# Objective
- Fix some small errors in the documentation of the `OwningPtr` struct.
## Solution
- Change comments with 4 slashes `////` to doc comments with 3 slashes `///`.
- Fix typos.
# Objective
Update the `calculate_bounds` system to update `Aabb`s
for entities who've either:
- gotten a new mesh
- had their mesh mutated
Fixes https://github.com/bevyengine/bevy/issues/4294.
## Solution
There are two commits here to address the two issues above:
### Commit 1
**This Commit**
Updates the `calculate_bounds` system to operate not only on entities
without `Aabb`s but also on entities whose `Handle<Mesh>` has changed.
**Why?**
So if an entity gets a new mesh, its associated `Aabb` is properly
recalculated.
**Questions**
- This type is getting pretty gnarly - should I extract some types?
- This system is public - should I add some quick docs while I'm here?
### Commit 2
**This Commit**
Updates `calculate_bounds` to update `Aabb`s of entities whose meshes
have been directly mutated.
**Why?**
So if an entity's mesh gets updated, its associated `Aabb` is properly
recalculated.
**Questions**
- I think we should be using `ahash`. Do we want to do that with a
direct `hashbrown` dependency or an `ahash` dependency that we
configure the `HashMap` with?
- There is an edge case of duplicates with `Vec<Entity>` in the
`HashMap`. If an entity gets its mesh handle changed and changed back
again it'll be added to the list twice. Do we want to use a `HashSet`
to avoid that? Or do a check in the list first (assuming iterating
over the `Vec` is faster and this edge case is rare)?
- There is an edge case where, if an entity gets a new mesh handle and
then its old mesh is updated, we'll update the entity's `Aabb` to the
new geometry of the _old_ mesh. Do we want to remove items from the
`Local<HashMap>` when handles change? Does the `Changed` event give us
the old mesh handle? If not we might need to have a
`HashMap<Entity, Handle<Mesh>>` or something so we can unlink entities
from mesh handles when the handle changes.
- I did the `zip()` with the two `HashMap` gets assuming those would
be faster than calculating the Aabb of the mesh (otherwise we could do
`meshes.get(mesh_handle).and_then(Mesh::compute_aabb).zip(entity_mesh_map...)`
or something). Is that assumption way off?
## Testing
I originally tried testing this with `bevy_mod_raycast` as mentioned in the
original issue but it seemed to work (maybe they are currently manually
updating the Aabbs?). I then tried doing it in 2D but it looks like
`Handle<Mesh>` is just for 3D. So I took [this example](https://github.com/bevyengine/bevy/blob/main/examples/3d/pbr.rs)
and added some systems to mutate/assign meshes:
<details>
<summary>Test Script</summary>
```rust
use bevy::prelude::*;
use bevy::render:📷:ScalingMode;
use bevy::render::primitives::Aabb;
/// Make sure we only mutate one mesh once.
#[derive(Eq, PartialEq, Clone, Debug, Default)]
struct MutateMeshState(bool);
/// Let's have a few global meshes that we can cycle between.
/// This way we can be assigned a new mesh, mutate the old one, and then get the old one assigned.
#[derive(Eq, PartialEq, Clone, Debug, Default)]
struct Meshes(Vec<Handle<Mesh>>);
fn main() {
App::new()
.add_plugins(DefaultPlugins)
.init_resource::<MutateMeshState>()
.init_resource::<Meshes>()
.add_startup_system(setup)
.add_system(assign_new_mesh)
.add_system(show_aabbs.after(assign_new_mesh))
.add_system(mutate_meshes.after(show_aabbs))
.run();
}
fn setup(
mut commands: Commands,
mut meshes: ResMut<Assets<Mesh>>,
mut global_meshes: ResMut<Meshes>,
mut materials: ResMut<Assets<StandardMaterial>>,
) {
let m1 = meshes.add(Mesh::from(shape::Icosphere::default()));
let m2 = meshes.add(Mesh::from(shape::Icosphere {
radius: 0.90,
..Default::default()
}));
let m3 = meshes.add(Mesh::from(shape::Icosphere {
radius: 0.80,
..Default::default()
}));
global_meshes.0.push(m1.clone());
global_meshes.0.push(m2);
global_meshes.0.push(m3);
// add entities to the world
// sphere
commands.spawn_bundle(PbrBundle {
mesh: m1,
material: materials.add(StandardMaterial {
base_color: Color::hex("ffd891").unwrap(),
..default()
}),
..default()
});
// new 3d camera
commands.spawn_bundle(Camera3dBundle {
projection: OrthographicProjection {
scale: 3.0,
scaling_mode: ScalingMode::FixedVertical(1.0),
..default()
}
.into(),
..default()
});
// old 3d camera
// commands.spawn_bundle(OrthographicCameraBundle {
// transform: Transform::from_xyz(0.0, 0.0, 8.0).looking_at(Vec3::default(), Vec3::Y),
// orthographic_projection: OrthographicProjection {
// scale: 0.01,
// ..default()
// },
// ..OrthographicCameraBundle::new_3d()
// });
}
fn show_aabbs(query: Query<(Entity, &Handle<Mesh>, &Aabb)>) {
for thing in query.iter() {
println!("{thing:?}");
}
}
/// For testing the second part - mutating a mesh.
