS
Commit f10a984 added `EventLoopWindowTarget::set_device_event_filter`
with for a mutable reference, however most winit APIs work with
immutable references, so altering API to play nicely with existing APIs.
This also disables device event filtering on debug example.
The `ndk` crate now implements [`HasRawWindowHandle` directly on
`NativeWindow`], relieving the burden to reimplement it on `winit`.
[`HasRawWindowHandle` directly on `NativeWindow`]: https://github.com/rust-windowing/android-ndk-rs/pull/274
Previously on X11, by default all global events were broadcasted to
every winit application. This unnecessarily drains battery due to
excessive CPU usage when moving the mouse.
To resolve this, device events are now ignored by default and users must
manually opt into it using
`EventLoopWindowTarget::set_filter_device_events`.
Fixes (#1634) on Linux.
This reverts commit 78e5a395da.
It was discovered that in some cases mesa will lock the back
buffer, e.g. when making context current, leading to resize
missing. Given that applications can restructure their rendering
to account for that, and that winit isn't limited to playing
nice with mesa reverting the original commit.
Since there's a "Platform-specific" header, it makes sense to put the
Linux-specific part under it. On the other hand, "Can only be called on
the main thread." is true for all platforms, not just iOS, so there is
no reason to call it out for iOS specifically.
When the window switches mode from normal to tabbed one, it doesn't
get resized, however the frame gets resized. This commit makes
winit to track resizes when frame changes instead of window.
Fixes#2191.
While most compositors provide server side decorations, the GNOME
does not, and won't provide them. Also Wayland clients must render
client side decorations.
Winit was already drawing some decorations, however they were bad
looking and provided no text rendering, so the title was missing.
However this commit makes use of the SCTK external frame similar to
GTK's Adwaita theme supporting text rendering and looking similar to
other GTK applications.
Fixes#1967.
This commit brings new Ime event to account for preedit state of input
method, also adding `Window::set_ime_allowed` to toggle IME input on
the particular window.
This commit implements API as designed in #1497 for desktop platforms.
Co-authored-by: Artur Kovacs <kovacs.artur.barnabas@gmail.com>
Co-authored-by: Markus Siglreithmaier <m.siglreith@gmail.com>
Co-authored-by: Murarth <murarth@gmail.com>
Co-authored-by: Yusuke Kominami <yukke.konan@gmail.com>
Co-authored-by: moko256 <koutaro.mo@gmail.com>
`wl_pointer::set_cursor` expects a serial number of the last
`wl_pointer::enter` event. However other calls expect latest
observed pointer serial, so this commit tracks both and
use them as required by specification.
Fixes#2273.
This makes X11 and Wayland follow Windows and macOS, so the size of the
window could be set even though it has resizable attribute set to false.
Fixes#2242.
Both APIs are used to set application name. This commit unifies the API
between Wayland and X11, so downstream applications can remove platform
specific code when using `WindowBuilderExtUnix`.
Fixes#1739.
When all the receive from the compositor is `TouchEvent::Down` and
`TouchEvent::Up` for the same id, we would record the touch position in
the touch_points vector the first time. On `TouchEvent::Up` we'd find it
and report `TouchPhase::Ended` with that location. The next time we
receive `TouchEvent::Down` for the same id, we'd however unconditionally
append a new `TouchPoint` to `inner.touch_points`, with the new
position. On release however we'd find the earlier point and report its
location, which basically means that `TouchPhase::Ended` always and
forever reported the location of the very first touch down event.
Instead, this patch updates an existing touch point location with the
same id on `TouchDown`.
Fixes#1996
The commit without buffer attached should be done after setting all
top level size/application related properties. While c4df7ad7a added
commit after setting decorations, it accounted just for decorations,
however we should account for the min/max size and other attributes.
This commit brings `is_resizable` and `is_decorated`. Since the client
is responsible for both of them, they could be tracked without deep
syncing with server.
The scale factor being sent when the only monitor is disconnected and
reconnected is hard coded to 1.0. That may work by chance, if that's the
scale factor in use currently, but it does not work in the general case.
As a result, clients may end up with wrongly scaled or laid out window
contents after reconnect, as was reported over in
https://github.com/alacritty/alacritty/issues/5703, for example.
The problem was introduced by change 125ee0b, which caused an additional
ScaleFactorChanged event to be sent on monitor reconnect, but got the
scale factor wrong when the only monitor is disconnected and
reconnected.
This change fixes the problem by using the current monitor's scale
factor in this case. The event is still being sent as intended by
125ee0b.
Fixes#2123.
* Use NSView's inputContext instead of creating our own
This means that `set_ime_position` now properly invalidates the character coordinates.
* Make `set_ime_position` robust against moving windows
If you try to use `EventLoop::run_return` API in a way that you do on
demand polling of events it won't actually poll, since in such strategy
the `ControlFlow::Exit` is sent right before Wayland backend starts to
poll.
This was observed with smithay compositor Anvil which was doing this
particular thing leading to GNOME thinking that app isn't responding,
due to connection not being polled.
* Improve web example
* Implement basic logger into the example webpage
* Repace bash script with xtask
* replace wasm-bindgen-cli with wasm-bindgen-cli-support
* refactor
* Move logic into external crate.
* Remove CI changes
* Review feedback
This commit adds an `EventLoopBuilder` struct to simplify event loop
customization and providing options to it upon creation. It also
deprecates the use of `EventLoop::with_user_event` in favor of the same
method on new builder, and replaces old platforms specific extension
traits with the new ones on the `EventLoopBuilder`.
* Add TraceGuard to make tracing simpler
* Add SharedStateMutexGuard to make tracing simpler
* Add trace_scope macro
* Add missing let binding in trace_scope!
* Update changelog guidelines to prevent conflicts from blocking PRs
As per consensus in https://github.com/rust-windowing/winit/issues/2135
* Add note about whitespace in changelog
* Add note about maintainer creating new tag
* Add exit code to control flow and impl on linux
* Fix examples to have an exit code
* Fix doc examples to use an exit code
* Improve documentation wording on the exit code
* Add exit code example
* Add exit code on windows
* Change i32 as exit code to u8
This avoids nasty surprises with negative numbers on some unix-alikes
due to two's complement.
* Fix android usages of ControlFlow::Exit
* Fix ios usages of ControlFlow::Exit
* Fix web usages of ControlFlow::Exit
* Add macos exit code
* Add changelog note
* Document exit code on display server disconnection
* Revert "Change i32 as exit code to u8"
This reverts commit f88fba0253.
* Change Exit to ExitWithCode and make an Exit const
* Revert "Add exit code example"
This reverts commit fbd3d03de9.
* Revert "Fix doc examples to use an exit code"
This reverts commit daabcdf9ef.
* Revert "Fix examples to have an exit code"
This reverts commit 0df486896b.
* Fix unix-alike to use ExitWithCode instead of Exit
* Fix windows to use ExitWithCode rather than Exit
* Silence warning about non-uppercase Exit const
* Refactor exit code handling
* Fix macos Exit usage and recover original semantic
* Fix ios to use ExitWithCode instead of Exit
* Update documentation to reflect ExitWithCode
* Fix web to use ExitWithCode when needed, not Exit
* Fix android to use ExitWithCode, not Exit
* Apply documenation nits
* Apply even more documentation nits
* Move change in CHANGELOG.md under "Unreleased"
* Try to use OS error code as exit code on wayland
The repository is configured such that specific checks are required to pass before merging; since CI doesn't run on PRs that don't change code, they can't be merged without administrator intervention.
Also, while this definition is currently correct, we might in the future change something so that changes to other files (like markdown files) really should be run through CI.
And example of that could be:
```rust
#[cfg(doctest)]
#[doc = include_str!("../README.md")]
extern "C" {}
```
See also https://github.com/rust-windowing/winit/pull/1626.
The `cocoa` crate links to AppKit, which made the symbol `CGDisplayCreateUUIDFromDisplayID` from ApplicationServices/ColorSync (which AppKit uses internally) available to us on macOS 10.8 to 10.13.
However, this does not work on macOS 10.7 (where AppKit does not link to ColorSync internally). Instead of relying on this, we should just link to ApplicationServices directly.
* Add cursor grab
* Update feature matrix, changelog and platform information
* Add proper error propagation
* Remove "expect" from fallible code
code would crash if handling pointer capture outside of winit on
every mouse click
we swallow the error since we could not think of a case where this
would fail in a way that would want to handle that it fails
* Remove unnecessary implementation comment
Co-authored-by: Will Crichton <wcrichto@cs.stanford.edu>
Maybe the transparent setting in WM_NCCREATE on Windows 11 will cause a block when calling DwmEnableBlureBehindWindow and will crash. Puts them into WM_CREATE and it works.
This reverts commit 8afeb910bd.
Reverting because this change made Pinyin input unusable
(only latin characters showed even after selecting the
desired Chinese character)
* examples: Fix unused `Result` that must be used when initializing console_log
* examples: Fix unused imports
* Fix unread name field warning in linux x11 ime InputMethod struct
* Fix unread name field warning in linux x11 Device struct
* Ignore unread field warning in macos/ios MonitorHandle struct
* ci: Add `--deny warnings` to `RUSTFLAGS`
* In MacOS, only disable menu bar in exclusive fullscreen
* Save and restore fullscreen options in set_fullscreen
* Don't always cache presentation options when entering exclusive fullscreen
This commit caches presentation options when entering exclusive fullscreen
only if we're coming from borderless fullscreen.
Then, when transitioning from exclusive -> borderless, if no cached presentation
options are present, then the default borderless options are applied.
This fixes the menu bar being unavailable when taking the following path:
[not fullscreen] -> [exclusive fullscreen] -> [borderless fullscreen].
Without this commit, the presentation options from [not fullscreen] were being
cached and then applied to [borderless fullscreen].
* Restore the window level when switching to exclusive fullscreen
The hack of using `CGShieldingWindowLevel() + 1` in borderless fullscreen needs
to be undone when switching from [borderless] -> [exclusive] fullscreen,
otherwise there are menu bar glitches when following a path through
[borderless] -> [exclusive] -> [borderless] modes.
Now, this might appear to conflict with the 'always on top' feature which uses
the 'floating window' level, but this feature appears to be broken anyway when
entering and exiting fullscreen with an always-on-top window. So, rather than
introducing logic to attempt to restore to the 'floating' level here, I think
it's better to do the simple thing for now and then introduce logic for
always-on-top windows when fixing the overall fullscreen behaviour.
* Update the changelog
Co-authored-by: Ehden Sinai <ehdens@gmail.com>
Co-authored-by: Francesca Lovebloom <francesca@brainiumstudios.com>
* macOS: Ignore all events while in the callback
Previously all native dialogs, such as [rfd](https://github.com/PolyMeilex/rfd), would cause the event loop (event_loop.run) to freeze.
* Update changelog
* spelling
* Fix merge mistake
* Add link to issue in the code
Co-authored-by: Poly <marynczak.bartlomiej@gmail.com>
While winit was always using dlopen for opening system libs, it
provides a way now to disable dlopen feature helping with linking
on some targets.
Fixes#2037.
Some video drivers could set display metrics to odd values, which can result in
extra large scale factors (e.g. winit is sending 720 for 5k screen on nvidia
binary drivers), so let's just drop them to prevent clients from using them.
The value 20 was picked, because the DPR for 8k @ 5 inch is ~18.36.
Fixes#1983.
* Fix X11 memory leak and remove mio-misc
I also fixed a couple of clippy lints.
Fixes#1984
* Send the redraw event before waking up the main event
* Use .map instead of a match, and remove comments saved by git
* Remove unnecessary pub keywords on `WakeSender` in x11/mod.rs
When disconnect the only monitor, scale factor is reset to 1.0. We need
to set it back when the monitor is reconnected.
We previously assume current window must be on an existing monitor, but
that's not true in case of reconnecting the only one monitor.
This commit also drops 'Theme' trait with its support types
in favor of 'FallbackFrame' meaning that winit will use some
predefined frame for the time being, since porting 'ConceptFrame'
will require adding font rendering librarires right into winit,
which is not desired.
Fixes#1889.
The inability to set the cursor using any of the named cursor files
likely indicates an error in the system on which we are running.
'WinitPointer::set_cursor' also does not have any way of returning an
error so just log the error to assist users in diagnosing the problem.
Fixes: #1988.
* Adds Android winit<->ndk_glue version match table
* Fixes justification
* Adds crosses
* Address review and instead of a m:n table it shows a 1:1 version compatibility
* Addresses review
* Enable to show text when IME is active
* Remove unnecessary variable
* Enable to use IME
* fmt
* Remove println! for debug
* Fix handling of utf-8 string
* clear_marked_text should be rust function, not member function
* Store state information in ViewState
* Remove unnecessary function
* format
* Remove mut
* format
* Remove duplicate marked text
* Remove unused `is_preediting` field
Co-authored-by: Artúr Kovács <kovacs.artur.barnabas@gmail.com>
* Remove window subclassing
* Always call `DefWindowProcW` when we don't process a message
* Improve window initialization
Note that the error path in `init` is kind of cursed at the moment.
* Rename `ThreadMsgTargetCallbackData` to `ThreadMsgTargetData`
* Simplify window initialization
* Fix compilation on 32-bit targets
* Simplify the creation of the event target window
* Use `.clone()` rather than `Rc::clone()`
* Use concrete types for args to `SetWindowLongPtrW`
On 32-bit targets, `SetWindowLongPtrW` is an alias to `SetWindowLongW`,
which returns `LONG` (`i32`) rather than `LONG_PTR` (`isisze`).
* Minor comment adjustments
Fixes#1488.
`is_key_down` was only set to true when `insertText` was called.
Therefore `is_key_down` was actually meant to store whether the most
recently pressed key generated an `insertText` event, at which, winit
produces a `ReceivedCharacter`. The issue is that `insertText` is *not*
called for "key repeat", but winit wants to send `ReceivedCharacter`
events for "key repeat" too. To solve this, the `is_key_down` variable
was then checked in the `key_down` function to determine whether it was
valid to produce repeated `ReceivedCharacter` events during "key
repeat". However this check is not needed since `ReceivedCharacter` must
always be called if the key event has a non-empty `characters` property.
Furthermore `is_key_down` didn't actually store whether the previously
pressed character had an `insertText` event, because it was incorrectly
set to false on every "key up". This meant that if two keys were pressed
consecutively and then the first was released, then `is_key_down` was
set to false even if the most recent keypress did actually produce an
`insertText`.
Update changelog.
* changed description of window scale factor in Events
* Update src/dpi.rs
Co-authored-by: Markus Røyset <maroider@protonmail.com>
* Fix intra-doc link
Co-authored-by: Markus Røyset <maroider@protonmail.com>
* Drop the event callback before exiting
* Update the changelog
* Apply suggestion from review
Co-authored-by: Markus Røyset <maroider@protonmail.com>
* Apply review suggestions
Co-authored-by: Markus Røyset <maroider@protonmail.com>
* Remove support for `stdweb`
* Expunge `stdweb`; make `web-sys` the default
* Mark this change as a breaking change
* Re-insert accidental removal of space
* Use the correct cargo feature syntax
* Re-add some `cfg` attributes
* Remove `web-sys` feature from CI
* Require setting the activation policy on the event loop
* Run cargo fmt
* Update changelog
* Fixes and tweaks from review
* Correct comment in app_state.rs
Co-authored-by: Mads Marquart <mads@marquart.dk>
* MacOS: Only activate after the application has finished launching
This fixes the main menu not responding until you refocus, at least from what I can tell - though we might have to do something similar to https://github.com/linebender/druid/pull/994 to fix it fully?
