# Objective
- Another step towards unifying our orthonormal basis construction
#20050
- Preserve behavior but fix a bug. Unification will be a followup after
these two PRs and will need more thorough testing.
## Solution
- Make shadow cubemap sampling orthonormalize have the same function
signature as the other orthonormal basis functions in bevy
## Testing
- 3d_scene + lighting examples
# Objective
Picking was changed in the UI transform PR to walk up the tree
recursively to check if an interaction was on a clipped node.
`OverrideClip` only affects a node's local clipping rect, so valid
interactions can be ignored if a node has clipped ancestors.
## Solution
Add a `Without<OverrideClip>` query filter to the picking systems'
`child_of_query`s.
## Testing
This modified `button` example can be used to test the change:
```rust
//! This example illustrates how to create a button that changes color and text based on its
//! interaction state.
use bevy::{color::palettes::basic::*, input_focus::InputFocus, prelude::*, winit::WinitSettings};
fn main() {
App::new()
.add_plugins(DefaultPlugins)
// Only run the app when there is user input. This will significantly reduce CPU/GPU use.
.insert_resource(WinitSettings::desktop_app())
// `InputFocus` must be set for accessibility to recognize the button.
.init_resource::<InputFocus>()
.add_systems(Startup, setup)
.add_systems(Update, button_system)
.run();
}
const NORMAL_BUTTON: Color = Color::srgb(0.15, 0.15, 0.15);
const HOVERED_BUTTON: Color = Color::srgb(0.25, 0.25, 0.25);
const PRESSED_BUTTON: Color = Color::srgb(0.35, 0.75, 0.35);
fn button_system(
mut input_focus: ResMut<InputFocus>,
mut interaction_query: Query<
(
Entity,
&Interaction,
&mut BackgroundColor,
&mut BorderColor,
&mut Button,
&Children,
),
Changed<Interaction>,
>,
mut text_query: Query<&mut Text>,
) {
for (entity, interaction, mut color, mut border_color, mut button, children) in
&mut interaction_query
{
let mut text = text_query.get_mut(children[0]).unwrap();
match *interaction {
Interaction::Pressed => {
input_focus.set(entity);
**text = "Press".to_string();
*color = PRESSED_BUTTON.into();
*border_color = BorderColor::all(RED.into());
// The accessibility system's only update the button's state when the `Button` component is marked as changed.
button.set_changed();
}
Interaction::Hovered => {
input_focus.set(entity);
**text = "Hover".to_string();
*color = HOVERED_BUTTON.into();
*border_color = BorderColor::all(Color::WHITE);
button.set_changed();
}
Interaction::None => {
input_focus.clear();
**text = "Button".to_string();
*color = NORMAL_BUTTON.into();
*border_color = BorderColor::all(Color::BLACK);
}
}
}
}
fn setup(mut commands: Commands, assets: Res<AssetServer>) {
// ui camera
commands.spawn(Camera2d);
commands.spawn(button(&assets));
}
fn button(asset_server: &AssetServer) -> impl Bundle + use<> {
(
Node {
width: Val::Percent(100.0),
height: Val::Percent(100.0),
align_items: AlignItems::Center,
justify_content: JustifyContent::Center,
..default()
},
children![(
Node {
width: Val::Px(0.),
height: Val::Px(0.),
overflow: Overflow::clip(),
..default()
},
children![(
//OverrideClip,
Button,
Node {
position_type: PositionType::Absolute,
width: Val::Px(150.0),
height: Val::Px(65.0),
border: UiRect::all(Val::Px(5.0)),
// horizontally center child text
justify_content: JustifyContent::Center,
// vertically center child text
align_items: AlignItems::Center,
..default()
},
BorderColor::all(Color::WHITE),
BorderRadius::MAX,
BackgroundColor(Color::BLACK),
children![(
Text::new("Button"),
TextFont {
font: asset_server.load("fonts/FiraSans-Bold.ttf"),
font_size: 33.0,
..default()
},
TextColor(Color::srgb(0.9, 0.9, 0.9)),
TextShadow::default(),
)]
)],
)],
)
}
```
On main the button ignores interactions, with this PR it should respond
correctly.
---------
Co-authored-by: Alice Cecile <alice.i.cecile@gmail.com>
Noticed that we're converting perceptual_roughness to roughness for SSAO
specular occlusion up here, _but_ that happens _before_ we sample the
metallic_roughness texture map. So we're using the wrong roughness. I
assume this is a bug and was not intentional.
Suggest reviewing while hiding the whitespace diff.
## Objective
Fixes#20058
## Solution
Fix the `dynamic_offsets` array being too small if a mesh has morphs and
skins and motion blur, and the renderer isn't using storage buffers
(i.e. WebGL2). The bug was introduced in #13572.
## Testing
- Minimal repro: https://github.com/M4tsuri/bevy_reproduce.
- Also examples `animated_mesh`, `morph_targets`,
`test_invalid_skinned_meshes`.
- As far as I can tell Bevy doesn't have any examples or tests that can
repro the problem combination.
Tested with WebGL and native, Win10/Chrome/Nvidia.
# Objective
- `component.rs` is becoming quite big, let's split it.