///
/// Without the fix we should see this mutate an old mesh and it affects the new mesh that the
/// entity currently has.
/// With the fix, the mutation doesn't affect anything until the entity is reassigned the old mesh.
fn mutate_meshes(
mut meshes: ResMut<Assets<Mesh>>,
time: Res<Time>,
global_meshes: Res<Meshes>,
mut mutate_mesh_state: ResMut<MutateMeshState>,
) {
let mutated = mutate_mesh_state.0;
if time.seconds_since_startup() > 4.5 && !mutated {
println!("Mutating {:?}", global_meshes.0[0]);
let m = meshes.get_mut(&global_meshes.0[0]).unwrap();
let mut p = m.attribute(Mesh::ATTRIBUTE_POSITION).unwrap().clone();
use bevy::render::mesh::VertexAttributeValues;
match &mut p {
VertexAttributeValues::Float32x3(v) => {
v[0] = [10.0, 10.0, 10.0];
}
_ => unreachable!(),
}
m.insert_attribute(Mesh::ATTRIBUTE_POSITION, p);
mutate_mesh_state.0 = true;
}
}
/// For testing the first part - assigning a new handle.
fn assign_new_mesh(
mut query: Query<&mut Handle<Mesh>, With<Aabb>>,
time: Res<Time>,
global_meshes: Res<Meshes>,
) {
let s = time.seconds_since_startup() as usize;
let idx = s % global_meshes.0.len();
for mut handle in query.iter_mut() {
*handle = global_meshes.0[idx].clone_weak();
}
}
```
</details>
## Changelog
### Fixed
Entity `Aabb`s not updating when meshes are mutated or re-assigned.
# Objective
- Allows conversion of mutable queries to immutable queries.
- Fixes#4606
## Solution
- Add `to_readonly` method on `Query`, which uses `QueryState::as_readonly`
- `AsRef` is not feasible because creation of new queries is needed.
---
## Changelog
### Added
- Allows conversion of mutable queries to immutable queries using `Query::to_readonly`.
# Objective
https://github.com/bevyengine/bevy/pull/4447 adds functions that can fetch resources/components as `*const ()` ptr by providing the `ComponentId`. This alone is not enough for them to be usable safely with reflection, because there is no general way to go from the raw pointer to a `&dyn Reflect` which is the pointer + a pointer to the VTable of the `Reflect` impl.
By adding a `ReflectFromPtr` type that is included in the type type registration when deriving `Reflect`, safe functions can be implemented in scripting languages that don't assume a type layout and can access the component data via reflection:
```rust
#[derive(Reflect)]
struct StringResource {
value: String
}
```
```lua
local res_id = world:resource_id_by_name("example::StringResource")
local res = world:resource(res_id)
print(res.value)
```
## Solution
1. add a `ReflectFromPtr` type with a `FromType<T: Reflect>` implementation and the following methods:
- ` pub unsafe fn as_reflect_ptr<'a>(&self, val: Ptr<'a>) -> &'a dyn Reflect`
- ` pub unsafe fn as_reflect_ptr_mut<'a>(&self, val: PtrMut<'a>) -> &'a mud dyn Reflect`
Safety requirements of the methods are that you need to check that the `ReflectFromPtr` was constructed for the correct type.
2. add that type to the `TypeRegistration` in the `GetTypeRegistration` impl generated by `#[derive(Reflect)]`.
This is different to other reflected traits because it doesn't need `#[reflect(ReflectReflectFromPtr)]` which IMO should be there by default.