* MacOS: Remove activation hack
* Stop unnecessarily calling `makeKeyWindow` on initially hidden windows
You can't make hidden windows the key window
* Add new, simpler activation hack
For activating multiple windows created before the application finished launching
* feat: added MacOS menu
* fix: ran fmt
* extracted function into variable
* idiomatic formatting
* Set the default menu only during app startup
* Don't set the activation policy in the menu init
Co-authored-by: Artur Kovacs <kovacs.artur.barnabas@gmail.com>
This is called internally by NSApplication.run, and is not something we should call - I couldn't find the reasoning behind this being there in the first place, git blame reveals c38110cac from 2014, so probably a piece of legacy code.
Removing this fixes creating new windows when you have assigned a main menu to the application.
We allow to have RunLoop running only on the main thread. Which means if
we call Window::request_redraw() from other the thread then we have to
wait until some other event arrives on the main thread. That situation
is even worse when we have ControlFlow set to the `Wait` mode then user
will not ever render anything.
* Fix for #1850
* Update changelog
* Fix for compilation warnings
* Apply suggestions from code review
Co-authored-by: Markus Røyset <maroider@protonmail.com>
* Improve code quality
* Change Arc<Mutex> to Rc<RefCell>
* Panicking in the user callback is now well defined
* Address feedback
* Fix nightly warning
* The panic info is now not a global.
* Apply suggestions from code review
Co-authored-by: Francesca Lovebloom <francesca@brainiumstudios.com>
* Address feedback
Co-authored-by: Markus Røyset <maroider@protonmail.com>
Co-authored-by: Francesca Lovebloom <francesca@brainiumstudios.com>
* Restore the ability to have fully transparent windows on Windows
Besides its original purpose, commit 6343059b "Fix Windows transparency
behavior to support fully-opaque regions (#1621)" also included some
changes considered cleanups, one of them was:
* Remove the `CreateRectRgn` call, since we want the entire window's region to
have blur behind it, and `DwnEnableBlurBehindWindow` does that by default.
But the original code actually disabled the blur effect for the whole
window by creating an empty region for it, because that allows for the
window to be truely fully transparent. With the blur effect in place,
the areas meant to be transparent either blur the things behind it
(until Windows 8) or are darkened (since Windows 8). This also means
that on Windows 8 and newer, the resulting colors are darker than
intended in translucent areas when the blur effect is enabled.
This restores the behaviour from winit <0.24 and fixes#1814.
Arguably, one might want to expose the ability to control the blur
region, but that is outside the scope of this commit.
* Remove useless WS_EX_LAYERED from transparent windows on Windows
`WS_EX_LAYERED` is not supposed to be used in combination with
`CS_OWNDC`. In winit, as it is currently used, `WS_EX_LAYERED` actually
has no effect at all. The only relevant call is to
`SetLayeredWindowAttributes`, which is required to make the window
visible at all with `WS_EX_LAYERED` set, but is called with full
opacity, i.e. there's no transparency involved at all.
The actual transparency is already achieved by using
`DwmEnableBlurBehindWindow`, so `WS_EX_LAYERED` and the call to
`SetLayeredWindowAttributes` can both be removed.
* Add DeviceEvent::MouseMove on web platform to support pointer lock
* Update changelog
* Add support for stdweb too
* Add mouse_delta to stdweb
* Remove reference to pointer lock
* Link CGDisplayCreateUUIDFromDisplayID through ColorSync instead of CoreGraphics
* Conditionally link through ColorSync only if WINIT_LINK_COLORSYNC is set
to true
* Document new macos env var in README
Following the changes in [1] this bumps ndk and ndk-glue to 0.3 and uses
the new constants. The minor version has been bumped to prevent
applications from running an older winit (without #1826) with a newer
ndk/ndk-glue that does not pass this `ident` through the `data` pointer
anymore.
[1]: https://github.com/rust-windowing/android-ndk-rs/pull/112
The issue was caused by calling SetCapture on a window which already
had the capture. The WM_CAPTURECHANGED handler assumed that it would
only run if the capture was lost, but that wasn't the case. This made
the handler to try to lock the window state mutex while it was already
locked.
The issue was introduced in #1797 (932cbe4).
ndk-glue currently sets both the `ident` field and user-data pointer to
`0` or `1` for the event pipe and input queue respectively, to tell
these sources apart. While it works to reinterpret this `data` pointer
as integer identifier it shouldn't be abused for that, in particular
when one may wish to provide extra information with an event in the
future; then the `data` field is used as pointer (or abused as abstract
value) for that.
This commit forwards "unknown" Wayland mouse buttons downstream via
'MouseButton::Other'. Possible values for those could be found in
<linux/input-event-codes.h>.
Also, since Wayland just forwards buttons from the kernel, which are
'u16', we must adjust 'MouseButton::Other' to take 'u16' instead of
'u8'.
This commit introduces a cross platform way to request a user attention
to the window via a 'request_user_attention' method on a Window struct.
This method is inspired by macOS's 'request_user_attention' method and
thus reuses its signature and semantics to some extent.
* Fix WindowEvent::ReceivedCharacter on web
The event was never sent to the application because of the unconditional
preventDefault() call on keydown.
Fixes#1741
* Don't scroll when pressing space on a focused canvas
After reaching keypress, we should prevent further propagation.
Relates to #1741
Aborting compilation by using 'compile_error!' macro in build.rs was resulting in failing cross
compilation, thus this commit removes build.rs. The compilation will now be aborted on existing 'compile_error!' macros in corresponding platform sources.
Building window with 'set_min_inner_size' was setting 'max_inner_size'
under the hood, thus completely disabling window resize, since
the window isn't resizeable on Wayland when its minimum size
is equal to its maximum size.
This patch removes an unneeded workaround for transparent windows on the
Windows platform. In addition, it simplifies a couple of related API calls:
* Remove the `CreateRectRgn` call, since we want the entire window's region to
have blur behind it, and `DwnEnableBlurBehindWindow` does that by default.
* Remove the `color_key` for `SetLayeredWindowAttributes`, since it's not used
(we're not passing `winuser::LWA_COLORKEY` to the flags).
The initial media query that's used to watch for device pixel ratio
changes should match. Unfortunately it doesn't with Safari and the
corresponding assertion fails. This is because resolution is not a
supported support property for queries. As a workaround, this patch
extends the query to optionally match for the webkit specific DPR
property directly.
This fixes#1734
* Update SCTK to 0.11.0
Updates smithay-client-toolkit to 0.11.0. The major highlight
of that updated, is update of wayland-rs to 0.27.0. Switching
to wayland-cursor, instead of using libwayland-cursor. It
also fixes the following bugs:
- Disabled repeat rate not being handled.
- Decoration buttons not working after tty switch.
- Scaling not being applied on output reenable.
- Crash when `XCURSOR_SIZE` is `0`.
- Pointer getting created in some cases without pointer capability.
- On kwin, fix space between window and decorations on startup.
- Incorrect size event when entering fullscreen when using
client side decorations.
- Client side decorations not being hided properly in fullscreen.
- Size tracking between fullscreen/tiled state changes.
- Repeat rate triggering multiple times from slow callback handler.
- Resizable attribute not being applied properly on startup.
- Not working IME
Besides those fixes it also adds a bunch of missing virtual key codes,
implements proper cursor grabbing, adds right click on decorations
to open application menu, disabled maximize button for non-resizeable
window, and fall back for cursor icon to similar ones, if the requested
is missing.
It also adds new methods to a `Theme` trait, such as:
- `title_font(&self) -> Option<(String, f32)>` - The font for a title.
- `title_color(&self, window_active: bool) -> [u8; 4]` - The color of
the text in the title.
Fixes#1680.
Fixes#1678.
Fixes#1676.
Fixes#1646.
Fixes#1614.
Fixes#1601.
Fixes#1533.
Fixes#1509.
Fixes#952.
Fixes#947.
This changes 'Fullscreen::Borderless' enum variant from
'Fullscreen::Borderless(MonitorHandle)' to
'Fullscreen::Borderless(Option<MonitorHandle>)'. Providing
'None' to it will result in picking the current monitor.
* web: Allow event to be queued from inside the EventLoop handler
The Runner is behind a RefCell, which is mutably borrowed when the event
handler is being called. To queue events, `send_events` needs to check
`is_closed()` and the `is_busy` flag, but it cannot be done since the
RefCell is already locked. This commit changes the conditions to work
without needing a successful borrow.
* web: Emit WindowEvent::Resized on Window::set_inner_size
* Update changelog
* web-sys: Impl. event listeners removal for canvas
* web-sys: Impl. media query listeners cleanup
* web: Emit WindowEvent::Destroyed after Window is dropped
* web-sys: Fix unload event closure being dropped early
* web: Impl. cleanup on ControlFlow::Exit
- Drops the Runner, which causes the event handler closure to be
dropped.
- (web-sys only:) Remove event listeners from DOM.
* web: Do not remove canvas from DOM when dropping Window
The canvas was inserted by the user, so it should be up to the user
whether the canvas should be removed.
* Update changelog
This commit is a follow up to a2db4c0a32
to make it clear which virtual key codes are located on numeric pad.
It also adds Asterisk and Plus virtual key codes.
Certain platforms like Wayland don't have a concept of
primary Monitor in particular. To indicate that
'primary_monitor' will return 'None' as well as in cases
where the primary monitor can't be detected.
Fixes#1683.
On certain platforms window couldn't be on any monitor
resulting in failures of 'current_monitor' function.
Such issue was happening on Wayland, since the window
isn't on any monitor, unless the user has drawn something into it.
Returning 'Option<MonitorHandle>' will give an ability to
handle such situations gracefully by properly indicating that
there's no current monitor.
Fixes#793.
In winit the swipe from left to right on touchpad should
generate positive horizontal delta change, however on
macOS it was the other way around without
natural scrolling.
This commit inverses the horizontal scrolling delta
in 'MouseScrollDelta' events to match other platforms.
Fixes#1695.
* Change web backend `event_handler` to without 'static lifetime
* Refactor web runner and fix ControlFlow::Exit not being sticky
* Impl. scaling change event for web-sys backend
* Improve `dpi` docs regarding the web backend
* Add changes to changelog
* Update features.md
On all platforms other than Linux/X11, the Subtract key was uniformly
used only for the Numpad. To make this cross-platform compatible, the
`-` key will now map to `Minus` on X11 instead of `Subtract`.
Since people have been confused about the difference between `Minus` and
`Subtract` in the past, the `Subtract` key has also been renamed to
`NumpadSubtract`. This is a breaking change that might be annoying to
downstream since there's no direct improvement, but it should help new
users in the future. Alternatively this could just be documented, rather
than explicitly mentioning the Numpad in the name.
* Impl. mouse capturing for web-sys with PointerEvent
* Impl. mouse capturing for web-sys with MouseEvent by manual tracking
* Reorganize web-sys backend mouse and pointer handling code
* Impl. mouse capturing for stdweb with PointerEvent
* Add mouse capturing for web target to changelog
The `_lwp_self` function cannot be used to reliably determine the main
thread, see
https://github.com/alacritty/alacritty/issues/2631#issuecomment-676723289.
It might always be equal to the PID, but it's certainly not always 1
when the thread is the main thread.
However, Rust's built in `Thread::id` and `Thread::name` function will
always return `ThreadId(1)` and `Some("main")`. Since converting the
thread's ID to a number is not supported on stable Rust, checking that
the thread is labeled `Some("main")` seems like the most reliable
option. It should also be a good fallback in general.
PhysicalSize is recorded as canvas.size, whereas LogicalSize is stored
as canvas.style.size.
The previous cursor behavior on stdweb clobbered all style - thus losing
the LogicalSize.
This resolves various problems with the documentation about platform
support on the Window struct.
It also completely removes pointless runtime errors in favor of
consistent no-ops on all platforms that do not support a certain
features.
* macos: Support click-dragging out of a window
* macos: Use NSEvent::pressedMouseButtons for click-dragging
* macos: Click-dragging: Move pressedMouseButtons inside
* Fix for fullscreen with run_return on mac
* Cleanup
* Removed a comment
* fmt
* This doesn't break exiting run_return anymore
* Now you can also transition from code
* Fmt & cleanup
* Now using a atomic instead of a static bool
* reinserted a line
* Fmt
* Added support for dialogs and child windows
* Cargo fmt
* Dialogs are now being shutdown properly
* Cargo fmt
* Update CHANGELOG.md
* Report mouse motion before click
This fixes an issue on macOS where a mouse click would be generated,
without ever getting a mouse motion to the position before the click.
This leads to the application thinking the mouse click occurred at a
position other than the actual mouse location.
This happens due to mouse motion above the window not automatically
giving focus to the window, unless it is actually clicked, making it
possible to move the window without motion events.
Fixes#942.
* Add additional mouse motion events
Co-authored-by: Ryan Goldstein <ryan@ryanisaacg.com>
* Use requestAnimationFrame for polling wasm
* Implement `requestAnimationFrame` for stdweb
Co-authored-by: Ryan G <ryanisaacg@users.noreply.github.com>
x11-dl was using std::mem::uninitialized incorrectly and when
rustlang added MaybeUninit and intrinsic panics on UB caused
by improper use of uninitialized (see rust-lang/rust/pull/69922)
it caused issues with X11 initialization. x11-dl pr
erlepereira/x11-rs/pull/101 updated x11-dl to use MaybeUninit
correctly
There are two PRs I'm aware of that should be relatively trivial to get
merged, which would fix some issues. Other than those, I don't think it
makes sense to wait on anything.
- Fix Windows crash: https://github.com/rust-windowing/winit/pull/1459
- Fix macOS mouse reports: https://github.com/rust-windowing/winit/pull/1490
While #1459 seems pretty essential to actually make winit run, #1490 is
much less important and can probably be ignored if there aren't any
resources to merge it.
This reverts commit 9daa0738a9.
This commit introduced other bug #1453 with likely much more common bindings,
so reverting it for now.
Fixes#1453.
Co-authored-by: Osspial <osspial@gmail.com>
* On Windows, fix request_redraw() related panics
These panics were introduced by 6a330a2894
Fixes https://github.com/rust-windowing/winit/issues/1391
Fixes https://github.com/rust-windowing/winit/issues/1400
Fixes https://github.com/rust-windowing/winit/issues/1466
Probably fixes other related issues
See https://github.com/rust-windowing/winit/issues/1429
* On Windows, replace all calls to UpdateWindow by calls to InvalidateRgn
This avoids directly sending a WM_PAINT message,
which might cause buffering of RedrawRequested events.
We don't want to buffer RedrawRequested events because:
- we wan't to handle RedrawRequested during processing of WM_PAINT messages
- state transitionning is broken when handling buffered RedrawRequested events
Fixes https://github.com/rust-windowing/winit/issues/1469
* On Windows, panic if we are trying to buffer a RedrawRequested event
* On Windows, move modal loop jumpstart to set_modal_loop() method
This fixes a panic.
Note that the WM_PAINT event is now sent to the modal_redraw_method
which is more correct and avoids an unecessary redraw of the window.