## Solution
- Split `component.rs` into the following:
- `component/mod.rs`: the new root of the `component` module: contains
the `Component` trait and a few other very generic items
- `component/clone.rs`: contains component cloning related items
- `component/tick.rs`: contains `Tick` and other tick related items
- future possibilities: move these into `change_detection`; however I
wanted to keep this PR without breaking changes
- `component/register.rs`: contains component registration (included
queued) related items
- `component/required.rs`: contains items and functions for properly
computing required components
- `component/info.rs`: contains items for storing component info and
metadata in the `World`
# Objective
The current Entity Debug impl prints the bit representation. This is an
"overshare". Debug is in many ways the primary interface into Entity, as
people derive Debug on their entity-containing types when they want to
inspect them. The bits take up too much space in the console and
obfuscate the useful information (entity index and generation).
## Solution
Use the Display implementation in Debug as well. Direct people
interested in bits to `Entity::to_bits` in the docs.
# Objective
- Rebase of https://github.com/bevyengine/bevy/pull/12561 , note that
this is blocked on "up-streaming
[iyes_perf_ui](https://crates.io/crates/iyes_perf_ui)" , but that work
seems to also be stalled
> Frame time is often more important to know than FPS but because of the
temporal nature of it, just seeing a number is not enough. Seeing a
graph that shows the history makes it easier to reason about
performance.
## Solution
> This PR adds a bar graph of the frame time history.
>
> Each bar is scaled based on the frame time where a bigger frame time
will give a taller and wider bar.
>
> The color also scales with that frame time where red is at or bellow
the minimum target fps and green is at or above the target maximum frame
rate. Anything between those 2 values will be interpolated between green
and red based on the frame time.
>
> The algorithm is highly inspired by this article:
https://asawicki.info/news_1758_an_idea_for_visualization_of_frame_times
## Testing
- Ran `cargo run --example fps_overlay --features="bevy_dev_tools"`
---------
Co-authored-by: IceSentry <c.giguere42@gmail.com>
Co-authored-by: Alice Cecile <alice.i.cecile@gmail.com>
# Objective
The extraction systems for materials, meshes, and skins previously
iterated over `RemovedComponents<ViewVisibility>` in addition to more
specific variants like `RemovedComponents<MeshMaterial3d<M>>`. This
caused each system to loop through and check many irrelevant despawned
entities—sometimes multiple times. With many material types, this
overhead added up and became noticeable in frames with many despawns.
<img width="1091" alt="Screenshot 2025-02-21 at 10 28 01 AM"
src="https://github.com/user-attachments/assets/63fec1c9-232c-45f6-9150-daf8751ecf85"
/>
## Solution
This PR removes superfluous `RemovedComponents` iteration for
`ViewVisibility` and `GlobalTransform`, ensuring that we only iterate
over the most specific `RemovedComponents` relevant to the system (e.g.,
material components, mesh components). This is guaranteed to match what
the system originally collected.
### Before (red) / After (yellow):
<img width="838" alt="Screenshot 2025-02-21 at 10 46 17 AM"
src="https://github.com/user-attachments/assets/0e06b06d-7e91-4da5-a919-b843eb442a72"
/>
Log plot to highlight the long tail that this PR is addressing.
# Objective
Make the schedule graph code more understandable, and replace some
panics with `Result`s.
I found the `check_edges` and `check_hierarchy` functions [a little
confusing](https://github.com/bevyengine/bevy/pull/19352#discussion_r2181486099),
as they combined two concerns: Initializing graph nodes for system sets,
and checking for self-edges on system sets. It was hard to understand
the self-edge checks because it wasn't clear what `NodeId` was being
checked against! So, let's separate those concerns, and move them to
more appropriate locations.
Fix a bug where `schedule.configure_sets((SameSet, SameSet).chain());`
would panic with an unhelpful message: `assertion failed: index_a <
index_b`.
## Solution
Remove the `check_edges` and `check_hierarchy` functions, separating the
initialization behavior and the checking behavior and moving them where
they are easier to understand.
For initializing graph nodes, do this on-demand using the `entry` API by
replacing later `self.system_set_ids[&set]` calls with a
`self.system_sets.get_or_add_set(set)` method. This should avoid the
need for an extra pass over the graph and an extra lookup.
Unfortunately, implementing that method directly on `ScheduleGraph`
leads to borrowing errors as it borrows the entire `struct`. So, split
out the collections managing system sets into a separate `struct`.
For checking self-edges, move this check later so that it can be
reported by returning a `Result` from `Schedule::initialize` instead of
having to panic in `configure_set_inner`. The issue was that `iter_sccs`
does not report self-edges as cycles, since the resulting components
only have one node, but that later code assumes all edges point forward.
So, check for self-edges directly, immediately before calling
`iter_sccs`.
This also ensures we catch *every* way that self-edges can be added. The
previous code missed an edge case where `chain()`ing a set to itself
would create a self-edge and would trigger a `debug_assert`.