Co-authored-by: Jakob Hellermann <hellermann@sipgate.de>
Co-authored-by: Carter Anderson <mcanders1@gmail.com>
after #5355, three methods were added on world:
* `send_event`
* `send_event_batch`
* `send_default_event`
rename `send_default_event` to `send_event_default` for better discoverability
# Objective
- Provide better compile-time errors and diagnostics.
- Add more options to allow more textures types and sampler types.
- Update array_texture example to use upgraded AsBindGroup derive macro.
## Solution
Split out the parsing of the inner struct/field attributes (the inside part of a `#[foo(...)]` attribute) for better clarity
Parse the binding index for all inner attributes, as it is part of all attributes (`#[foo(0, ...)`), then allow each attribute implementer to parse the rest of the attribute metadata as needed. This should make it very trivial to extend/change if needed in the future.
Replaced invocations of `panic!` with the `syn::Error` type, providing fine-grained errors that retains span information. This provides much nicer compile-time errors, and even better IDE errors.
Updated the array_texture example to demonstrate the new changes.
## New AsBindGroup attribute options
### `#[texture(u32, ...)]`
Where `...` is an optional list of arguments.
| Arguments | Values | Default |
|-------------- |---------------------------------------------------------------- | ----------- |
| dimension = "..." | `"1d"`, `"2d"`, `"2d_array"`, `"3d"`, `"cube"`, `"cube_array"` | `"2d"` |
| sample_type = "..." | `"float"`, `"depth"`, `"s_int"` or `"u_int"` | `"float"` |
| filterable = ... | `true`, `false` | `true` |
| multisampled = ... | `true`, `false` | `false` |
| visibility(...) | `all`, `none`, or a list-combination of `vertex`, `fragment`, `compute` | `vertex`, `fragment` |
Example: `#[texture(0, dimension = "2d_array", visibility(vertex, fragment))]`
### `#[sampler(u32, ...)]`
Where `...` is an optional list of arguments.
| Arguments | Values | Default |
|----------- |--------------------------------------------------- | ----------- |
| sampler_type = "..." | `"filtering"`, `"non_filtering"`, `"comparison"`. | `"filtering"` |
| visibility(...) | `all`, `none`, or a list-combination of `vertex`, `fragment`, `compute` | `vertex`, `fragment` |
Example: `#[sampler(0, sampler_type = "filtering", visibility(vertex, fragment)]`
## Changelog
- Added more options to `#[texture(...)]` and `#[sampler(...)]` attributes, supporting more kinds of materials. See above for details.
- Upgraded IDE and compile-time error messages.
- Updated array_texture example using the new options.
# Objective
- With access to `World`, it's not obvious how to send an event.
- This is especially useful if you are writing a `Command` that needs to send an `Event`.
- `Events` are a first-class construct in bevy, even though they are just `Resources` under the hood. Their methods should be discoverable.
## Solution
- Provide a simple helpers to send events through `Res<Events<T>>`.
---
## Changelog
> `send_event`, `send_default_event`, and `send_event_batch` methods added to `World`.
# Objective
Fixes#5362
## Solution
Add the attribute `#[label(ignore_fields)]` for `*Label` types.
```rust
#[derive(SystemLabel)]
pub enum MyLabel {
One,
// Previously this was not allowed since labels cannot contain data.
#[system_label(ignore_fields)]
Two(PhantomData<usize>),
}
```
## Notes
This label makes it possible for equality to behave differently depending on whether or not you are treating the type as a label. For example:
```rust
#[derive(SystemLabel, PartialEq, Eq)]
#[system_label(ignore_fields)]
pub struct Foo(usize);
```
If you compare it as a label, it will ignore the wrapped fields as the user requested. But if you compare it as a `Foo`, the derive will incorrectly compare the inner fields. I see a few solutions
1. Do nothing. This is technically intended behavior, but I think we should do our best to prevent footguns.
2. Generate impls of `PartialEq` and `Eq` along with the `#[derive(Label)]` macros. This is a breaking change as it requires all users to remove these derives from their types.
3. Only allow `PhantomData` to be used with `ignore_fields` -- seems needlessly prescriptive.
---
## Changelog
* Added the `ignore_fields` attribute to the derive macros for `*Label` types.
* Added an example showing off different forms of the derive macro.
<!--
## Migration Guide
> This section is optional. If there are no breaking changes, you can delete this section.
- If this PR is a breaking change (relative to the last release of Bevy), describe how a user might need to migrate their code to support these changes
- Simply adding new functionality is not a breaking change.