Relates to but does does not fix https://github.com/rust-windowing/winit/issues/1484
* On Window, filter by paint messages when draining paint messages
This seems to prevent PeekMessage from dispatching unrelated sent messages
* Change recently added panic/assert calls with warn calls
This makes the code less panicky...
And actually, winit's Windoww callbacks should not panic
because the panic will unwind into Windows code.
It is currently undefined behavior to unwind from Rust code into foreign code.
See https://doc.rust-lang.org/std/panic/fn.catch_unwind.html
* add comments to clarify WM_PAINT handling in non modal loop
* made redraw_events_cleared more explicit and more comments
* Remove assertions from Windows dark mode code
In general, winit should never assert on anything unless it means that
it is impossible to continue the execution of the program. There are
several assertions in the Windows dark mode code where this is not the
case.
Based on surface level inspection, all existing assertions could be
easily replaced with just simple conditional checks, allowing the
execution of the program to proceed with sane default values.
Fixes#1458.
* Add changelog entry
* Format code
* Pass dark mode by mutable reference
* Format code
* Return bool instead of mutable reference
* Fix dark mode success reply
Co-Authored-By: daxpedda <daxpedda@gmail.com>
* Fix dark mode success reply
* Replace magic integers with constants
Co-authored-by: daxpedda <daxpedda@gmail.com>
* Move `ModifiersChanged` variant to `WindowEvent`
* macos: Fix flags_changed for ModifiersChanged variant move
I haven't look too deep at what this does internally, but at least
cargo-check is fully happy now. :)
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* macos: Fire a ModifiersChanged event on window_did_resign_key
From debugging, I determined that macOS' emission of a flagsChanged
around window switching is inconsistent. It is fair to assume, I think,
that when the user switches windows, they do not expect their former
modifiers state to remain effective; so I think it's best to clear that
state by sending a ModifiersChanged(ModifiersState::empty()).
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* windows: Fix build
I don't know enough about the code to implement the fix as it is done on
this branch, but this commit at least fixes the build.
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS
Very similar to the changes made in [1], as focus is lost, send an event
to the window indicating that the modifiers have been released.
It's unclear to me (without a Windows device to test this on) whether
this is necessary, but it certainly ensures that unfocused windows will
have at least received this event, which is an improvement.
[1]: f79f21641a
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* macos: Add a hook to update stale modifiers
Sometimes, `ViewState` and `event` might have different values for their
stored `modifiers` flags. These are internally stored as a bitmask in
the latter and an enum in the former.
We can check to see if they differ, and if they do, automatically
dispatch an event to update consumers of modifier state as well as the
stored `state.modifiers`. That's what the hook does.
This hook is then called in the key_down, mouse_entered, mouse_exited,
mouse_click, scroll_wheel, and pressure_change_with_event callbacks,
which each will contain updated modifiers.
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* Only call event_mods once when determining whether to update state
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* flags_changed: Memoize window_id collection
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* window_did_resign_key: Remove synthetic ModifiersChanged event
We no longer need to emit this event, since we are checking the state of
our modifiers before emitting most other events.
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* mouse_motion: Add a call to update_potentially_stale_modifiers
Now, cover all events (that I can think of, at least) where stale
modifiers might affect how user programs behave. Effectively, every
human-interface event (keypress, mouse click, keydown, etc.) will cause
a ModifiersChanged event to be fired if something has changed.
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* key_up: Add a call to update_potentially_stale_modifiers
We also want to make sure modifiers state is synchronized here, too.
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* mouse_motion: Remove update_potentially_stale_modifiers invocation
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* Retry CI
* ViewState: Promote visibility of modifiers to the macos impl
This is so that we can interact with the ViewState directly from the
WindowDelegate.
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* window_delegate: Synthetically set modifiers state to empty on resignKey
This logic is implemented similarly on other platforms, so we wish to
regain parity here. Originally this behavior was implemented to always
fire an event with ModifiersState::empty(), but that was not the best as
it was not necessarily correct and could be a duplicate event.
This solution is perhaps the most elegant possible to implement the
desired behavior of sending a synthetic empty modifiers event when a
window loses focus, trading some safety for interoperation between the
NSWindowDelegate and the NSView (as the objc runtime must now be
consulted in order to acquire access to the ViewState which is "owned"
by the NSView).
Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com>
* Check for modifiers change in window events
* Fix modifier changed on macOS
Since the `mouse_entered` function was generating a mouse motion, which
updates the modifier state, a modifiers changed event was incorrectly
generated.
The updating of the modifier state has also been changed to make sure it
consistently happens before events that have a modifier state attached
to it, without happening on any other event.
This of course means that no `CursorMoved` event is generated anymore
when the user enters the window without it being focused, however I'd
say that is consistent with how winit should behave.
* Fix unused variable warning
* Move changelog entry into `Unreleased` section
Co-authored-by: Freya Gentz <zegentzy@protonmail.com>
Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com>
Co-authored-by: Christian Duerr <contact@christianduerr.com>
User can switch between Wait, WaitUntil and Poll modes with key '1', '2' and '3' respectivly.
User can toggle request_redraw calls with the 'R' key.
Helpful for testing all control flow modes and use of request_redraw.
This restores default portable 'C' locale when target locale is unsupported
by X11 backend (Xlib).
When target locale is unsupported by X11, some locale-dependent Xlib
functions like `XSetLocaleModifiers` fail or have no effect triggering
later failures and panics.
When target locale is not valid, `setLocale` should normally leave the
locale unchanged (`setLocale` returns 'C'). However, in some situations,
locale is accepted by `setLocale` (`setLocale` returns the new locale)
but the accepted locale is unsupported by Xlib (`XSupportsLocale` returns
`false`).
Fix#636
* On Wayland, fix coordinates in touch events when scale factor isn't 1
* Explicitly state that Wayland is using LogicalPosition internally
* Fix CHANGELOG
* Remove Wayland theme intermediates
This removes the intermediate struct for passing a Wayland theme to
allow the user direct implementation of the trait.
By passing the trait directly, it is possible for downstream users to
have more freedom with customization without relying on winit to offer
these options as fields.
It should also make maintenance easier, since winit already doesn't
implement all the functions which are offered by the smithay client
toolkit.
* Reimplement SCTK's Theme and ButtonState
* Fix style issues
* Remove public signature
* Format code
* Add change log entry
Co-authored-by: Murarth <murarth@gmail.com>
The current implementation of the event loop runner has some significant
problems. It can't handle multiple events being emitted at once (for
example, when a keyboard event causes a key input, a text input, and a
modifier change.) It's also relatively easy to introduce bugs for the
different possible control flow states.
The new model separates intentionally emitting a NewEvents (poll
completed, wait completed, init) and emitting a normal event, as well as
providing a method for emitting multiple events in a single call.
* Don't discard high-precision cursor position data
Most platforms (X11, wayland, macos, stdweb, ...) provide physical
positions in f64 units, which can contain meaningful fractional
data. For example, this can be empirically observed on modern X11
using a typical laptop touchpad. This is useful for e.g. content
creation tools, where cursor motion might map to brush strokes on a
canvas with higher-than-screen resolution, or positioning of an object
in a vector space.
* Update CHANGELOG.md
Co-Authored-By: Murarth <murarth@gmail.com>
Co-authored-by: Murarth <murarth@gmail.com>
* Fix bug causing RedrawRequested events to only get emitted every other iteration of the event loop.
* Initialize simple_logger in examples.
This PR's primary bug was discovered because a friend of mine reported
that winit was emitting concerning log messages, which I'd never seen
since none of the examples print out the log messages. This addresses
that, to hopefully reduce the chance of bugs going unnoticed in the
future.
* Add changelog entry
* Format
* Update the DPI module docs
* Fix HiDpiFactorChanged doc link
* Incorporate lokathor and icefox feedback
* Adjust documented desktop resolution range
* X11 is one of the reasons I use Windows
* Address DPI generics and float->int rounding
* Revise DPI value statement to better reflect best practices
* Address some of freya's feedback
* phrasing
* Rephrase X11 DPI stuff
* Rename hidpi_factor to scale_factor
* Deprecate WINIT_HIDPI_FACTOR environment variable in favor of WINIT_X11_SCALE_FACTOR
* Rename HiDpiFactorChanged to DpiChanged and update docs
I'm renaming it to DpiChanged instead of ScaleFactorChanged, since I'd
like Winit to expose the raw DPI value at some point in the near future,
and DpiChanged is a more apt name for that purpose.
* Format
* Fix macos and ios again
* Fix bad macos rebase
* fix: use a 'static lifetime for the web backend's `Event` types
* implement hidpi for stdweb (web-sys wip?)
* fix: make all canvas resizes go through backend::set_canvas_size
* update Window docs for web, make `inner/outer_position` return the position in the viewport
* Begin implementing DPI generics
* Fix multithreaded example
* Format
* Fix serde test
* hopefully fix most of the errors
* Fix dpi module errors
* More error fixings
* Format
* fix macos errors
* Another error pass
* Replace bad type signatures
* more fixins
* WIP - Make EL2 DPI changes and implement on Windows (#895)
* Modify DPI API publicly and on Windows
* Add generic Position and make dpi creation functions const
* Make examples work
* Fix fullscreen windows not appearing
* Replace Logical coordinates in window events with Physical coordinates
* Update HiDpiFactorChanged
* Document to_static
* On Windows, make AdjustRect calls DPI-aware when possible (#1015)
* Use AdjustWidowRectExForDPI when available
* Prioritize presevering logical size when handling WM_DPICHANGED
* Format
* Add changelog entry
* macOS: Dpi overhaul (#997)
* WIP - Make EL2 DPI changes and implement on Windows (#895)
* Modify DPI API publicly and on Windows
* Add generic Position and make dpi creation functions const
* Make examples work
* Fix fullscreen windows not appearing
* Replace Logical coordinates in window events with Physical coordinates
* Update HiDpiFactorChanged
* Document to_static
* fix app_state errors
* fixes hidpi related errors in window_delegate
* fix bad merge
* dpi_factor edits in window_delegate
* fixes type and lifetime errors in window and window_delegate
* applies fmt
* complies with @aleksijuvani requested changes
* modifies Handler lifetimes
* fixes lifetime isues, adds propper handling for HiDpiChanged
* applies fmt
* restore original lifetimes
* solution is somewhere out there
* applies fmt
* pass as references
* resolves issue with HANDLER
* crate visible type error
* fixes visibility issues
* applies fmt
* deals with warnings
* simplifies new_inner_size setting algorthm
* moves proxy instead of referencing it and removes double deref from proxy.ns_window
* makes @Osspial tests (https://github.com/rust-windowing/winit/pull/997\#discussion_r301852354) pass
* complies with @aleksijuvani suggested changes
* makes max window size std::f32::MAX
* On Windows, fix new DPI API not setting window size properly (#1130)
* First attempt
* Second attempt
* Maintain cursor horizontal ratio
* Fix DPI change handling when maximized
* Revert window example
* Make new DPI code more understandable
* Format
* Implement DPI Usability Upgrades for X11 and Wayland (#1098)
* Fix compile errors
* Use `mio` for the X11 event loop
* Removes `calloop` from the X11 event loop, as the method of draining a
source using a closure provided to the `calloop::EventLoop` instance
conflicts with the need to deliver events directly to the callback
provided to `EventLoop::run`, in order to respond to the value provided by
`WindowEvent::HiDpiFactorChanged`.
* Implement interactive `HiDpiFactorChanged` event for X11
* Implement interactive `HiDpiFactorChanged` event for Wayland
* Run cargo fmt
* Fix Wayland not processing events from EventQueue
* Backport #981
* some lifetime tinkering
* finishes lifetime tinkering
* fixes all type errors
* adds support ffi functions
* adds wrappers for nonstatic events
* replaces events with event wrappers
* reimplementing hidpichanged event in app_state
* implements HiDpiFactorChanged for iOS
* applies formatter
* complies with @aleksijuvani requested changes
* resolves conflicts
* applies fmt
* removes merge blurp
* corrects state of CHANGELOG
* fix fmt check error
* fixes hidpi_factor for armv7-apple-ios
* WIP - Make EL2 DPI changes and implement on Windows (#895)
* Modify DPI API publicly and on Windows
* Add generic Position and make dpi creation functions const
* Make examples work
* Fix fullscreen windows not appearing
* Replace Logical coordinates in window events with Physical coordinates
* Update HiDpiFactorChanged
* Document to_static
* fix app_state errors
* fixes hidpi related errors in window_delegate
* fix bad merge
* dpi_factor edits in window_delegate
* fixes type and lifetime errors in window and window_delegate
* applies fmt
* complies with @aleksijuvani requested changes
* modifies Handler lifetimes
* fixes lifetime isues, adds propper handling for HiDpiChanged
* applies fmt
* restore original lifetimes
* solution is somewhere out there
* applies fmt
* pass as references
* resolves issue with HANDLER
* crate visible type error
* fixes visibility issues
* applies fmt
* deals with warnings
* simplifies new_inner_size setting algorthm
* moves proxy instead of referencing it and removes double deref from proxy.ns_window
* makes @Osspial tests (https://github.com/rust-windowing/winit/pull/997\#discussion_r301852354) pass
* complies with @aleksijuvani suggested changes
* makes max window size std::f32::MAX
Changes from rebasing:
* fixes compile errors
* applies fmt
* reimplements HiDpiFactorChanged after #1173 merge
* uses EventWrappers
* Fix compile errors
* Use `mio` for the X11 event loop
* Removes `calloop` from the X11 event loop, as the method of draining a
source using a closure provided to the `calloop::EventLoop` instance
conflicts with the need to deliver events directly to the callback
provided to `EventLoop::run`, in order to respond to the value provided by
`WindowEvent::HiDpiFactorChanged`.
* Implement interactive `HiDpiFactorChanged` event for X11
* Implement interactive `HiDpiFactorChanged` event for Wayland
* Run cargo fmt
* Fix Wayland not processing events from EventQueue
* Backport #981
* First attempt
* Second attempt
* Maintain cursor horizontal ratio
* Fix DPI change handling when maximized
* Revert window example
* Make new DPI code more understandable
* Format
* Modify DPI API publicly and on Windows
* Add generic Position and make dpi creation functions const
* Make examples work
* Fix fullscreen windows not appearing
* Replace Logical coordinates in window events with Physical coordinates
* Update HiDpiFactorChanged
* Document to_static
* Added some "how" and "why" docs to event handling.
Basically I had these questions when I started exploring the new
event API's, and as I figured out the answers I put down more info
about how everything works. This is not final, and suggestions
are welcome -- the code example in the `event` module docs is
particularly dubious, but it's how I'm used to thinking abou things
so it only made sense to me once I wrote that.
Note that my bias is towards using winit for writing games, so that's
the sort of things I was interested in. This may not be valid for
more general use cases.
* cargo fmt
* Fix minor typos
* Revise event documentation
* Update lib.rs docs
* Update root docs
Co-authored-by: Osspial <osspial@gmail.com>
* On X11, make `WINIT_HIDPI_FACTOR` dominate `Xft.dpi` in some cases
This commit makes `WINIT_HIDPI_FACTOR` dominate `Xft.dpi` in general and
adds a special value `0` for `WINIT_HIDPI_FACTOR` to use winit computed
DPI factor with randr over Xft.dpi.
* Use `randr` instead of `0` for auto dpi scaling
* Update CHANGELOG
* blow up on wrong env var
* Allow empty string for env var
* Move DeviceEvent handling to the message target window.