# Objective
- `bevy_math` allows the `dead_code` lint on some private structs when
`alloc` is not enabled
- allowing lints is not allowed, we should use expect
## Solution
- Don't even compile the code if its expected to be dead instead of
allowing or expecting the lint
# Objective
- `bevy_render_macros` fails to build on its own:
```
error[E0432]: unresolved import `syn::Pat`
--> crates/bevy_render/macros/src/specializer.rs:13:69
|
13 | DeriveInput, Expr, Field, Ident, Index, Member, Meta, MetaList, Pat, Path, Token, Type,
| ^^^
| |
| no `Pat` in the root
| help: a similar name exists in the module: `Path`
|
note: found an item that was configured out
--> /home/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/syn-2.0.104/src/lib.rs:485:15
|
485 | FieldPat, Pat, PatConst, PatIdent, PatLit, PatMacro, PatOr, PatParen, PatPath, PatRange,
| ^^^
note: the item is gated behind the `full` feature
--> /home/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/syn-2.0.104/src/lib.rs:482:7
|
482 | #[cfg(feature = "full")]
| ^^^^^^^^^^^^^^^^
```
## Solution
- Enable the `full` feature of `syn`
# Objective
- bevy_log is not "others". it's part of Bevy
## Solution
- Move it
- Also use the same format as other Bevy dependency (path first) as I
use it in some regexes to run tests on the repo
# Objective
- Related to #19024.
## Solution
- This is a mix of several ways to get rid of weak handles. The primary
strategy is putting strong asset handles in resources that the rendering
code clones into its pipelines (or whatever).
- This does not handle every remaining case, but we are slowly clearing
them out.
## Testing
- `anti_aliasing` example still works.
- `fog_volumes` example still works.
# Objective
The false and true arguments for the select statement in `lerp_hue_long`
are misordered, resulting in it taking the wrong hue path:
## Solution
Swap the arguments around.
Also fixed another case I found during testing. The hue was interpolated
even when it is undefined for one of the endpoints (for example in a
gradient from black to yellow). In those cases it shouldn't interpolate,
instead it should return the hue of the other end point.
## Testing
I added a `linear_gradient` module to the testbed `ui` example, run
with:
```
cargo run --example testbed_ui
```
In the linear gradients screen (press space to switch) it shows a column
of red to yellow linear gradients. The last gradient in the column uses
the OKLCH long path, which should look like this:
matching the same gradient in CSS:
https://jsfiddle.net/fevshkdy/14/
if the correct hue path is chosen.
# Objective
`ThreadLocal::<T>::default()` doesn't require `T: Default`, so
`Parallel<T>` shouldn't require it either.
## Solution
- Replaced the `Default` derive with a manually specified impl.
- Added `Parallel::borrow_local_mut_or` as a non-`T: Default`-requiring
alternative to `borrow_local_mut`.
- Added `Parallel::scope_or` as a non-`T: Default`-requiring alternative
to `scope`.
# Objective
Add interpolation in HSL and HSV colour spaces for UI gradients.
## Solution
Added new variants to `InterpolationColorSpace`: `Hsl`, `HslLong`,
`Hsv`, and `HsvLong`, along with mix functions to the `gradients` shader
for each of them.
#### Limitations
* Didn't include increasing and decreasing path support, it's not
essential and can be done in a follow up if someone feels like it.
* The colour conversions should really be performed before the colours
are sent to the shader but it would need more changes and performance is
good enough for now.
## Testing
```cargo run --example gradients```
## Objective
Fixes#19884.
## Solution
- Add an internal entity command `insert_with`, which takes a function
returning a component and checks if the component would actually be
inserted before invoking the function.
- Add the same check to `insert_from_world`, since it's a similar
situation.
- Update the `or_insert_with`, `or_try_insert_with`, and `or_default`
methods on `EntityEntryCommands` to use the new command.
Since the function/closure returning the component now needs to be sent
into the command (rather than being invoked before the command is
created), the function now has `Send + 'static` bounds. Pretty typical
for command stuff, but I don't know how/if it'll affect existing users.
---------
Co-authored-by: Chris Russell <8494645+chescock@users.noreply.github.com>
# Objective
To implement fmt::Display for the direction types. The reason that this
would be a good addition is that I often find myself using println! to
debug things with directions and adding the extra ":?" was getting a
little annoying. It would also be better for any potential CLI apps that
might need to output a direction.
## Solution
Copied glam's implementation of Display for each length of direction.
I.E Vec3's display for Dir3.
## Testing
- Did you test these changes? If so, how?
Yes, I wrote a little script that printed out the different directions
and compared it to their vector counterparts.
Here it is if anyone's interested
```
use bevy_math::*;
fn main() {
let dir2 = Dir2::from_xy(0.0, 1.0).unwrap();
let dir3 = Dir3::from_xyz(0.0, 1.0, 0.0).unwrap();
let dir3a = Dir3A::from_xyz(0.0, 1.0, 0.0).unwrap();
let dir4 = Dir4::from_xyzw(0.0, 1.0, 0.0, 0.0).unwrap();
let vec2 = Vec2::new(0.0, 1.0);
let vec3 = Vec3::new(0.0, 1.0, 0.0);
let vec4 = Vec4::new(0.0, 1.0, 0.0, 1.0);
println!("{dir2} {dir3} {dir3a} {dir4}");
println!("{vec2}, {vec3}, {vec4}")
}
```
- Are there any parts that need more testing?
Perhaps
# Objective
#19649 introduced new `*_if_new` and `*_by_bundle_id_*` variations to
`EntityClonerBuilder` filtering functionality, which resulted in
increase in method permutations - there are now 8 allow variants to
support various id types and 2 different insert modes.