- Fixing behavior that was definitely a bug, rather than a questionable design choice is not a breaking change.
-->
# Objective
remove `QF` generics from a bunch of types and methods on query related items. this has a few benefits:
- simplifies type signatures `fn iter(&self) -> QueryIter<'_, 's, Q::ReadOnly, F::ReadOnly>` is (imo) conceptually simpler than `fn iter(&self) -> QueryIter<'_, 's, Q, ROQueryFetch<'_, Q>, F>`
- `Fetch` is mostly an implementation detail but previously we had to expose it on every `iter` `get` etc method
- Allows us to potentially in the future simplify the `WorldQuery` trait hierarchy by removing the `Fetch` trait
## Solution
remove the `QF` generic and add a way to (unsafely) turn `&QueryState<Q1, F1>` into `&QueryState<Q2, F2>`
---
## Changelog/Migration Guide
The `QF` generic was removed from various `Query` iterator types and some methods, you should update your code to use the type of the corresponding worldquery of the fetch type that was being used, or call `as_readonly`/`as_nop` to convert a querystate to the appropriate type. For example:
`.get_single_unchecked_manual::<ROQueryFetch<Q>>(..)` -> `.as_readonly().get_single_unchecked_manual(..)`
`my_field: QueryIter<'w, 's, Q, ROQueryFetch<'w, Q>, F>` -> `my_field: QueryIter<'w, 's, Q::ReadOnly, F::ReadOnly>`
# Objective
- Help user when they need to add both a `TransformBundle` and a `VisibilityBundle`
## Solution
- Add a `SpatialBundle` adding all components
# Objective
- Make `game_of_life.wgsl` easier to read and understand
## Solution
- Remove unused code in the shader
- `location_f32` was unused in `init`
- `color` was unused in `update`
# Objective
- Running examples on windows crash due to full disk
- The prebuild step was not being reused and consuming extra space
## Solution
- Use the exact same command to prebuild to ensure it will be reused
- Also on linux
Birbs no longer bounce too low, not coming close to their true bouncy potential.
Birbs also no longer bonk head when window is smaller. (Will still bonk head when window is made smaller too fast! pls no)
*cough cough*
Make the height of the birb-bounces dependent on the window size so they always bounce elegantly towards the top of the window.
Also no longer panics when closing the window q:
~~Might put a video here if I figure out how to.~~
<sup> rendering video is hard. birbrate go brr
Co-authored-by: devil-ira <justthecooldude@gmail.com>
# Objective
[This unwrap()](de484c1e41/crates/bevy_pbr/src/pbr_material.rs (L195)) in pbr_material.rs will be hit if a StandardMaterial normal_map image has not finished loading, resulting in an error message that is hard to debug.
## Solution
~~This PR improves the error message including a potential indication of why the unwrap() could have panic'd by using expect() instead of unwrap().~~
This PR removes the panic by only proceeding if the image is found.
---
## Changelog
Don't panic when StandardMaterial normal_map images have not finished loading.
# Objective
- Fixes #5338
- Allow the usage of `use bevy::ui::Size` (see migration guide in #4285)
## Solution
- Remove the `use crate::Size` import so that the `pub use geometry::*` import also publicly uses the `Size` struct.
# Objective
- 0.8 is coming soon, and our mikktspace implementation is unsound - see https://github.com/gltf-rs/mikktspace/issues/26
- Best not to ship that
## Solution
- Fix the unsoundness in a minimal way
- Obviously there might be others, but it seems unlikely we have any way to know about those
# Objective
- Add capability to use `Affine3A`s for some `GlobalTransform`s. This allows affine transformations that are not possible using a single `Transform` such as shear and non-uniform scaling along an arbitrary axis.
- Related to #1755 and #2026
## Solution
- `GlobalTransform` becomes an enum wrapping either a `Transform` or an `Affine3A`.
- The API of `GlobalTransform` is minimized to avoid inefficiency, and to make it clear that operations should be performed using the underlying data types.
- using `GlobalTransform::Affine3A` disables transform propagation, because the main use is for cases that `Transform`s cannot support.
---
## Changelog
- `GlobalTransform`s can optionally support any affine transformation using an `Affine3A`.
Co-authored-by: Carter Anderson <mcanders1@gmail.com>
Port changes made to Material in #5053 to Material2d as well.
This is more or less an exact copy of the implementation in bevy_pbr; I
simply pretended the API existed, then copied stuff over until it
started building and the shapes example was working again.