Previously, device events seem to have only been sent to one particular
window, and when that window was closed Winit would stop receiving
device events. This also allows users to create windowless event loops
that process device events - an intriguing idea, to say the least.
* Emit LWin and RWin VirtualKeyCodes on Windows
* Implement ModifiersChanged on Windows
* Make ModifiersChanged a tuple variant instead of a struct variant
* Add changelog entries
* Format
* Update changelog entry
* Fix AltGr handling
* Reformat
* Publicly expose ModifiersChanged and deprecate misc. modifiers fields
* Change ModifiersState to a bitflags struct
* Make examples work
* Add modifier state methods
* all things considered, only erroring out in one file throughout all of these changes is kinda impressive
* Make expansion plans more clear
* Move changelog entry
* Try to fix macos build
* Revert modifiers println in cursor_grab
* Make serde serialization less bug-prone
* X11: Sync key press/release with window focus
* When a window loses focus, key release events are issued for all pressed keys
* When a window gains focus, key press events are issued for all pressed keys
* Adds `is_synthetic` field to `WindowEvent` variant `KeyboardInput`
to indicate that these events are synthetic.
* Adds `is_synthetic: false` to `WindowEvent::KeyboardInput` events issued
on all other platforms
* Implement windows focus key press/release on Windows
* Docs
Co-authored-by: Murarth <murarth@gmail.com>
* Register windowWillExitFullScreen
* On macOS: Do not toggle fullscreen during fullscreen transition
* Add CHANGELOG
Co-authored-by: Freya Gentz <zegentzy@protonmail.com>
* Add support for Windows Dark Mode
* Add is_dark_mode() getter to WindowExtWindows
* Add WindowEvent::DarkModeChanged
* Add support for dark mode in Windows 10 builds > 18362
* Change strategy for querying windows 10 build version
* Drop window state before sending event
Co-Authored-By: daxpedda <daxpedda@gmail.com>
* Change implementation of windows dark mode support
* Expand supported range of windows 10 versions with dark mode
* Use get_function! macro where possible
* Minor style fixes
* Improve documentation for ThemeChanged
* Use `as` conversion for `BOOL`
* Correct CHANGELOG entry for dark mode
Co-authored-by: daxpedda <daxpedda@gmail.com>
Co-authored-by: Osspial <osspial@gmail.com>
* Implement revamped `RedrawRequested` on iOS
* Added RedrawEventsCleared to events_cleared logic
* Fixed from comments
* Added RedrawEventsCleared to draw_rect handler.
* Fixed out of order `RedrawEventsCleared` events.
* cargo fmt
* Implement changes to `RedrawRequested` event
Implements the changes described in #1041 for the X11 platform and for
platform-independent public-facing code.
* Fix `request_redraw` example
* Fix examples in lib docs
* Only issue `RedrawRequested` on final `Expose` event
Mutter can reposition window on resize, if it is behind mutter's "bounding box".
So, when you start winit window in maximized mode with CSD, mutter places its CSD
behind the GNOME's status bar initially, and then sends configure with the
exact same size as your current window. If winit decides to optimize calling
frame.resize(..) in this case, we won't call set_geometry(we're calling
it through resize) and GNOME won't reposition your window to be inside the
"bounding box", which is not a desired behavior for the end user.
* Add ModifiersChanged event for macOS
This implements the macOS portion of #1124.
* Fix ModifiersChanged event import
* Fix event passing window instead of device id
* X11: Report `CursorMoved` when touch event occurs
* Only trigger CursorMoved events for the first touch ID
* Fix testing for current touch events
* Fix first touch logic
* X11: Sync key press/release with window focus
* When a window loses focus, key release events are issued for all pressed keys
* When a window gains focus, key press events are issued for all pressed keys
* Adds `is_synthetic` field to `WindowEvent` variant `KeyboardInput`
to indicate that these events are synthetic.
* Adds `is_synthetic: false` to `WindowEvent::KeyboardInput` events issued
on all other platforms
* Clarify code with comments
* X11: Fix incorrect modifiers when events are missed
* Syncs modifier state with state data in X key/button/motion events.
* Fixes modifier state in XWayland, as xinput2 raw input events will
not be received when a window does not have focus.
* Removes `impl From<_> for ModifiersState` on X11/Wayland API types,
replacing them with `pub(crate)` methods.
* Cleanup modifier state update using a macro
* Remove keys from modifier state when updating
* X11: Fix window creation hangs when another application is fullscreen
Previously, the X11 backend would block until a `VisibilityNotify` event
is received when creating a Window that is visible or when calling
`set_visible(true)` on a Window that is not currently visible. This
could cause winit to hang in situations where the WM does not quickly
send this event to the application, such as another window being
fullscreen at the time.
This behavior existed to prevent an X protocol error caused by setting
fullscreen state on an invisible window. This fix instead stores desired
fullscreen state when `set_fullscreen` is called (iff the window is not
visible or not yet visible) and issues X commands to set fullscreen
state when a `VisibilityNotify` event is received through the normal
processing of events in the event loop.
* Add window_debug example to facilitate testing
* Add a CHANGELOG entry
* Call `XUnmapWindow` if `VisibilityNotify` is received on an invisible window
* Prevent EventLoop from getting initialized outside the main thread
This only applies to the cross-platform functions. We expose functions
to do this in a platform-specific manner, when available.
* Add CHANGELOG entry
* Formatting has changed since the latest stable update...
* Fix error spacing
* Unix: Prevent initializing EventLoop outside main thread
* Updates libc dependency to 0.2.64, as required by BSD platforms
* Update CHANGELOG.md for Linux implementation
* Finish sentence
* Consolidate documentation
* Fix so the compiler can infer msg_send! return type
Currently, due to a quirk in Rust's type inference interacting with the
structure of the msg_send! macro, a () return type will be inferred when
the compiler cannot otherwise determine the return type. This behavior
is expected to change, and in the future could resolve to a ! return
type, which results in undefined behavior.
Linting has previously been added for this in rust-lang/rust#39216, but
it did not catch these cases due to SSheldon/rust-objc#62. An upcoming
version of objc will be fixed to stop hiding these errors, at which
point they will become compile errors.
This change fixes these errors and allows winit to compile with the
fixed version of objc.
* Bump cocoa to 0.19.1
* Add web targets to travis
* Fix some bash syntax
* Avoid crash on providing no features
* Fix syntax error
* Syntax?
* Fix pleas
* Fix features check
* Add a comment
* Add nightly builds
Adds fullscreen using native web APIs to the stdweb and web-sys backends.
Due to limitations of browser APIs, requests for fullscreen can only be fulfilled during a short-lived user-triggered event. This commit does automatically handle that under the hood, but it does introduce unavoidable latency in full-screening the canvas.
* Expose HINSTANCE now using a getter function
* Missing changes
* remove unused import
* Required changes for the PR
* Rust fmt
* Use GetWindowLong
* Use GetWindowLong
* [#1111] Use consistent return types for available_monitors()
Always use `impl Iterator<Item = MonitorHandle>` instead of
`AvailableMonitorsIter`. Fix an example that used the Debug
implementation of `AvailableMonitorsIter`.
* [#1111] Update changelog
* [#1111] Remove AvailableMonitorsIter type completely
* [#1111] Remove doc references to AvailableMonitorsIter
* Fixed relative_pointer not being set up when the "zwp_relative_pointer_manager_v1" callback comes after the "wl_seat" callback
* Ran cargo fmt
* Updated changelog
* Added wayland support for set_grab_cursor and set_cursor_visible
* Updated changelog
* Ran cargo fmt
* Fixed set_cursor_visible and set_cursor_grab so they can be called from any thread.
* Ran cargo_fmt
* Improved CHANGELOG
* Added workaround so that when cursor is hidden it takes effect before the cursor enters the surface. Making the cursor visible again still only happens once the cursor re-enters the surface
* Switched to using Rc<RefCell> instead of Arc<Mutex> since all accesses to the relative_pointer_manager_proxy will happen on the same thread.
* Forgot to run cargo fmt
* Switched to using Rc and RefCell instead of Arc and Mutex where applicable.
* Improved comments and documentation relating to changing a hidden cursor back to visible on wayland.
* Wayland: Fixed cursor not appearing immendiately when setting the cursor to visible.
* Forgot to run cargo fmt
* Switched to only storing the pointers in CursorManager as AutoPointer.
* Fixed typo and removed println
* Update CHANGELOG.md
Co-Authored-By: Kirill Chibisov <wchibisovkirill@gmail.com>
* Fixed relative_pointer not being set up when the "zwp_relative_pointer_manager_v1" callback comes after the "wl_seat" callback
* Ran cargo fmt
* Updated changelog
* Improved CHANGELOG
* Switched to using Rc<RefCell> instead of Arc<Mutex> since all accesses to the relative_pointer_manager_proxy will happen on the same thread.
* Forgot to run cargo fmt
* X11: Fix panic when no monitors are available
* Set dummy monitor's dimensions to `(1, 1)`
* X11: Avoid panicking when there are no monitors in Window::new
* Allow using multiple `XWindowType`s on X11 (#1140)
* Update documentation to make combining window types clearer
* Update build flags because X11 runs on more than just Linux
* Revert "Update build flags because X11 runs on more than just Linux"
This reverts commit 882b910046.
* Revert "Update documentation to make combining window types clearer"
This reverts commit da00ad391a.
* Revert "Allow using multiple `XWindowType`s on X11 (#1140)"
This reverts commit a230333456.
* Allow using multiple `XWindowType`s on X11 (slice variant) (#1140)
* Multiple `XWindowType`s, with non-static lifetime.
* Multiple `XWindowType`s (#1140) (`Vec` variant)
* Append change to changelog.
* Fix formatting.
* Keep track of what windows have requested redraw
Instead of using request_animation_frame and sending redraw request
events, just keep track of all windows that have asked for a redraw.
This doesn't handle dispatching the events
* Issue redraw events to windows that request it
* Cargo fmt
* Use actual numeric IDs to differentiate Windows
This is generally important to identifying which window should
recieve which event, but is also specifically crucial for fixing
RedrawRequested on web.
* Cargo fmt
* Flush high surrogate if not followed by low surrogate
* Remove transmute from WM_CHAR handler
* Fix window_state being locked while dispatching ReceivedCharacter for surrogate codepoints.
* Format
Just before starting to poll/wait on calloop(mio), check if there
are already events pending in the internal buffer of our wayland
event queue. If so, dispatch them and force an instant wakeup from
the polling, in order to behave as if we were instantly woken up by
incoming wayland events.
When using OpenGL, mesa shares our wayland socket, and also reads
from it, especially if vsync is enabled as it'll do blocking reads.
When doing so, it may enqueue events in the internal buffer of our
event queue.
As the socket has been read, mio will thus not notify it to calloop
as read, and thus calloop will not know it needs to dispatch. In some
cases this can lead to some events being delivered much later than
they should. Combined with key repetition this can actually cause some
flooding of the event queue making this effect event worse.
Fixes#1148
* Add touch pressure information for touch events on Windows
* Modified CHANGELOG.md and FEATURES.md to reflect changes
* Updated documentation of struct Touch to reflect changes
* Replaced mem::uninitalized() with mem::MaybeUninit
Fixed warnings in platform_impl/windows/dpi.rs
* fix#1087. the CFRunLoopTimer was never started if the user never changed the controlflow.
* RedrawRequested ordering matches the new redraw api
consistent asserts
lots of appstate refactoring to rely less on unsafe, and hopefully make it easier to maintain
* ios: dpi bugfix. inputs to setContentScaleFactor are not to be trusted as iOS uses 0.0 as a sentinel value for "default device dpi".
the fix is to always go through the getter.
* move touch handling onto uiview
* update changelog
* rustfmt weirdness
* fix use option around nullable function pointers in ffi
* Document why gl and metal views don't use setNeedsDisplay
* change main events cleared observer priority to 0 instead of magic number
log when processing non-redraw events when we expect to only be processing redraw events
* X11: Fix performance issue with rapidly resetting cursor icon
* When setting cursor icon, if the new icon value is the same as the
current value, no messages are sent the X server.
* X11: Cache cursor objects in XConnection
* Add changelog entry
* iOS os version checking
* iOS, fix some incorrect msg_send return types
* address nits, and fix OS version check for unsupported os versions
* source for 60fps guarantee
* macOS/iOS: Fix auto trait impls of `EventLoopProxy`
`EventLoopProxy<T>` allows sending `T` from an arbitrary thread that
owns the proxy object. Thus, if `T` is `!Send`, `EventLoopProxy<T>` must
not be allowed to leave the main thread.
`EventLoopProxy<T>` uses `std::sync::mpsc::Sender` under the hood,
meaning the `!Sync` restriction of it also applies to
`EventLoopProxy<T>`. That is, even if `T` is thread-safe, a single
`EventLoopProxy` object cannot be shared between threads.
* Update `CHANGELOG.md`
* Stop appending canvas to document in web platform
* Remove `tabindex` TODO in web backend
* Return `OsError` instead of panicking on web canvas creation
* Add exclusive fullscreen mode
* Add `WindowExtMacOS::set_fullscreen_presentation_options`
* Capture display for exclusive fullscreen on macOS
* Fix applying video mode on macOS after a fullscreen cycle
* Fix compilation on iOS
* Set monitor appropriately for fullscreen on macOS
* Fix exclusive to borderless fullscreen transitions on macOS
* Fix borderless to exclusive fullscreen transition on macOS
* Sort video modes on Windows
* Fix fullscreen issues on Windows
* Fix video mode changes during exclusive fullscreen on Windows
* Add video mode sorting for macOS and iOS
* Fix monitor `ns_screen` returning `None` after video mode change
* Fix "multithreaded" example on macOS
* Restore video mode upon closing an exclusive fullscreen window
* Fix "multithreaded" example closing multiple windows at once
* Fix compilation on Linux
* Update FEATURES.md
* Don't care about logical monitor groups on X11
* Add exclusive fullscreen for X11
* Update FEATURES.md
* Fix transitions between exclusive and borderless fullscreen on X11
* Update CHANGELOG.md
* Document that Wayland doesn't support exclusive fullscreen
* Replace core-graphics display mode bindings on macOS
* Use `panic!()` instead of `unreachable!()` in "fullscreen" example
* Fix fullscreen "always on top" flag on Windows
* Track current monitor for fullscreen in "multithreaded" example
* Fix exclusive fullscreen sometimes not positioning window properly
* Format
* More formatting and fix CI issues
* Fix formatting
* Fix changelog formatting
* Process WM_SYSCOMMAND to forbid screen savers in fullscreen mode
Fixes#1047
* Update CHANGELOG.md and documentation to reflect changes from issue #1065
* Updated documentation of window.Window.set_fullscreen to match the documentation of window.WindowBuilder.with_fullscreen.
* Touch events emit screen coordinates instead of client coordinates on Windows
Fixes#1002
* Don't lose precision of WM_TOUCH events when converting from screen space to client space
* Updated CHANGELOG.md to reflect changes from issue: #1042
* Replace `std::mem::uninitialized` with `MaybeUninit`
* Avoid undefined behavior when using `MaybeUninit`
* Restore unused `PointerState` fields as internally public
* Zero-initialize some struct values in Xlib FFI calls
* Reform usage of `MaybeUninit` in Xlib FFI
* Prefer safe zero-initialization using `Default`, when possible
* Zero-initialize integers and floats using `0` or `0.0`
* Use `MaybeUninit::uninit` for large byte buffers and union types
* Use `MaybeUninit::uninit` when the resulting value is ignored
* Fix old `use` declarations
* Fix hidden lifetime parameter
* Fix missing methods in `web::Monitor`.