## Solution
This PR introduces a new trait `FilterableIds` to unify all id types and
their `IntoIterator` implementations, which is somewhat similar to
`WorldEntityFetch`. It supports `TypeId`, `ComponentId` and `BundleId`,
allowing us to reduce the number of `allow` methods to 4: `allow`,
`allow_if_new`, `allow_by_ids`, `allow_by_ids_if_new`. The function
signature is a bit less readable now, but the docs mention types that
can be passed in.
## Testing
All existing tests pass, performance is unchanged.
---------
Co-authored-by: urben1680 <55257931+urben1680@users.noreply.github.com>
# Objective
There is a pattern that appears in multiple places, involving
`reflect_clone`, followed by `take`, followed by `map_err` that produces
a `FailedDowncast` in a particular form.
## Solution
Introduces `reflect_clone_and_take`, which factors out the repeated
code.
## Testing
`cargo run -p ci`
---------
Co-authored-by: Alice Cecile <alice.i.cecile@gmail.com>
# Objective
All the derived reflection methods currently have multiple trait bounds
on non-generic field types, which serve no purpose. The are emitted
because "emit bounds on all fields" is easier than "emit bounds on
fields that need them". But improving things isn't too hard.
Similarly, lots of useless `Any + Send + Sync` bounds exist on
non-generic types.
Helps a lot with #19873.
## Solution
Remove the unnecessary bounds by only emitting them if the relevant type
is generic.
## Testing
I used `cargo expand` to confirm the unnecessary bounds are no longer
produced.
`-Zmacro-stats` output tells me this reduces the size of the `Reflect`
code produced for `bevy_ui` by 21.2%.
Fixes#19594
The exact problem is described in that issue.
I improved the docs to guide anyone who has the the same issue I had.
I kept myself minimal, since the problem is relatively niche, hopefully
it will be enough if anyone else has that problem
I noticed that the `SpatialListener` asks to have a `Transform`
attached. It seemed weird that we didnt just use a require macro, so i
went ahead and did that
I also tweaked the system that plays audio to use a `&GlobalTransform`
instead of an `Option<&GlobalTransform>`
# Objective
`PickingPlugin` and `PointerInputPlugin` were kinda weird being both a
plugin and a resource.
## Solution
Extract the resource functionality of `PickingPlugin` and
`PointerInputPlugin` into new resources
## Testing
`mesh_picking` and `sprite_picking`
---------
Co-authored-by: Alice Cecile <alice.i.cecile@gmail.com>
Co-authored-by: Jan Hohenheim <jan@hohenheim.ch>
# Objective
- nice bevy::camera bevy::mesh bevy::light imports
- skip bevy_light in 2d
## Solution
- add optional crates to internal
- make light only included when building pbr
## Testing
- 3d_scene
# Objective
Fixes#16525Fixes#19710
## Solution
Not allocating a mesh if it is empty.
## Testing
I tested using the following minimum repro from #16525
```rust
use bevy::{asset::RenderAssetUsages, prelude::*, render::mesh::PrimitiveTopology};
fn main() {
App::new()
.add_plugins(DefaultPlugins)
.add_systems(Startup, setup)
.run();
}
fn setup(
mut commands: Commands,
mut meshes: ResMut<Assets<Mesh>>,
mut materials: ResMut<Assets<ColorMaterial>>,
) {
commands.spawn(Camera2d);
let mesh = Mesh::new(
PrimitiveTopology::TriangleList,
RenderAssetUsages::default(),
);
commands.spawn((
Mesh2d(meshes.add(mesh)),
MeshMaterial2d(materials.add(Color::hsl(180.0, 0.95, 0.7))),
));
}
```
I was able to test on webgl2 and windows native and the issue seems to
be resolved. I am not familiar with how mesh rendering works and feel
like just skipping meshes should cause issues but I did not notice any.
# Objective
- Fixes#19910.
## Solution
- First, allow extraction function to be FnMut instead of Fn. FnMut is a
superset of Fn anyway, and we only ever call this function once at a
time (we would never call this in parallel for different pairs of worlds
or something).
- Run the `RenderStartup` schedule in the extract function with a flag
to only do it once.
- Remove all the `MainRender` stuff.
One sad part here is that now the `RenderStartup` blocks extraction. So
for pipelined rendering, our simulation will be blocked on the first
frame while we set up all the rendering resources. I don't see this as a
big loss though since A) that is fundamentally what we want here -
extraction **has to** run after `RenderStartup`, and the only way to do
better is to somehow run `RenderStartup` in parallel with the first
simulation frame, and B) without `RenderStartup` the **entire** app was
blocked on initializing render resources during Plugin construction - so
we're not really losing anything here.
## Testing
- I ran the `custom_post_processing` example (which was ported to use
`RenderStartup` in #19886) and it still works.