Originally fixed by @ryanisaacg in 94387c4bf5.
* Disable some tests and examples on `wasm32`
* Fix issues with redraw_requested when called during EventsCleared
* Format
* Fix event dispatch after RedrawRequested but before EventsCleared
This could happen if the event queue was cleared, we processed WM_PAINT,
but the event queue got re-filled before we checked to see it was empty.
* Fix paint ordering issues when resizing window
* Format
* Make FileDropHandler::iterate_filenames more robust
by replacing the call to mem::uninitialized with mem::zeroed and change
file name retrieval to use buffers of exact length as reported
by DragQueryFileW instead of relying on MAX_PATH.
* Change remaining calls of uninitialized to zeroed
* Run rustfmt
* Add CHANGELOG entry and comment
Internally, `XFlush` calls `_XSend` to write data. It then calls
`XEventsQueued(display, QueuedAfterReading)`, which reads data from the
X server connection. This prevents the event loop source callback from
being run, as there is no longer data waiting on the socket.
Ideally, we would want to call `_XSend` directly to ensure that no
output is buffered by Xlib. However, this function is not exported as
part of Xlib's public API.
Testing with the `XFlush` call removed does not appear to adversely
affect the performance of an application. If any bugs should eventually
arise from this change, perhaps another function may be used in place of
`XFlush`, such as `XPending`, which writes buffered output but does not
so aggressively read from the X server connection.
Closes#865
* Add relative pointer movement for Wayland
* Format changed code with rustfmt
* Wayland: merge window and device event queues into one
* Replace map_or_else call for simplification
* README: Use shields.io instead of Herokuapp (#859)
* README: Link to FEATURES.md and missing features wiki page (#860)
Closes#854
* Update URLs (#863)
* CHANGELOG.md: Add line from #861 (legacy) that is missing from equivalent #964 (EL 2)
* Format everything and add rustfmt to travis
* Remove extern crate winit from examples and add force_multiline_blocks
* Format the code properly
* Fix inconsistent period in PULL_REQUEST_TEMPLATE.md
* Only run rustfmt on nightly
* Travis fixings
This decorelates the window management from the actual user content,
meaning:
- the created window no longer needs the user to draw something to
start existing
- it reduces our need to do roundtrips during initialization to
avoid protocol errors
None of the platforms that use it are supported right now, so it seems
silly to fail builds for other platforms when these ones can't possibly
succeed.
This should be reverted in stages as the platforms regain support.
* Support listing available video modes for a monitor
* Use derivative for Windows `MonitorHandle`
* Update FEATURES.md
* Fix multiline if statement
* Add documentation for `VideoMode` type
The cursor visibility docs were still outdated from the rename and
talking about setting the invisibility, rather than visibility.
The platform-specific docs are unchanged since those should be fine and
the rest has been adapted using similar docs.
* Fix request_redraw with Poll and WaitUntil(time_in_the_past) on Windows
`Window::request_redraw` now fires a `RedrawRequested` event when
called from an `Event::EventsCleared` callback while the control
flow is set to `Poll`. A control flow of `WaitUntil(resume_time)`,
will now also fire the `RedrawRequested` event when `resume_time`
is in the past.
* Prevent panic on x11 when WaitUntil(resume_time) is in the past
* Prevent panic on wayland when WaitUntil(resume_time) is in the past
* First name consistency pass. More to come!
* Remove multitouch variable (hopefully this compiles!)
* Remove CreationError::NotSupported
* Add new error handling types
* Remove `get_` prefix from getters.
This is as per the Rust naming conventions recommended in
https://rust-lang-nursery.github.io/api-guidelines/naming.html#getter-names-follow-rust-convention-c-getter
* Make changes to Window position and size function signatures
* Remove CreationError in favor of OsError
* Begin updating iOS backend
* Change MonitorHandle::outer_position to just position
* Fix build on Windows and Linux
* Add Display and Error implementations to Error types
* Attempt to fix iOS build.
I can't actually check that this works since I can't cross-compile to
iOS on a Windows machine (thanks apple :/) but this should be one of
several commits to get it working.
* Attempt to fix iOS errors, and muck up Travis to make debugging easier
* More iOS fixins
* Add Debug and Display impls to OsError
* Fix Display impl
* Fix unused code warnings and travis
* Rename set_ime_spot to set_ime_position
* Add CHANGELOG entry
* Rename set_cursor to set_cursor_icon and MouseCursor to CursorIcon
* Organize Window functions into multiple, categorized impls
* Improve clarity of function ordering and docs in EventLoop
* port ios winit el2.0 implementation to the new rust-windowing repo
* unimplemented! => unreachable
trailing comma in CFRunLoopTimerCallback
* implement get_fullscreen
* add iOS specific platform documentation. Add a TODO about how to possibly extend the iOS backend to work have methods callable from more than just the main thread
* assert that window is only dropped from the main thread
* assert_main_thread called from fewer places
The previous attempt to update parking_lot missed the windows platform.
The parking_lot dependency is now no longer specified twice to help
prevent that mistake from happening again.
This includes de-listing @francesca64 from the table, as I suggested. I
realize that this is a controversial decision, and it's not a decision I
make lightly, but I believe I have justification for doing so:
I contacted @francesca64 last month asking her about her inactivity as a
maintainer on this project. She replied on March 26th as follows. For the
sake of her privacy, I've removed removed certain sections of her response,
as it contains some personal details that I'm not comfortable sharing with
the world at large without her explicit permission.
> Hello! Thanks for reaching out!❤
> In short, I've moved on. <removed> ...in November, I was hired by a
> company that recently started using Rust. I'm very happily building
> infrastructure there!
> I'm simply not interested in spending more than 8 hours a day
> programming. You couldn't even pay me to do it! My time is best spent
> going on adventures with my beloved.
> I'll still be active in the Rust ecosystem, but only insofar as the
> company I work for is. We use winit on iOS, Android, and (to a limited
> extent) macOS, so I'll work on those backends as needed.
> Thank you for taking care of winit. I hope you're taking care of
> yourself too; <removed>.
I don't begrudge her for her decision, and others shouldn't either - I
firmly believe that, as this is unpaid, volunteer work, everybody should
have the right to move on when they decide they no longer have the time
or will to contribute.
The exact impliciation of this in regards to her status as maintainer
is open to interpretation. I would argue that this means, should she in
her work stumble upon an issue in any of her listed backends, she would be
willing to submit PRs addressing those issues. However, it also means
that she is not able to put in the time to be active as a maintainer on
those platforms, or review PRs and issues for those platforms.
On March 28th I responded to her email as follows:
> Hey, thanks for responding.
> <removed, personal details>
> In the meanwhile, there are still a few loose ends from your time as
> maintainer that I'd like to get cleaned up. It's okay if you aren't
> going to be spending much time on Winit, but there's still a broad
> assumption that you're able to review PRs for them and that doesn't seem
> to be the case. Would you be able to do a couple things to ease the
> transition to whoever next takes over the macOS, X11, and Android
> backends?
> 1) Submit a PR downgrading yourself from maintainer for macOS, X11, and
> Android, so somebody else can more actively take them over.
> 2) Post your WIP macOS backend for EL2.0 as well as the issues it
> currently has, so whoever next maintains macOS can finish it.
> Going forward, I'd like to reach out to the broader Rust community and
> find more active maintainers so Winit can get to 1.0 and I can mostly
> move on from it.
I recieved no response to that email. On April 4th, I followed up on
that email:
> If you aren't able to act as a maintainer for Winit, and aren't able to
> submit a PR updating your official status as maintainer to reflect
> reality, would you mind if I submitted a PR removing you as maintainer?
> That's not something I want to do since it's a bad image for me, a bad
> image for Winit, and sets an extremely uncomfortable precedent, but I'd
> like to start more aggressive outreach to ensure each backend is less
> dependent on one specific person and I don't want to see new
> contributors pinging you for help when you're unable to provide it.
> If you don't reply by the 11th that's the path I'm going to take, but I
> consider it the nuclear option and I want to avoid invoking it if at all
> possible.
Up to this date (April 13th), I have recieved no response. Given the
amount of time I've given her to respond, as well as her lack of
response, I believe we have the justification to remove her from the
table. Should she show back up again, any clarifications on her status
would be welcome, and she is welcome to submit a PR re-listing herself
on the table with a more accurate description of her current contributor
status. However, once we begin the contributor marketing push discussed
in #830, I don't want new contributors to attempt to ask her questions
on the macOS, X11, or Android backends when she isn't able to give a
response.
-----------------------------------------------------------------------
This PR also introduces HALL_OF_CHAMPIONS.md, which commends the efforts
of former maintainers that have contributed greatly to the Winit project.
This wasn't discussed previously, but I think it's important to recognize
the people that brought us to where we are today. It currently lists
@tomaka and @francesca64, as they are the two individuals I'm aware of
that both deserve such recognition and no longer actively contribute to
Winit, but if there's anybody I missed feel free to suggest them and a
blurb describing their work.
* Add initial draft of SCOPE document
* Rephrase/rename feature tiers
* Rename to FEATURES and add a few annotations
* Fix API Reworks table
* Add more annotations
* Some phrasing
* Split compat matrix into seperate section, to be moved into wiki
* Mention compatibility in CONTRIBUTING
* Remove some discuss annotations
* Apply review changes and rename child windows feature to popup windows
* Update based on discussion
* Add issue for Android HiDPI
* Update FEATURES.md
* Update FEATURES.md
* Update PULL_REQUEST_TEMPLATE.md
* Update PULL_REQUEST_TEMPLATE.md
* Reformat FEATURES.MD
* Remove comments
* Improve formatting and add guide for extending #Features
* Prevent the event loop from pausing after entering modal loop
After clicking the window title bar or border (for a drag or resize),
the event loop pauses until the mouse is moved. This change relays
the WM_NCLBUTTONDOWN message to the dummy window where it queues
a redraw and consumes the message. This effectively jumpstarts
the modal loop and it continues to fire draw requests.
* Handle WM_NCLBUTTONDOWN in public_window_callback instead of relaying.
Relaying the WM_NCLBUTTONDOWN message to the modal window turned out
to be unnecessary.
* Add additional numpad key mappings
Since some platforms have already used the existing `Add`, `Subtract`
and `Divide` codes to map numpad keys, the X11 and Wayland platform has
been updated to achieve parity between platforms. On macOS only the
`Subtract` numpad key had to be added.
Since the numpad key is different from the normal keys, an alternative
option would be to add new `NumpadAdd`, `NumpadSubtract` and
`NumpadDivide` actions, however I think in this case it should be fine
to map them to the same virtual key code.
* Add Numpad PageUp/Down, Home and End on Wayland
XNextEvent will block for input while holding the global Xlib mutex.
This will cause a deadlock in even the most trivial multi-threaded
application because OpenGL functions will need to hold the Xlib mutex
too.
Add EventsLoop::poll_one_event and EventsLoop::wait_for_input to provide
thread-safe functions to poll and wait events from the X11 event queue
using unix select(2) and XCheckIfEvent.
This is a somewhat ugly workaround to an ugly problem.
Fixes#779
* Fix incorrect keycodes when using a non-US keyboard layout.
This commit fixes the issue described in #752, and uses the advised
method to fix it.
* Style fixes
Co-Authored-By: Toqozz <toqoz@hotmail.com>
* Refactoring of macOS `virtualkeycode` fix (#752)
* Applies requested changes as per pull request discussion (#755).
* match unix common API to evl 2.0
* wayland: eventloop2.0
* make EventLoopProxy require T: 'static
* Fix linux build and tests
* wayland: update sctk & small fixes
* Rename EventsLoop and associated types to EventLoop
* Rename WindowEvent::Refresh to WindowEvent::Redraw
* Remove second thread from win32 backend
* Update run_forever to hijack thread
* Replace windows Mutex with parking_lot Mutex
* Implement new ControlFlow and associated events
* Add StartCause::Init support, timer example
* Add ability to send custom user events
* Fully invert windows control flow so win32 calls into winit's callback
* Add request_redraw
* Rename platform to platform_impl
* Rename os to platform, add Ext trait postfixes
* Add platform::desktop module with EventLoopExt::run_return
* Re-organize into module structure
* Improve documentation
* Small changes to examples
* Improve docs for run and run_return
* Change instances of "events_loop" to "event_loop"
* Rename MonitorId to MonitorHandle
* Add CHANGELOG entry
* Improve WaitUntil timer precision
* When SendEvent is called during event closure, buffer events
* Fix resize lag when waiting in some situations
* Update send test and errors that broke some examples/APIs
* Improve clarity/fix typos in docs
* Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events.
* Fix crash when running in release mode
* Remove crossbeam dependency and make drop events work again
* Remove serde implementations from ControlFlow
* Fix 1.24.1 build
* Fix freeze when setting decorations
* Replace &EventLoop in callback with &EventLoopWindowTarget
* Document and implement Debug for EventLoopWindowTarget
* Fix some deadlocks that could occur when changing window state
* Fix thread executor not executing closure when called from non-loop thread
* Fix buffered events not getting dispatched
* Fix crash with runner refcell not getting dropped
* Address review feedback
* Fix CHANGELOG typo
* Catch panics in user callback
* Add support for generating dummy DeviceIDs and WindowIDs
* Fix linux
* Improve docs and move dummy to unsafe
* Strengthen guarantees a bit
* Add backticks to CHANGELOG.md
Co-Authored-By: Xaeroxe <xaeroxe@amethyst-engine.org>
* feat: add macos simple fullscreen
* move impl to WindowExt
* feedback: remove warning, unused file and rename param
* feedback: combine fullscreen examples into one example
* fix: ensure decorations and maximize do not toggle while in fullscreen
* fix: prevent warning on non-macos platforms
* feedback: make changelog more explicit
* fix: prevent unconditional construction of NSRect
* fix: don't try to set_simple_fullscreen if already using native fullscreen
* fix: ensure set_simple_fullscreen plays nicely with set_fullscreen
* fix: do not enter native fullscreen if simple fullscreen is active
This commit restricts an Xfwm4-specific DPI-preserving hack to Xfwm4
only. The hack saves and restores the DPI-adjusted size until the actual
size matches. On tiling WMs like i3 this fails, since the size is
constrained by the layout. This in turn causes a never-ending
XResizeWindow vs. XConfigureWindow fight between the WM and the client,
making the WM, winit client, and Xorg consume all CPU cycles available.
* Fix resize border appearing in some cases after leaving fullscreen.
* On fullscreen, save client rect instead of window rect
* Add CHANGELOG entry
* Revert test changes to fullscreen example
* Update panic message when unable to get client area
* Move the event managent to the closure
In preparation of more events not relating to the SeatManager being
captured.
* Handle wl_output remove events
In some cases, wl_outputs can be removed without the compositor
notifying the surfaces using leave/enter events. This breaks the DPI and
resize stuff since the windows' list of monitors were not updated.
Now, wl_output removals are handled and windows are updated accordingly.
* Add changelog entry for disappearing wl_outputs
* Clearer changelog message for wl_output removal changes
* Fix panic when dragging text onto a window on Windws (#697)
* Changed `panic` to `debug` (log) when unknow error occurs in `GetData` while processing a drag-drop / hover event. Plus added appropriate cursor effect if hovered item can not be processed.