# Objective
- the fast inverse sqrt trick hasnt been useful on modern hardware for
over a decade now
## Solution
- just use sqrt, modern hardware has a dedicated instruction which will
outperform approximations both in efficiency and accuracy
## Testing
- ran `atmosphere`
# Objective
- make lights usable without bevy_render
## Solution
- make a new crate for lights to live in
## Testing
- 3d_scene, lighting, volumetric_fog, ssr, transmission, pcss,
light_textures
Note: no breaking changes because of re-exports, except for light
textures, which were introduced this cycle so it doesn't matter anyways
# Objective
- Calculating gradients in variable-termination loop is bad, and we dont
need to here
## Solution
- Sample mip 0 always
## Testing
- volumetric_fog example
# Objective
- prepare bevy_light for split
- make struct named better
- put it where it belongs
## Solution
- do those things
## Testing
- 3d_scene, lighting
---------
Co-authored-by: Alice Cecile <alice.i.cecile@gmail.com>
# Objective
- prepare bevy_light for split
## Solution
- split render world extract related cluster code from main world ecs
stuff
re-exports make this not breaking
# Objective
- for smallvec some crates specify default features false, other dont.
turns out we dont need them
## Solution
- remove
## Testing
- 3d_scene
# Objective
- prepare bevy_light for split
## Solution
- extract cascade module (this is not strictly necessary for bevy_light)
- clean up imports to be less globby and tangled
- move light specific stuff into light modules
- move light system and type init from pbr into new LightPlugin
## Testing
- 3d_scene, lighting
NOTE TO REVIEWERS: it may help to review commits independently.
# Objective
- Make bevy_light possible by making it possible to split out
clusterable into bevy_camera
## Solution
- move ClusteredDecal to cluster module
- Depends on #19957 (because of the imports shuffling around) (draft
until thats merged)
## Testing
- 3d_scene runs
Note: no breaking changes thanks to re-exports
# Objective
- Make bevy_light possible
## Solution
- Move non-light stuff out of light module (its a marker for whether a
material should cast shadows: thats a material property not a light
property)
## Testing
- 3d_scene runs
# Objective
- Make bevy_light possible
## Solution
- Move some stuff it needs out of somewhere it cant depend on. Plus it
makes sense, cubemap stuff goes next to cubemap stuff.
## Testing
- 3d_scene runs
Note: no breaking changes thanks to re-exports
# Objective
- Make bevy_light possible by making it possible to split out
clusterable into bevy_camera
## Solution
- Move some stuff so i can split it out cleanly.
## Testing
- 3d_scene runs
# Objective
- Make bevy_light possible by making it possible to split out
clusterable into bevy_camera
## Solution
- Move cubemap stuff next to cubemap stuff.
## Testing
- 3d_scene runs
Note: no breaking changes thanks to re-exports
# Objective
- Make bevy_light possible
## Solution
- Move some stuff it needs out of somewhere it cant depend on. Plus it
makes sense, spotlight stuff goes in spotlight file.
## Testing
- 3d_scene runs
Note: no breaking changes thanks to re-exports
# Objective
- Make bevy_light possible by making it possible to split out
clusterable into bevy_camera
## Solution
- Use a resource to store cluster settings instead of recalculating it
every time from the render adapter/device
## Testing
- 3d_scene runs
# Objective
- define scenes without bevy_render
## Solution
- Move Camera2d/3d components out of bevy_core_pipeline
## Testing
- 3d_scene runs fine
Note: no breaking changes thanks to re-exports
# Objective
- Make bevy_light possible
## Solution
- Move some stuff it needs out of somewhere it cant depend on. Plus it
makes sense, visibility stuff goes in visibility.
## Testing
- 3d_scene runs
Note: no breaking changes thanks to re-exports
# Objective
- Progress towards #19887.
## Solution
- Convert `FromWorld` impls into systems that run in `RenderStartup`.
- Move `UiPipeline` init to `build_ui_render` instead of doing it
separately in `finish`.
Note: I am making several of these systems pub so that users could order
their systems relative to them. This is to match the fact that these
types previously were FromWorld so users could initialize them.
## Testing
- Ran `ui_material`, `ui_texture_slice`, `box_shadow`, and `gradients`
examples and it still worked.
# Objective
- get closer to being able to load gltfs without using bevy_render
## Solution
- Split bevy_camera out of bevy_render
- Builds on #19943
- Im sorry for the big diff, i tried to minimize it as much as i can by
using re-exports. This also prevents most breaking changes, but there
are still a couple.
## Testing
- 3d_scene looks good
# Objective
- The usage of ComponentId is quite confusing: events are not
components. By newtyping this, we can prevent stupid mistakes, avoid
leaking internal details and make the code clearer for users and engine
devs reading it.
- Adopts https://github.com/bevyengine/bevy/pull/19755
---------
Co-authored-by: oscar-benderstone <oscarbenderstone@gmail.com>
Co-authored-by: Oscar Bender-Stone <88625129+oscar-benderstone@users.noreply.github.com>
# Objective
- I accidentally left a `register_component_hooks` without actually
adding a hook and didnt notice
## Solution
- mark it must_use so it doesnt happen to other people (maybe this is
just skill issue on me though)
# Objective
- another step towards splitting out bevy_camera, this is needed by
visibility systems
## Solution
- move mesh stuff into mesh place
## Testing
- 3d_scene looks fine
No migration needed because of the re-export, that can be another PR
after i split bevy_camera
# Objective
Move Bevy UI's rendering into a dedicated crate.
Motivations:
* Allow the UI renderer to be used with other UI frameworks than
`bevy_ui`.
* Allow for using alternative renderers like Vello with `bevy_ui`.
* It's difficult for rendering contributors to make changes and
improvements to the UI renderer as it requires in-depth knowledge of the
UI implementation.