* Improved code clarity.
* Add documentation to clarify behaviour of `DroppedFile`, `HoveredFile`, and `HoveredFileCancelled`
* Add period at the end of sentences in documentation.
Motivation:
This allows VirtualKeyCode variants to be stored in a BTreeSet.
Unlike HashSets, BTreeSets implement Ord and Hash, allowing them to be
keys in a {Hash|BTree}Maps. This is nice, e.g. when implementing
keyboard shortcuts functionality, which maps a set of pressed keys to
some action.
Fixes#691
Dropping a window before running the `EventsLoop` results in events
still being queued when `XDestroyWindow` is called, so events like
`XI_Enter` (the culprit in this case) will still be processed.
Simply checking that the window still exists before calling
`query_pointer` was enough to solve the problem.
On Mojave, views automatically become layer-backed shortly after being added to
a window. Changing the layer-backedness of a view breaks the association
between the view and its associated OpenGL context. To work around this, on
Mojave we explicitly make the view layer-backed up front so that AppKit doesn't
do it itself and break the association with its context.
This was breaking the `window` example in `glutin`.
* Add optional serde feature
* Document features in README
* Add changelog entry
* Implement some missing derivable traits
* Add changelog entry for std derives
* Remove extraneous space on serde doc comments
* Add period to end of serde line in readme
* Remove serde impls from WindowAttributes
* Add serde impls for TouchPhase
* Add serde test file
* Add feature lines to testing CIs
* Remove WindowAttributes from changelog
* Implement HoveredFile and HoveredFileCancelled on Windows (#448)
* Update CHANGELOG.
* Applied code organizational corrections and fixed IDropHandler leak on window destroy.
* Moved FileDropHandle to a separate file.
* x11: compute resize logical size with new dpi
Whenever a dpi change occurs, trigger a Resized event as well with the
new logical size. Given X11 primarily deals in physical pixels, a change
in DPI also changes the logical size (as the physical size remains
fixed).
* Doc tweaks
* Windows: Fix window.set_maximized()
* Add window.set_maximized fix to CHANGELOG.
* Windows: use same style for set_maximized as other set_x(bool) methods
* Added support for F16-F19 keys.
* Documented support for F16-F19 keys
* Added support for F20 key
* Added support for F21-F24 on platforms except macOs
* Added support for F21-F24 on macOs
* Documented addition of F16-F24 keys
* Added missing ref qualifier
* Fixed compilation error on 1.24.1
* Refactored methods in macOs events_loop and view files
* fix <C-Tab>
* fix <CMD-{key}>
* move the NSKeyUp special handling into the match statement
* add special handling for `<Cmd-.>`
* formatting
* add return type to msg_send!
* Raised minimum version of `x11-dl`.
This fixes a compilation error in release mode on X11.
Issue #620
* Updated `CHANGELOG.md` about X11 release mode compilation issue.
* Fix unsoundness in windows backend
* Synchronize window state properly
* update changelog and add a comment to execute_in_thread
* Formatting fixes and improve changelog message
* iOS: Fix the `longjmp`/`setjmp` ffi. `jmp_buf` was the wrong size (too small) causing crashes on application launch, make longjmp return Never
* remove extra parentheses around JBLEN, and add a changelog entry about the JmpBuf fix
* remove opengl code from winit
* iOS: restrict EventsLoop to be created on the main thread
iOS: Window can only be made once, make Drop on Window thread safe
iOS: make DelegateState owned by Window, cleanup
iOS: fixes from merge (class! macro)
* update the changelog
* Fixed nitpicks
* set the UIViewController's view to the one that was just created
* capture the return value in a Unit to avoid SIGILL/SIGSEGV's.
change whitespace to be more idiomatic of Obj-C
* CHANGELOG entry
* Disabled maximize button
When creating a non resizable window in win32.
Also added example code to test.
* Added to changelog
* Added to documentation
* Removed non_resizable test
* Other suggested PR changes
* Documentation changes
* CHANGELOG nits
* Change set_cursor_position to return Result<(), String>
This is now consistent with `grab_cursor`, and
enables `window.set_cursor_position(x, y)?` in functions
that return `Result<_, Box<Error>>`.
* Adjust error handling of unimplemented cusor opertions in wayland
* The final nitpick
* Actually one more
* Windows: Use new cursor state API
* X11: Use new cursor state API
* macOS: Use new cursor state API
* Android+iOS: Stubbed new cursor state API
* Emscripten: Use new cursor state API
* Prevent multiple inc/dec of display count on Windows
* Fixed missing imports (no idea where those went)
* Remove NoneCursor
* Improved documentation
* Fix Emscripten build
* Windows: Re-grab before and after fullscreen
loops on the same thread.
This is needed for adding shared context support to glutin, as contexts
must be made with the same native display (and therefore the same
connection.)
Signed-off-by: Hal Gentz <zegentzy@protonmail.com>
* macOS: Monitor list methods on Window
* X11+Wayland: Monitor list methods on Window
* Windows: Monitor list methods on Window
* iOS: Monitor list methods on Window
* Android: Monitor list methods on Window
* Emscripten: Monitor list methods on Window
* Fixed Wayland implementation
* Fix DPI with 0 width/hight reported by xorg
* Add `WINIT_HIDPI_FACTOR` env variable
It is now possible to override the DPI factor using the
`WINIT_HIDPI_FACTOR` environment variable on X11.
The changelog also has been updated to introduce all current changes
made.
* Add documentation for the environment variable
* Fix nitpicks
* Learning the alphabet
* Panic with error message if DPI env var is <= 0
* Windows: CursorState improvements
Fixes#523
Prior to changing the cursor state, we now check the current grab
state, since it can be invalidated by alt-tabbing and other things.
`CursorState::Hide` is also implemented now.
The cursor name is now wrapped in a `Cursor` struct to allow
multithreaded access.
`Window::set_cursor_state` has been reworked to use
`execute_in_thread`. Two unneeded `transmute` calls were also
removed.
The `WM_SETCURSOR` handler is much more readable now.
`MonitorId::get_adapter_name` has been removed, since it's dead
code and appears to be a relic from 4 years ago.
* Windows: CursorState::Grab no longer hides cursor
`MouseCursor::NoneCursor` has been implemented to allow for
equivalent behavior to the older implementation.
Windows and X11 now have consistent cursor grabbing behavior.
macOS still needs to be updated.
* Windows: Grabbing auto-hides again (for now)
This API needs more work, so let's stick to a bug fix and some
refactoring. However, it now hides using a different technique
than it did originally, which applies instantly instead of after
mouse movement.
* Windows: Fix panic for set_fullscreen(None) (#501)
* Add condition to prevent panic
Trying to call set_fullscreen(None) on a window that has never been in
fullscreen mode caused a panic before this change.
The responsible method now simply checks if this precondition is met and
returns (does nothing) otherwise.
* Add entry to CHANGELOG
* Add platform specification to CHANGELOG entry
Forgot to add that the to_fullscreen(None) bugfix is Windows only in
CHANGELOG.
* X11: General cleanup
This is almost entirely internal changes, and as usual, doesn't actually
fix any problems people have complained about.
- `XSetInputFocus` can't be called before the window is visible. This
was previously handled by looping (with a sleep) and querying for the
window's state until it was visible. Now we use `XIfEvent`, which blocks
until we receive `VisibilityNotify`. Note that this can't be replaced
with an `XSync` (I tried).
- We now call `XSync` at the end of window creation and check for
errors, assuring that broken windows are never returned. When creating
invisible windows, this is the only time the output buffer is flushed
during the entire window creation process (AFAIK). For visible windows,
`XIfEvent` will generally flush, but window creation has overall been
reduced to the minimum number of flushes.
- `check_errors().expect()` has been a common pattern throughout the
backend, but it seems that people (myself included) didn't make a
distinction between using it after synchronous requests and asynchronous
requests. Now we only use it after async requests if we flush first,
though this still isn't correct (since the request likely hasn't been
processed yet). The only real solution (besides forcing a sync *every
time*) is to handle asynchronous errors *asynchronously*. For future
work, I plan on adding logging, though I don't plan on actually
*handling* those errors; that's more of something to hope for in the
hypothetical async/await XCB paradise.
- We now flush whenever it makes sense to. `util::Flusher` was added to
force contributors to be aware of the output buffer.
- `Window::get_position`, `Window::get_inner_position`,
`Window::get_inner_size`, and `Window::get_outer_size` previously all
required *several* round-trips. On my machine, it took an average of
around 80µs. They've now been reduced to one round-trip each, which
reduces my measurement to 16µs. This was accomplished simply by caching
the frame extents, which are expensive to calculate (due to various
queries and heuristics), but change infrequently and predictably. I
still recommend that application developers use these methods sparingly
and generally prefer storing the values from `Resized`/`Moved`, as
that's zero overhead.
- The above change enabled me to change the `Moved` event to supply
window positions, rather than client area positions. Additionally, we no
longer generate `Moved` for real (as in, not synthetic)
`ConfigureNotify` events. Real `ConfigureNotify` events contain
positions relative to the parent window, which are typically constant
and useless. Since that position would be completely different from the
root-relative positions supplied by synthetic `ConfigureNotify` events
(which are the vast majority of them), that meant real `ConfigureNotify`
events would *always* be detected as the position having changed, so the
resultant `Moved` was multiple levels of misleading. In practice, this
meant a garbage `Moved` would be sent every time the window was resized;
now a resize has to actually change the window's position to be
accompanied by `Moved`.
- Every time we processed an `XI_Enter` event, we would leak 4 bytes via
`util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a
dynamically-allocated mask field which we weren't freeing. As this event
occurs with fairly high frequency, long-running applications could
easily accumulate substantial leaks. `util::PointerState::drop` now
takes care of this.
- The `util` module has been split up into several sub-modules, as it
was getting rather lengthy. This accounts for a significant part of this
diff, unfortunately.
- Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't
typically be a round-trip anyway, but the added complexity is
negligible.
- Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this
backend). There appears to be no downside to this, but if anyone finds
one, this would be easy to revert.
- The WM name and supported hints are now global to the application, and
are updated upon `ReparentNotify`, which should detect when the WM was
replaced (assuming a reparenting WM was involved, that is). Previously,
these values were per-window and would never update, meaning replacing
the WM could potentially lead to (admittedly very minor) problems.
- The result of `Window2::create_empty_cursor` will now only be used if
it actually succeeds.
- `Window2::load_cursor` no longer re-allocates the cursor name.
- `util::lookup_utf8` previously allocated a 16-byte buffer on the heap.
Now it allocates a 1024-byte buffer on the stack, and falls back to
dynamic allocation if the buffer is too small. This base buffer size is
admittedly gratuitous, but less so if you're using IME.
- `with_c_str` was finally removed.
- Added `util::Format` enum to help prevent goofs when dealing with
format arguments.
- `util::get_property`, something I added way back in my first winit PR,
only calculated offsets correctly for `util::Format::Char`. This was
concealed by the accomodating buffer size, as it would be very rare for
the offset to be needed; however, testing with a buffer size of 1,
`util::Format::Long` would read from the same offset multiple times, and
`util::Format::Short` would miss data. This function now works correctly
for all formats, relying on the simple fact that the offset increases by
the buffer size on each iteration. We also account for the extra byte
that `XGetWindowProperty` allocates at the end of the buffer, and copy
data from the buffer instead of moving it and taking ownership of the
pointer.
- Drag and drop now reliably works in release mode. This is presumably
related to the `util::get_property` changes.
- `util::change_property` now exists, which should make it easier to add
features in the future.
- The `EventsLoop` device map is no longer in a mutex.
- `XConnection` now implements `Debug`.
- Valgrind no longer complains about anything related to winit (with
either the system allocator or jemalloc, though "not having valgrind
complain about jemalloc" isn't something to strive for).
* X11: Add better diagnostics when initialization fails
* X11: Handle XIQueryDevice failure
* X11: Use correct types in error handler
* Add Copy/Paste keys
This is only a tiny update which introduces the `Copy` and `Paste` keys
which are present on X11/Wayland/Windows. I'm not sure if this exists on
MacOS too, but I'm not able to test that and it doesn't have names but
just matches on the hex key values.
The "Copy" element is a reserved keyword in Rust but shouldn't cause any
conflicts in this scenario, this behavior falls in line with
https://docs.rs/winit/0.13.1/winit/enum.MouseCursor.html#variant.Copy,
but it would be possible to rename it. However `Copy` seems like the
most intuitive choice.
* Add Cut key, fix windows and update CHANGELOG
This introduces a bunch of minor fixes:
* The changes introduced by this branch have been added to the changelog
* Since related, the `Cut` key has also been added
* An attempt has been made to fix Windows
* Fix position of fallback comment
The new keys have been inserted at the wrong position, so the fallback
comment has been moved to the `_ => ...` section again.
* Fix windows build
Apparently there are no keys for Cut/Paste on Windows, so for now those
have been removed on Windows and only the `Copy` key has been added on
Windows, the changelog has been updated to reflect that.
Linux still implements Copy/Clone/Paste, but `Copy` is now working
properly on Wayland.
MacOS still does not have any of these keys.
* Remove Windows changes
Because the Windows design wasn't completely clear the VirtualKeyCode
variants are now only used on Linux with X11 and Wayland and ignored on
both MacOS and Windows.
The CHANGELOG has also been updated. Windows has been removed from it
and the Linux section has been clarified a bit.
* macOS: fix regression in 03c3e79409
fixed !decorations case and refactored logic to be a little easier to
read
* fix default case to inlude closable mask
* add comment to default case of macos platform attrs
* x11: Always receive Awakened event in run_forever
Do not reset the pending_wakeup boolean at the start of run_forever so
that each call to EventsLoopProxy::wakeup results in an Awakened event.
Fixes#462
* Update CHANGELOG.md
Fixes#467
All variants other than Text have been implemented. While Text can
be implemented using ToUnicode, that doesn't play nice with dead
keys, IME, etc.
Most of the mouse DeviceEvents were already implemented, but due
to the flags that were used when registering for raw input events,
they only worked when the window was in the foreground.
This is also a step forward for #338, as DeviceIds are no longer
useless on Windows. On DeviceEvents, the DeviceId contains that
device's handle. While that handle could ostensibly be used by
developers to query device information, my actual reason for
choosing it is because it's simply a very easy way to handle this.
As a fun bonus, this enabled me to create this method:
DevideIdExt::get_persistent_identifier() -> Option<String>
Using this gives you a unique identifier for the device that
persists across replugs/reboots/etc., so it's ideal for something
like device-specific configuration.
There's a notable caveat to the new DeviceIds, which is that the
value will always be 0 for a WindowEvent. There doesn't seem to be
any straightforward way around this limitation.
I was concerned that multi-window applications would receive n
copies of every DeviceEvent, but Windows only sends them to one
window per application.
Lastly, there's a chance that these additions will cause
antivirus/etc. software to detect winit applications as keyloggers.
I don't know how likely that is to actually happen to people, but
if it does become an issue, the raw input code is neatly
sequestered and would be easy to make optional during compilation.
* adding a multiwindow example
* Added NSAutoReleasepool for WindowDelegate::Drop
as setDelegate:nil autoreleases WindowDelegate during work.
Added NSAutoReleasepool for Window2::Create,
as it uses autorelease on objects while doing work.
Added NSAutoreleasepool for Window2::Drop
as nswindow::close uses autorelease on objects.
Added NSAutoreleasepool for IdRef.
Moved Window2 WinitWindow objc class to a static var, as we are creating
multiple windows.
* specifying return type for msg_send!
* removing example/recreate_window_leak.rs
* EventLoop, Shared, no need to retain dead weak ptr
* Change log entry added
* added comment about Shared.find_and_remove_window
* fixed code style errors
* Remove executable flag from os/macos.rs
This was causing me some grief while working on Windows, and it
doesn't belong here to begin with.
* Windows: get_position returns screen coordinates instead of workspace coordinates
Previously, get_position used GetWindowPlacement. As per the
documentation of WINDOWSTRUCT, the returned coordinates are in
workspace space, meaning they're relative to the taskbar. It's
also explicitly remarked that these coordinates should only be
used in conjunction with SetWindowPlacement, as mixing them with
functions expecting screen coordinates can cause unpleasantness.
Since our set_position (correctly) uses SetWindowPos, this meant
that passing the return of get_position to set_position would
cause the window to move.
We now use GetWindowRect, which returns screen coordinates. This
gives us both better consistency within the Windows backend and
across platforms.
Note that this only makes a difference if the taskbar is visible.
With the taskbar hidden, the values are exactly the same as before.
* Windows: Moved event position values are consistent with get_position
The old Moved values had two problems:
* They were obtained by casting a WORD (u16) straight to an i32.
This meant wrap-around would never be interpreted as negative,
thus negative positions (which are ubiquitous when using multiple
monitors) would result in positions around u16::MAX.
* WM_MOVE supplies client area positions, not window positions.
Switching to handling WM_WINDOWPOSCHANGED solves both of these
problems.
* Better documentation for Moved and Resized
Without this pin, an existing cargo.lock for an older winit will not
update the x11-dl dependency, and thus will select a version that is
missing required new XIM features.
* bump minor version
* update changelog date and fix typo
* Updated date (we're going to release for real this time)
* Update version in README for the first time in a long time
* Replace Closed event with CloseRequested and Destroyed
Implements #434
The existing Closed event had ambiguous meaning, both in name and in
cross-platform behavior. Closed is now split into two more precise events:
* CloseRequested - the window has been requested to close, most commonly by
having clicked the window's close button. Whether or not you respond by
closing the window is up to you.
* Destroyed - the window has been destroyed, and can no longer be safely
used.
Most notably, now you can reliably implement classic patterns like
prompting the user to save their work before closing, and have the
opportunity to perform any necessary cleanup.
Migrating to the new API is straightforward. In most cases, you can simply
replace all existing usages of Closed with CloseRequested. For more
information, see the example programs, particularly handling_close and
multiwindow.
iOS applications must replace all usages of Closed with Destroyed, and
require no other changes.
* x11: Windows are Sync again
Fixes#472
* Add test ensuring that Window is Sync
Window must be Sync for Vulkano's Arc<FramebufferAbstract> to be usable.
This reverts commit 19cd53193b.
Testing fullscreen functionality revealed that windowDidResize is invoked in more cases than previously thought, causing the user's events to be eaten and HiDPI problems.
* Added helper function for make monitor from display.
* Implement get_current_monitor for macos
* Implemented with_fullscreen and set_fullscreen for macos
* Implemented set_decorations for macos
* Implement set_maximized and with_maximized for macos
* Changed fullscreen example fullscreen keypress from F11 to F
* Update CHANGELOG.md
* Add and fixed some comments
* Reformat and add more comments
* Better handling window and maximized state
* Reformat and typo fix
* Add get_inner_position for windows, prototypes for other platforms
* Fix linux builds
* Implement get_inner_position for osx
* Add get_inner_pos implementations for other platforms
* Fixed get_inner_position on macOS
* Corrected set_position on macOS
* Added CHANGELOG entry
* Discard mouse down after Cocoa window resize
We are sending the mouse down event after the window resize has
completed, because Cocoa uses a modal event loop to implement window
resize. This leads to a mouse down without a matching mouse up.
* Also handle event discard in poll_events
Add some explanatory comments and a changelog entry.
* Implement set_fullscreen for windows
* Implement get_current_monitor for windows
* Implement set_maximized
* Implement set_decorations for windows
* Update CHANGELOG.md
* Fixed minor syntax bug for stable rust version
* Added support for WindowBuilder::with_maximized
* Move all window sized related functions to main thread
* Refactor and formatting force_window_active
* Remove unused code
* Update CHANGELOG.md
* Refactor and change keyboard handling code
* Reformatting and refactoring
* Added back missing link for comment
* Fixed set_maximized and set_fullscreen wrong order bug
* Call ShowWindow(SW_RESTORE) when restore_saved_window
* Sync system maximized status when set_fullscreen
* Fixed wrong function name
Fixes#195Fixes#277Fixes#455
* Read `XMODIFIERS` explicitly/directly instead of calling `XSetLocaleModifiers` with an
empty string. This is useful for debugging purposes, and more clear to read and handle.
* Fallback to local input method if the one specified in `XMODIFIERS` is later closed on the
server end (i.e. if ibus/fcitx is terminated). Previously, that would cause the event loop
to freeze and usually also segfault.
* If using the fallback input method, respond to the `XMODIFIERS` input method later
becoming available. This means that the input method restarting is handled, and that even if
the program was started while ibus/fcitx/etc. was unavailable, it will start using it as
soon as it becomes available.
* Only one input method is opened for the whole event loop, with each window having its own
input context.
* IME works completely out of the box now, no longer requiring application developers to
call `setlocale` or `XSetLocaleModifiers`.
* Detailed error messages are provided if no input method could be opened. However, no
information is provided to the user if their intended `XMODIFIERS` input method failed to
open but the fallbacks (which will ostensibly always succeed) succeeded; in my opinion, this
is something that is best filled by adding a logging feature to winit.
XIM isn't thread-safe at all. Any call made to it from another thread will result in the
event loop freezing (this is why the old implementation of Drop for Window had that
problem).
XIM is now confined to one thread, and the existing API is maintained using channels. In
testing this with Alacritty, I initially thought the occasional slight lag on updating the
spot location was due to this change, but it's present without it as well.
* Try XOpenIM with different locale modifiers
Implements the solution suggested in
https://github.com/tomaka/winit/issues/277#issuecomment-337751136.
* Use empty XSetLocaleModifiers beforehand
Also, for modifiers, convert from length-based UTF-8 strings to
null-terminated bytestrings.
* Add CHANGELOG entry and comments
* make windows without decorations resizable and movable in macos
fixes#368
The subclassing logic was copied from servo's fork of glutin:
63026a0f4c/src/api/cocoa/mod.rs (L418)
* remove `isMovableByWindowBackground` and `mouseDownCanMoveWindow`
* revert example changes
* remove resizable mask from decoration: false
* avoid duplicate class declarations
* update changelog
* fix changelog
* changelog whitespace
* Add min/max size setting for win32 and wayland backends
* Implement dynamic min/max size on macos
* Add min/max size setting for x11
* Add empty functions for remaining platforms
* Improved min/max size setting for x11
* Added CHANGELOG entry for new min/max methods
* Added documentation for new min/max methods
* On win32, bound window size to min/max dimensions on window creation
* On win32, force re-check of window size when changing min/max dimensions
* Fix freeze when setting min and max size
Fixes#79#414
This changes the implementation of Drop for Window to send a WM_DELETE_WINDOW ClientMessage,
offloading all the cleanup and window destruction to the event loop. Unsurprisingly, this
entails that the event loop now handles WM_DELETE_WINDOW using the behavior that was
previously contained in Window's Drop implementation, along with destroying the Window.
Not only does this mean that dropped windows are closed, but also that clicking the × button
on the window actually closes it now.
The previous implemention of Drop was also broken, as the event loop would be (seemingly
permenanently) frozen after its invocation. That was caused specifically by the mutex
locking, and is no longer an issue now that the locking is done in the event loop.
While I don't have full confidence that it makes sense for the Drop implementation to behave
this way, this is nonetheless a significant improvement. The previous behavior led to
inconsistent state, panics, and event loop breakage, along with not actually destroying the
window.
This additionally makes the assumption that users don't need Focused or CursorLeft events
for the destroyed window, as Closed is adequate to indicate unfocus, and users may not
expect to receive events for closed/dropped windows. In my testing, those specific events
were sent immediately after the window was destroyed, though this sort of behavior could be
WM-specific. I've opted to explicitly suppress those events in the case of the window no
longer existing.
* macOS: Allow hiding the title from the titlebar
* macOS: Allow making the titlebar transparent
* macOS: Give control over content view size
Allows setting `NSFullSizeContentViewWindowMask` with WindowBuilder.
* macOS: add `.with_titlebar_hidden` to WindowBuilderExt
* macOS: adds `titlebar_buttons_hidden` platform specific attribute
Due to the recent changes in the Rust compiler, unconstrained type
variables are now deduced to `!` instead of `()`. There are some
occurrences where `msg_send!` is used without constraining its return
type (relying on the assumption that they would be deduced to be `()`).
As a result, the macOS port of winit stopped working.
This PR fixes this issue (#426) by adding explicit return types to
such uses of `msg_send!`.
The fix for returning accurate window position lead to a regression
computing inner size in pixels. This commit resolves that by getting
inner size from the window ID winit caches and still resolving position
by climbing the window hierarchy.
Resolves#398
* x11: Support XRandR versions older than 1.5
Fixes#392
Previously, initializing the member `xrandr` of `XConnection` resulted
in a panic when symbols from XRandR version 1.5 were missing. There was
already code to handle older versions of XRandR but it was never
executed because of the panic.
The member `XConnection.xrandr` now contains only functions that can
safely be used with older versions. Additionally, this commit adds a new
member to `XConnection` of type `Option<ffi::XRandr>` that only contains
a value if version 1.5 functionality is present.
* x11: Document the xrandr* members of XConnection
Some window managers like i3wm will actually nest application windows
(like those opened by winit) within other windows to, for example, add
decorations. Initially when debugging this method on i3, the x and y
positions were always returned as "2".
The solution that other xlib abstractions use is to climb up the window
hierarchy until just below the root window, and that window must be used
to determine the appropriate position.
This patch doesn't take into account borders or the window's offset
within its parent, but it's much more usable than the original
implementation on certain WMs.
* macOS: Move the window if there is no title bar
On macOS by default windows can only be moved by clicking and
dragging on the titlebar, if we spawn a window without one we
need to set the `movableByWindowBackground` property.
Partial fix for #368
* macOS: Make moveByWindowBackground optional
Implements setting the property via WindowBuilderExt:
WindowBuilder::new()
.with_decorations(false)
.with_movable_by_window_background(true)
* Update CHANGELOG
This was previously hardcoded to 1.0. The values for physical size in
millimeters and pixel counts on each axis are used to compute the dpi
per monitor.
The `CursorMoved` events that are used to send position updates alongside `Focused` and
`CursorEntered` events were using incorrect values for the window ID. This is a direct
result of the X11 backend being hard to understand, as those values came from variables in
the top-level scope of the function, which one would assume to be valid throughout the
entirety of their scope. In reality, their validity is dependent on the event belonging to
the `XEvent` union, so very surprising things can happen if those variables are read in the
case of XInput2/XKB/etc. events. To prevent future accidents, the aforementioned variables
have been removed, and are now defined per-event instead.
Additionally, the `CursorMoved` event sent alongside `Focused` now uses the correct device
ID; it previously used the ID of a master keyboard, but now uses the ID of the pointer
paired to that keyboard. Note that for those using multi-pointer X, the correctness of this
ID is dependent on the correctness of the window manager's focus model.
This has been stubbed on all platforms other than X11. The X11 implementation has also been
revised to toggle correctly, as it was previously only able to remove decorations.
Fixes#256
`get_xlib_window` and `get_xlib_screen_id` previously returned `Option<*mut c_void>` by
casting integer IDs into pointers, which while producing no functionality issues, is
semantically incorrect and rather surprising. Worse still, the docs for `get_xlib_window`
stated that it was in fact a valid pointer.
Additionally, now all `unix::WindowExt` methods return `std::os::raw` types rather than
`libc` types; note that the two versions of `c_void` are not interchangeable in the eyes of
the compiler, so those wanting the `libc` type will need to explicitly cast.
This is a breaking change, and will require some trivial changes to glutin.
Previously, the maximization hints were being sent as two separate client messages: one for
horizontal, and one for vertical. That resulted in the window only being maximized
horizontally (at least with my WM). The corrected client message sets both of these hints at
once.
In the process of implementing that, the relevant components were refactored to use the util
module, as we gradually move towards a hopeful future of a more readable X11 backend.
Fixes#282
Some tiling window managers (i3, dwm, etc.) determine how a window should behave based on
its name. If the name is set after mapping, then window managers will check the name before
we set it, followed by them detecting it as a change when the name is actually set. That
results in the window briefly behaving in an unexpected way, followed by a rapid switch to
the expected behavior.
In accordance to section 4.1.2 of ICCCM, the name, decorations, size hints, and window
deletion redirection have all been moved up to be set before mapping.
* Update mouse pos after cursor enter event
* Update mouse position on windows focus
* Send device_id
* Update other device id
* Fix windows import
* Remove deque for vec
* Just send event
* Use correct push_back method
* Push correct event
* Explicit mouse-related DeviceEvents
This makes the API more intuitive for common use-cases and allows us
to better propagate platform knowledge of motion semantics.
* Improve event naming consistency
* Clarify axis event forwards-compatibility
* Rename WindowEvent::MouseMoved/Entered/Left to CursorMoved/...
This emphasizes the difference between motion of the host GUI cursor,
as used for clicking on things, and raw mouse(-like) input data, as
used for first-person controls.
* Add support for windows and OSX, fix merging
* Fix warnings and errors on Linux
* Remove unnecessary breaking changes
* Add MouseWheel events to windows and OSX
* Fix bad push call.
* Fix docs, naming, and x11 events
* Remove mutability warning
* Add changelog entry
* wayland: upgrade wayland-window
This new version of wayland window considerably simplifies the
window handling for winit, meaning much of the previous juggling
is no longer needed, and the windows will appear even if nothing is
drawn.
* wayland: cleanup unused stuff
* Fix no primary monitor panic in XWayland
In this case try to use the first existing monitor instead of panicking.
Fixes#317
* Shift no monitor panic to x11::get_primary_monitor
* Update changelog with xll get_primary_monitor fallback
* Implement public API for high-DPI #105
* Recover get_inner_size_points and get_inner_size_pixels and change their implementation assuming get_inner_size() returns size in pixels
* Update changelog for high-DPI changes
This should trigger the compositor's mechanism for sending a
configure event, which should most of the time be processed
before any winit user actually tries to draw.
* Add an i386 target to travis
* Fix X11 on 32bit architectures
One would hope 32bit X11 was dead by now but apparently not :). Fix
the window hint setting code to not assume window IDs are 64bit as
apparently they are not in 32bit arches.
* wayland: don't create a second event_queue
As each EventsLoop has its own context, this is no longer necessary.
* wayland: buffer events rather than direct dispatch
Changes the behavior of the event loop to first internally
buffer the events generated by the wayland handlers, and then
dispatch them to the client's closure.