## Solution
Move the `render` and `ui_material` modules from `bevy_ui` into a new
crate `bevy_ui_render`.
## Testing
Important examples to check are `testbed_ui`, `testbed_full_ui`,
`ui_material`, `viewport_node` and `gradients`.
---------
Co-authored-by: Alice Cecile <alice.i.cecile@gmail.com>
# Objective
The generated `GetTypeRegistration::get_type_registration` method has an
unnecessary `allow(unused_mut)` attribute. It used to be necessary
because it was possible for `registration` to not be modified, but now
there is always at least one modification.
## Solution
Remove the attribute.
## Testing
I checked the `cargo expand` output.
# Objective
```
2025-07-03T11:48:34.039501Z ERROR panic: thread 'IO Task Pool (6)' panicked at 'byte index 9 is not a char boundary; it is inside '个' (bytes 7..10) of `展示_个人收款码.png`': [...]\crates\bevy_asset\src\path.rs:475
```
## Solution
char_indices
Adds support for:
- `WGPU_ADAPTER_NAME` which will attempt to select a specific adapter
with a given name.
- `WGPU_FORCE_FALLBACK_ADAPTER` which will force fallback to a fallback
(software) renderer, if available.
The first has higher specificity than the second.
# Objective
Allow combinator and pipe systems to delay validation of the second
system, while still allowing the second system to be skipped.
Fixes#18796
Allow fallible systems to be used as one-shot systems, reporting errors
to the error handler when used through commands.
Fixes#19722
Allow fallible systems to be used as run conditions, including when used
with combinators. Alternative to #19580.
Always validate parameters when calling the safe
`run_without_applying_deferred`, `run`, and `run_readonly` methods on a
`System`.
## Solution
Have `System::run_unsafe` return a `Result`.
We want pipe systems to run the first system before validating the
second, since the first system may affect whether the second system has
valid parameters. But if the second system skips then we have no output
value to return! So, pipe systems must return a `Result` that indicates
whether the second system ran.
But if we just make pipe systems have `Out = Result<B::Out>`, then
chaining `a.pipe(b).pipe(c)` becomes difficult. `c` would need to accept
the `Result` from `a.pipe(b)`, which means it would likely need to
return `Result` itself, giving `Result<Result<Out>>`!
Instead, we make *all* systems return a `Result`! We move the handling
of fallible systems from `IntoScheduleConfigs` and `IntoObserverSystem`
to `SystemParamFunction` and `ExclusiveSystemParamFunction`, so that an
infallible system can be wrapped before being passed to a combinator.
As a side effect, this enables fallible systems to be used as run
conditions and one-shot systems.
Now that the safe `run_without_applying_deferred`, `run`, and
`run_readonly` methods return a `Result`, we can have them perform
parameter validation themselves instead of requiring each caller to
remember to call them. `run_unsafe` will continue to not validate
parameters, since it is used in the multi-threaded executor when we want
to validate and run in separate tasks.
Note that this makes type inference a little more brittle. A function
that returns `Result<T>` can be considered either a fallible system
returning `T` or an infallible system returning `Result<T>` (and this is
important to continue supporting `pipe`-based error handling)! So there
are some cases where the output type of a system can no longer be
inferred. It will work fine when directly adding to a schedule, since
then the output type is fixed to `()` (or `bool` for run conditions).
And it will work fine when `pipe`ing to a system with a typed input
parameter.
I used a dedicated `RunSystemError` for the error type instead of plain
`BevyError` so that skipping a system does not box an error or capture a
backtrace.
...which previously used a HashSet, whos iter has no ordering guarantee
fixes#19687
i also discovered that the asserted order in the unit test is reversed,
so i fixed that. I dont know if that reversed order is intentional
Edit: i referenced the wrong issue oops
# Objective
- First step towards #279
## Solution
Makes the necessary internal data structure changes in order to allow
system removal to be added in a future PR: `Vec`s storing systems and
system sets in `ScheduleGraph` have been replaced with `SlotMap`s.
See the included migration guide for the required changes.
## Testing
Internal changes only and no new features *should* mean no new tests are
requried.
- renamed `spec_v2` related modules, that commit slipped through the
other pr #17373
- revised struct and trait docs for clarity, and gave a short intro to
specialization
- turns out the derive macro was broken, fixed that too
# Objective
Current implementation of `Sprite::compute_pixel_space_point` always
returns the sprite centre as an `Ok` point when the `custom_size` is set
to `Vec2::ZERO`. This leads to unexpected behaviour. For example, it
causes these sprites to block all interactions with other sprites in the
picking backend (under default settings). This small PR:
- Fixes sprite pixel space point computation for sprites with zero
custom_size
- Resolves issue #19880.
## Solution
We handle the zero custom_size case explicitly and return
`Err(point_relative_to_sprite_center)` instead of
`Ok(point_relative_to_texture)`.
## Testing
Implemented a new test for zero custom_size sprites within the
`bevy_sprite::sprite` module. Also verified that the example from issue
#19880 is behaving as expected.
No further testing is required.
- How can other people (reviewers) test your changes? Is there anything
specific they need to know?
Can run the simple application example from the linked issue. Or
evaluate the implemented test.