- It simplifies the event loop logic
- It makes it possible for the user to call window methods such as
`set_title()` or `set_inner_size()` without causing a deadlock
* wayland: add is_ready() & fix protocol errors
Adds a `is_ready()` method to the windows to advertize
when it is legal to start drawing, and fix a few wayland
protocol mishandling in the process.
* partial implementation for emscripten
this pull request contain a partial but working implementation of emscripten backend
some implementations may be controversial.
here some implementation detail:
* cursor state:
* on grab: emscripten request pointer lock deferred and also set a callback when pointer lock change
the callback request pointer lock deferred.
* on hide: `emscripten_hide_mouse` exist but not `emscripten_show_mouse`
a pull request has been open on october 2016 but never been merged
so I copied the javascript function and put it in emscripten_asm_const function
* fullscreen: if fullscreen is requested then it request fullscreen deferred and set a callback on fullscreen change
the callback request fullscreen deferred
* run forever: this method use emscripten main loop to run an infinite loop
* keyboard callback doesn't consume the event. I think it is more apopriate as in desktop environment it is the same, is it ?
* emscripten dir is added in example and contains html pages
Some things that are not implemented:
* lots of events
* min and max dimension can be implemented with a callback that listen to size change and resize if dimension out of bound
* title may be implemented using javascript to change document.title
* Use std::os::raw in the emscripten bindings
* Fix emscripten code
* Update code
* Add CI
* Remove the emscripten-specific examples
* Add some information to the README
* Fix X11 screen resolution change using XrandR
The previous XF86 resolution switching was broken and everything
seems to have moved on to xrandr. Use that instead while cleaning
up the code a bit as well.
* Use XRandR for actual multiscreen support in X11
* Use actual monitor names in X11
* Get rid of ptr::read usage in X11
* Use a bog standard Vec instead of VecDeque
* Get rid of the XRandR mode switching stuff
Wayland has made the decision that apps shouldn't change screen
resolutions and just take the screens as they've been setup. In the
modern world where GPU scaling is cheap and LCD panels are scaling
anyway it makes no sense to make "physical" resolution changes when
software should be taking care of it. This massively simplifies the
code and makes it easier to extend to more niche setups like MST and
videowalls.
* Rename fullscreen options to match new semantics
* Implement XRandR 1.5 support
* Get rid of the FullScreen enum
Moving to just having two states None and Some(MonitorId) and then
being able to set full screen in the current monitor with something
like:
window.set_fullscreen(Some(window.current_monitor()));
* Implement Window::get_current_monitor()
Do it by iterating over the available monitors and finding which
has the biggest overlap with the window. For this MonitorId needs
a new get_position() that needs to be implemented for all platforms.
* Add unimplemented get_position() to all MonitorId
* Make get_current_monitor() platform specific
* Add unimplemented get_current_monitor() to all
* Implement proper primary monitor selection in X11
* Shut up some warnings
* Remove libxxf86vm package from travis
Since we're no longer using XF86 there's no need to keep the package
around for CI.
* Don't use new struct syntax
* Fix indentation
* Adjust Android/iOS fullscreen/maximized
On Android and iOS we can assume single screen apps that are already
fullscreen and maximized so there are a few methods that are implemented
by just returning a fixed value or not doing anything.
* Mark OSX/Win fullscreen/maximized unimplemented()!
These would be safe as no-ops but we should make it explicit so
there is more of an incentive to actually implement them.
* Don't use UNIX_BACKEND in Window2::new
* Move get_available_monitors and get_primary_monitor to EventsLoop
* Remove UNIX_BACKEND
* Restore choosing the Linux backend
* Return a XNotSupported for new_x11()
* Fix fullscreen example
* Rework MonitorId::get_native_identifier
* Try fix compilation
* Returns the monitor ID on wayland as well
* Try fix compilation
* Fix iOS compilation
Use the enum to make a single fullscreen API that's much more
consistent. Both set_fullscreen() and with_fullscreen() take the
same enum and support all the variations so you can build the window
however you want and switch between the modes at runtime.
There are two kinds of fullscreen. One where you take over the whole
output the other where you just set the window size to the screen
size and get rid of decorations. The first one already existed,
implement the second which is more common for normal desktop apps.
Use an enum to consolidate all the fullscreen states.
When X's evdev input module is configured to emulate scroll events (as
used with e.g. trackpoints), it generates non-emulated scroll button
presses and does not generate motion events. This is contrary to the
behavior of all other hardware I've tested, and contrary to the
behavior of libinput, but nonetheless should be supported.
X11 always return the geometry in pixel units. Since
window.get_inner_size returns the size in points in other window manager
implementations X11 should also return in points instead of pixels.
This removes the need for the EventsLoop::interrupt method by inroducing
a ControlFlow type. This new type is to be returned by the user's
callback and indicates whether the `EventsLoop` should continue waiting
for events or break from the loop.
Only the wayland, x11 and api_transition backends have been updated so
far, and only the wayland backend has actually been tested.
X11 and Wayland implementations are now half implemented, however both
still do not correctly break from the inner blocking event dispatch
functions when `wakeup` is called, which they should do.
This commit only updates the top-level API to get some early feedback.
None of the platform-specific code has been updated yet. I'm hoping to
get around to this over the next couple days however if someone more
familiar with the windows backend would like to do a PR against this
fork that would be a great help.
Closes#187.
If the interrupted flag were set going into poll_events, it would only
ever handle the first event in the queue. Now, the flag is reset at the
start so events are processed until the caller requests otherwise.
This is the same behavior as with WindowProxy::wakeup_event_loop in
previous versions.
Unfortunately, `EventsLoop::interrupt` is also the recommend way to exit
a `run_forever` loop from within the event handler callback. Pushing an
extra event on the queue in that case is simply wasteful. Changing this
would require a refactor taking one of two possible forms:
1. Add a method *in addition* to interrupt intended for waking up the
event loop
2. Add a return type to the event callback like
enum Continue { True, False }
which would be used in lieu of the atomic interrupt flag.
It was only processing a single event per call. The docs say
> Fetches all the events that are pending, calls the callback function
> for each of them, and returns.
which suggests that was incorrect.
All platforms should now receive events in the following order:
1. KeyboardInput(ElementState::Pressed, ..)
2. ReceivedCharacter
3. KeyboardInput(ElementState::Released, ..)
cc https://github.com/tomaka/glutin/issues/878
ICCCM 4.1.2.5 (https://tronche.com/gui/x/icccm/sec-4.html#WM_CLASS)
states that:
> This property must be present when the window leaves the Withdrawn
> state and may be changed only while the window is in the Withdrawn
> state.
Previously, we would first map the window, and then set these
properties, causing sadness for window managers (#167,
tomaka/glutin#879). This patch changes that by setting the class and
name attributes immediately after the window is created, and before it
is mapped.
Fixes#167.
This expands input events to represent sub-pixel mouse positions, devices responsible for generating events, and raw
device-oriented events. The X11 back end is refactored to make full use of the new expressiveness. Other backends have
had new functionality minimally stubbed out, save for the macos backend which already supports sub-pixel mouse
positions.
2017-04-23 01:08:15 -07:00
351 changed files with 67378 additions and 10139 deletions
description:Description of the problem does this solve or what need does it fill? Please be mindful of the [scope](https://github.com/rust-windowing/winit/blob/master/FEATURES.md) of Winit.
validations:
required:true
- type:dropdown
id:platforms
attributes:
label:Relevant platforms
description:Onwhich platforms is this feature relevant?
For features _within_ the scope of winit, see [FEATURES.md](FEATURES.md).
For features _outside_ the scope of winit, see [Are we GUI Yet?](https://areweguiyet.com/) and [Are we game yet?](https://arewegameyet.rs/), depending on what kind of project you're looking to do.
## Contact Us
Join us in our [](https://matrix.to/#/#rust-windowing:matrix.org) room.
The maintainers have a meeting every friday at UTC 15. The meeting notes can be found [here](https://hackmd.io/@winit-meetings).
## Usage
Winit is a window creation and management library. It can create windows and lets you handle
events (for example: the window being resized, a key being pressed, a mouse mouvement, etc.)
produced by window.
events (for example: the window being resized, a key being pressed, a mouse movement, etc.)
produced by the window.
Winit is designed to be a low-level brick in a hierarchy of libraries. Consequently, in order to
show something on the window you need to use the platform-specific getters provided by winit, or
another library.
```rust
externcratewinit;
## CONTRIBUTING
fnmain(){
letwindow=winit::Window::new().unwrap();
For contributing guidelines see [CONTRIBUTING.md](./CONTRIBUTING.md).
## MSRV Policy
This crate's Minimum Supported Rust Version (MSRV) is **1.85**. Changes to
the MSRV will be accompanied by a minor version bump.
As a **tentative** policy, the upper bound of the MSRV is given by the following
formula:
foreventinwindow.wait_events(){
matchevent{
winit::Event::Closed=>break,
_=>()
}
}
}
```
min(sid, stable - 3)
```
Where `sid` is the current version of `rustc` provided by [Debian Sid], and
`stable` is the latest stable version of Rust. This bound may be broken in case of a major ecosystem shift or a security vulnerability.
# Using allow-invalid because this is platform-specific code
disallowed-macros=[
{path="std::print",reason="works badly on web",replacement="tracing::info"},
{path="std::println",reason="works badly on web",replacement="tracing::info"},
{path="std::eprint",reason="works badly on web",replacement="tracing::error"},
{path="std::eprintln",reason="works badly on web",replacement="tracing::error"},
{path="std::dbg",reason="leftover debugging aid, remove it or use tracing"},
]
disallowed-methods=[
{allow-invalid=true,path="objc2_app_kit::NSView::visibleRect",reason="We expose a render target to the user, and visibility is not really relevant to that (and can break if you don't use the rectangle position as well). Use `frame` instead."},
{allow-invalid=true,path="objc2_app_kit::NSWindow::setFrameTopLeftPoint",reason="Not sufficient when working with Winit's coordinate system, use `flip_window_screen_coordinates` instead"},
{allow-invalid=true,path="web_sys::Document::exit_fullscreen",reason="Doesn't account for compatibility with Safari"},
{allow-invalid=true,path="web_sys::Document::fullscreen_element",reason="Doesn't account for compatibility with Safari"},
{allow-invalid=true,path="web_sys::Element::request_fullscreen",reason="Doesn't account for compatibility with Safari"},
{allow-invalid=true,path="web_sys::HtmlCanvasElement::height",reason="Winit shouldn't touch the internal canvas size"},
{allow-invalid=true,path="web_sys::HtmlCanvasElement::set_height",reason="Winit shouldn't touch the internal canvas size"},
{allow-invalid=true,path="web_sys::HtmlCanvasElement::set_width",reason="Winit shouldn't touch the internal canvas size"},
{allow-invalid=true,path="web_sys::HtmlCanvasElement::width",reason="Winit shouldn't touch the internal canvas size"},
{allow-invalid=true,path="web_sys::HtmlElement::style",reason="cache this to reduce calls to JS"},
{allow-invalid=true,path="web_sys::MouseEvent::buttons",reason="Use `backend::event::cursor_buttons()` to avoid wrong conversions"},
{allow-invalid=true,path="web_sys::MouseEvent::button",reason="Use `backend::event::cursor_button()` to avoid wrong conversions"},
{allow-invalid=true,path="web_sys::PointerEvent::pointer_type",reason="Use `WebPointerType` to emit warnings"},
{allow-invalid=true,path="web_sys::Window::document",reason="cache this to reduce calls to JS"},
{allow-invalid=true,path="web_sys::Window::get_computed_style",reason="cache this to reduce calls to JS"},
{allow-invalid=true,path="web_sys::Window::navigator",reason="cache this to reduce calls to JS"},
{allow-invalid=true,path="web_sys::window",reason="is not available in every context"},
/// A hand (often used to indicate links in web browsers).
Hand,
/// Self explanatory.
Arrow,
/// Indicates something is to be moved.
Move,
/// Indicates text that may be selected or edited.
Text,
/// Program busy indicator.
Wait,
/// Help indicator (often rendered as a "?")
Help,
/// Progress indicator. Shows that processing is being done. But in contrast
/// with "Wait" the user may still interact with the program. Often rendered
/// as a spinning beach ball, or an arrow with a watch or hourglass.
Progress,
/// Cursor showing that something cannot be done.
NotAllowed,
ContextMenu,
NoneCursor,
Cell,
VerticalText,
Alias,
Copy,
NoDrop,
Grab,
Grabbing,
AllScroll,
ZoomIn,
ZoomOut,
/// Indicate that some edge is to be moved. For example, the 'SeResize' cursor
/// is used when the movement starts from the south-east corner of the box.
EResize,
NResize,
NeResize,
NwResize,
SResize,
SeResize,
SwResize,
WResize,
EwResize,
NsResize,
NeswResize,
NwseResize,
ColResize,
RowResize,
}
/// Describes how glutin handles the cursor.
#[derive(Debug, Copy, Clone, PartialEq)]
pubenumCursorState{
/// Normal cursor behavior.
Normal,
/// The cursor will be invisible when over the window.
Hide,
/// Grabs the mouse cursor. The cursor's motion will be confined to this
/// window and the window has exclusive access to further events regarding
/// the cursor.
///
/// This is useful for first-person cameras for example.
Grab,
}
/// Attributes to use when creating a window.
#[derive(Clone)]
pubstructWindowAttributes{
/// The dimensions of the window. If this is `None`, some platform-specific dimensions will be
/// used.
///
/// The default is `None`.
pubdimensions: Option<(u32,u32)>,
/// The minimum dimensions a window can be, If this is `None`, the window will have no minimum dimensions (aside from reserved).
///
/// The default is `None`.
pubmin_dimensions: Option<(u32,u32)>,
/// The maximum dimensions a window can be, If this is `None`, the maximum will have no maximum or will be set to the primary monitor's dimensions by the platform.
///
/// The default is `None`.
pubmax_dimensions: Option<(u32,u32)>,
/// If `Some`, the window will be in fullscreen mode with the given monitor.
///
/// The default is `None`.
pubmonitor: Option<platform::MonitorId>,
/// The title of the window in the title bar.
///
/// The default is `"glutin window"`.
pubtitle: String,
/// Whether the window should be immediately visible upon creation.
///
/// The default is `true`.
pubvisible: bool,
/// Whether the the window should be transparent. If this is true, writing colors
/// with alpha values different than `1.0` will produce a transparent window.
///
/// The default is `false`.
pubtransparent: bool,
/// Whether the window should have borders and bars.
///
/// The default is `true`.
pubdecorations: bool,
/// [iOS only] Enable multitouch, see [UIView#multipleTouchEnabled]
//! | `NativeActivity` | `android-native-activity` | Built-in to Android - it is possible to use without compiling any Java or Kotlin code. Java or Kotlin code may be needed to subclass `NativeActivity` to access some platform features. It does not derive from the [`AndroidAppCompat`] base class.|
//! | [`GameActivity`] | `android-game-activity` | Derives from [`AndroidAppCompat`], a defacto standard `Activity` base class that helps support a wider range of Android versions. Requires a build system that can compile Java or Kotlin and fetch Android dependencies from a [Maven repository][agdk_jetpack] (or link with an embedded [release][agdk_releases] of [`GameActivity`]) |
// The values are from the https://github.com/apple-oss-distributions/IOHIDFamily/blob/19666c840a6d896468416ff0007040a10b7b46b8/IOHIDSystem/IOKit/hidsystem/IOLLEvent.h#L258-L259
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