---------
Co-authored-by: James Lucas <jalucas@nvidia.com>
## Objective
Add a test that reproduces #11111 (and partially #18267). The bug is
that asset loader settings are effectively ignored if the same asset is
loaded multiple times with different settings.
## Solution
Add a unit test to `bevy_assets/lib.rs`. The test will be marked as
`#[ignore]` until #11111 is fixed.
```rust
// Load the same asset with different settings.
let handle_1 = load(asset_server, "test.u8", 1);
let handle_2 = load(asset_server, "test.u8", 2);
// Handles should be different.
assert_ne!(handle_1, handle_2);
```
## Concerns
I'm not 100% sure that the current behaviour is actually broken - I
can't see anything in the asset system design docs that explicitly says
different settings should create different asset ids.
UPDATE: Sentiment from issue comments and discord varies between "bug"
and "undesirable consequence of design decisions, alternatives should be
explored". So I've concluded that the test is valid and desirable.
## Testing
```sh
cargo test -p bevy_asset --features multi_threaded
# Or to repro the issue:
cargo test -p bevy_asset --features multi_threaded -- --ignored
```
# Objective
- Progress towards #19887.
## Solution
- Rewrite the FromWorld impls to systems that create the pipeline
resources.
## Testing
- Ran the `anti_aliasing` example and it still works.
# Objective
- Progress towards #19887.
## Solution
- Convert `FromWorld` impls into systems that run in `RenderStartup`.
- Add an ordering constraint to ensure that necessary resources exist.
## Testing
- Ran `2d_gizmos` and `3d_gizmos` examples and it still worked.
# Objective
for `BufferUsages::STORAGE` on webgpu (and maybe other contexts), buffer
sizes must be a multiple of 4. the skin uniform buffer starts at 16384
then increases by 1.5x, which eventually hits a number which isn't
## Solution
`.next_multiple_of(4)`
## Objective
Allow users to directly call ease functions rather than going through
the `EaseFunction` struct. This is less verbose and more efficient when
the user doesn't need the data-driven aspects of `EaseFunction`.
## Background
`EaseFunction` is a flexible and data-driven way to apply easing. But
that has a price when a user just wants to call a specific ease
function:
```rust
EaseFunction::SmoothStep.sample(t);
```
This is a bit verbose, but also surprisingly inefficient. It calls the
general `EaseFunction::eval`, which won't be inlined and adds an
unnecessary branch. It can also increase code size since it pulls in all
ease functions even though the user might only require one. As far as I
can tell this is true even with `opt-level = 3` and `lto = "fat"`.
```asm
; EaseFunction::SmoothStep.sample_unchecked(t)
lea rcx, [rip + __unnamed_2] ; Load the disciminant for EaseFunction::SmoothStep.
movaps xmm1, xmm0
jmp bevy_math::curve::easing::EaseFunction::eval
```
## Solution
This PR adds a struct for each ease function. Most are unit structs, but
a couple have parameters:
```rust
SmoothStep.sample(t);
Elastic(50.0).sample(t);
Steps(4, JumpAt::Start).sample(t)
```
The structs implement the `Curve<f32>` trait. This means they fit into
the broader `Curve` system, and the user can choose between `sample`,
`sample_unchecked`, and `sample_clamped`. The internals are a simple
function call so the compiler can easily estimate the cost of inlining:
```asm
; SmoothStep.sample_unchecked(t)
movaps xmm1, xmm0
addss xmm1, xmm0
movss xmm2, dword ptr [rip + __real@40400000] ; 3.0
subss xmm2, xmm1
mulss xmm2, xmm0
mulss xmm0, xmm2
```
In a microbenchmark this is around 4x faster. If inlining permits
auto-vectorization then it's 20-50x faster, but that's a niche case.
Adding unit structs is also a small boost to discoverability - the unit
struct can be found in VS Code via "Go To Symbol" -> "smoothstep", which
doesn't find `EaseFunction::SmoothStep`.
### Concerns
- While the unit structs have advantages, they add a lot of API surface
area.
- Another option would have been to expose the underlying functions.
- But functions can't implement the `Curve` trait.
- And the underlying functions are unclamped, which could be a footgun.
- Or there have to be three functions to cover unchecked/checked/clamped
variants.
- The unit structs can't be used with `EasingCurve`, which requires
`EaseFunction`.
- This might confuse users and limit optimisation.
- Wrong: `EasingCurve::new(a, b, SmoothStep)`.
- Right: `EasingCurve::new(a, b, EaseFunction::SmoothStep)`.
- In theory `EasingCurve` could be changed to support any `Curve<f32>`
or a more limited trait.
- But that's likely to be a breaking change and raises questions around
reflection and reliability.
- The unit structs don't have serialization.
- I don't know much about the motivations/requirements for
serialization.
- Each unit struct duplicates the documentation of `EaseFunction`.
- This is convenient for the user, but awkward for anyone updating the
code.
- Maybe better if each unit struct points to the matching
`EaseFunction`.
- Might also make the module page less intimidating (see screenshot).
## Testing
```
cargo test -p bevy_math
```
A few versions ago, wgpu made it possible to set shader entry point to
`None`, which will select the correct entry point in file where only a
single entrypoint is specified. This makes it possible to implement
`Default` for pipeline descriptors. This PR does so and attempts to
`..default()` everything possible.
# Objective
Generated `from_reflect` methods use closures in a weird way, e.g.:
```rust
x: (|| {
<f32 as ::bevy::reflect::FromReflect>::from_reflect(
::bevy::reflect::Struct::field(__ref_struct, "x")?,
)
})()?,
```
The reason for this is because when `#[reflect(Default)]` is used, you
instead get stuff like this:
```rust
if let ::core::option::Option::Some(__field) = (|| {
<f32 as ::bevy::reflect::FromReflect>::from_reflect(
::bevy::reflect::Struct::field(__ref_struct, "x")?,
)
})() {
__this.x = __field;
}
```
and the closure is necessary to contain the scope of the `?`. But the
first case is more common.
Helps with #19873.
## Solution
Avoid the closure in the common case.
## Testing
I used cargo expand to confirm the closures are no longer produced in
the common case.
`-Zmacro-stats` output tells me this reduces the size of the `Reflect`
code produced for `bevy_ui` by 0.5%.
# Objective
- This unblocks some work I am doing for #19887.
## Solution
- Rename `RenderGraphApp` to `RenderGraphExt`.
- Implement `RenderGraphExt` for `World`.
- Change `SubApp` and `App` to call the `World` impl.
# Objective
`WhereClauseOption` contains a reference to a `ReflectMeta`. Oddly
enough, a bunch of functions that take a `WhereClauseOption` argument
also take a `ReflectMeta` reference argument, which is exactly the same
as the reference in the `WhereClauseOption`.
## Solution
This commit removes the redundant `ReflectMeta` argument from these
functions. This requires adding a `WhereClauseOption::meta` getter
method.
## Testing
`cargo run -p ci`
# Objective
- `MapEntities` is not implemented for arrays, `HashMap`, `BTreeMap`,
and `IndexMap`.
## Solution
- Implement `MapEntities` for arrays, `HashMap`, `BTreeMap`, `IndexMap`
## Testing
- I didn't add a test for this as the implementations seems pretty
trivial
# Objective
Concise syntax docs on `Component`/`Event` derives. Partial fix for
#19537.
## Solution
Only document syntax. The doc tests are set to ignore because the macro
relies on the presence of `bevy_ecs`.
# Objective
- Progress towards #19024.
## Solution
- Remove `Handle::Weak`!
If users were relying on `Handle::Weak` for some purpose, they can
almost certainly replace it with raw `AssetId` instead. If they cannot,
they can make their own enum that holds either a Handle or an AssetId.
In either case, we don't need weak handles!
Sadly we still need Uuid handles since we rely on them for "default"
assets and "invalid" assets, as well as anywhere where a component wants
to impl default with a non-defaulted asset handle. One step at a time
though!
# Objective
- PrepassPipelineInternal used to exist to optimize compile time and
binary size when PrepassPipeline was generic over the material.
- After #19667, PrepassPipeline is no longer generic!
## Solution
- Flatten all the fields of `PrepassPipelineInternal` into
`PrepassPipeline`.
# Objective
During the migration to required components a lot of things were changed
around and somehow the draw order for some UI elements ended up
depending on the system ordering in `RenderSystems::Queue`, which can
sometimes result in the elements being drawn in the wrong order.
Fixes#19674
## Solution
* Added some more `stack_z_offsets` constants and used them to enforce
an explicit ordering.
* Removed the `stack_index: u32` field from `ExtractedUiNodes` and
replaced it with a `z_order: f32` field.
These changes should fix all the ordering problems.
## Testing
I added a nine-patched bordered node with a navy background color to the
slice section of the `testbed_ui` example.
The border should always be drawn above the background color.
# Objective
Change `ScrollPosition` to newtype `Vec2`. It's easier to work with a
`Vec2` wrapper than individual fields.
I'm not sure why this wasn't newtyped to start with. Maybe the intent
was to support responsive coordinates eventually but that probably isn't
very useful or straightforward to implement. And even if we do want to
support responsive coords in the future, it can newtype `Val2`.
## Solution
Change `ScrollPosition` to newtype `Vec2`.
Also added some extra details to the doc comments.
## Testing
Try the `scroll` example.
---------
Co-authored-by: Alice Cecile <alice.i.cecile@gmail.com>
Some misc cleanup in preparation for future PRs.
* Merge reservoir.wgsl with restir_di.wgsl, as the reservoir is going to
be DI-specific and won't be reused for GI
* Reformulate confidence weights to not multiply by INITIAL_SAMPLES. The
multiplication cancels out, it doesn't matter.
Previously, the specialize/queue systems were added per-material and the
plugin prepass/shadow enable flags controlled whether we added those
systems. Now, we make this a property of the material instance and check
for it when specializing. Fixes
https://github.com/bevyengine/bevy/issues/19850.
# Objective
- This plugin currently does nothing. That's because we add the plugin
to the `RenderApp`. Inside the plugin it then looks for the `RenderApp`
itself, but since it was added **to** the `RenderApp`, it will never
find the `RenderApp`.
## Solution
- Move the plugin into build, and more importantly, add it to the app
not the render_app.
# Objective
Because we want to be able to support more notification options in the
future (in addition to just using registered one-shot systems), the
`Option<SystemId>` notifications have been changed to a new enum,
`Callback`.
@alice-i-cecile