# Objective
The changes in #5053 makes it possible to add custom materials with a lot less boiler plate. However, the implementation isn't shared with Material 2d as it's a kind of fork of the bevy_pbr version. It should be possible to use AsBindGroup on the 2d version as well.
## Solution
This makes the same kind of changes in Material2d in bevy_sprite.
This makes the following work:
```rust
//! Draws a circular purple bevy in the middle of the screen using a custom shader
use bevy::{
prelude::*,
reflect::TypeUuid,
render::render_resource::{AsBindGroup, ShaderRef},
sprite::{Material2d, Material2dPlugin, MaterialMesh2dBundle},
};
fn main() {
App::new()
.add_plugins(DefaultPlugins)
.add_plugin(Material2dPlugin::<CustomMaterial>::default())
.add_startup_system(setup)
.run();
}
/// set up a simple 2D scene
fn setup(
mut commands: Commands,
mut meshes: ResMut<Assets<Mesh>>,
mut materials: ResMut<Assets<CustomMaterial>>,
asset_server: Res<AssetServer>,
) {
commands.spawn_bundle(MaterialMesh2dBundle {
mesh: meshes.add(shape::Circle::new(50.).into()).into(),
material: materials.add(CustomMaterial {
color: Color::PURPLE,
color_texture: Some(asset_server.load("branding/icon.png")),
}),
transform: Transform::from_translation(Vec3::new(-100., 0., 0.)),
..default()
});
commands.spawn_bundle(Camera2dBundle::default());
}
/// The Material2d trait is very configurable, but comes with sensible defaults for all methods.
/// You only need to implement functions for features that need non-default behavior. See the Material api docs for details!
impl Material2d for CustomMaterial {
fn fragment_shader() -> ShaderRef {
"shaders/custom_material.wgsl".into()
}
}
// This is the struct that will be passed to your shader
#[derive(AsBindGroup, TypeUuid, Debug, Clone)]
#[uuid = "f690fdae-d598-45ab-8225-97e2a3f056e0"]
pub struct CustomMaterial {
#[uniform(0)]
color: Color,
#[texture(1)]
#[sampler(2)]
color_texture: Option<Handle<Image>>,
}
```
# Objective
Fixes#4907. Fixes#838. Fixes#5089.
Supersedes #5146. Supersedes #2087. Supersedes #865. Supersedes #5114
Visibility is currently entirely local. Set a parent entity to be invisible, and the children are still visible. This makes it hard for users to hide entire hierarchies of entities.
Additionally, the semantics of `Visibility` vs `ComputedVisibility` are inconsistent across entity types. 3D meshes use `ComputedVisibility` as the "definitive" visibility component, with `Visibility` being just one data source. Sprites just use `Visibility`, which means they can't feed off of `ComputedVisibility` data, such as culling information, RenderLayers, and (added in this pr) visibility inheritance information.
## Solution
Splits `ComputedVisibilty::is_visible` into `ComputedVisibilty::is_visible_in_view` and `ComputedVisibilty::is_visible_in_hierarchy`. For each visible entity, `is_visible_in_hierarchy` is computed by propagating visibility down the hierarchy. The `ComputedVisibility::is_visible()` function combines these two booleans for the canonical "is this entity visible" function.
Additionally, all entities that have `Visibility` now also have `ComputedVisibility`. Sprites, Lights, and UI entities now use `ComputedVisibility` when appropriate.
This means that in addition to visibility inheritance, everything using Visibility now also supports RenderLayers. Notably, Sprites (and other 2d objects) now support `RenderLayers` and work properly across multiple views.
Also note that this does increase the amount of work done per sprite. Bevymark with 100,000 sprites on `main` runs in `0.017612` seconds and this runs in `0.01902`. That is certainly a gap, but I believe the api consistency and extra functionality this buys us is worth it. See [this thread](https://github.com/bevyengine/bevy/pull/5146#issuecomment-1182783452) for more info. Note that #5146 in combination with #5114 _are_ a viable alternative to this PR and _would_ perform better, but that comes at the cost of api inconsistencies and doing visibility calculations in the "wrong" place. The current visibility system does have potential for performance improvements. I would prefer to evolve that one system as a whole rather than doing custom hacks / different behaviors for each feature slice.
Here is a "split screen" example where the left camera uses RenderLayers to filter out the blue sprite.
Note that this builds directly on #5146 and that @james7132 deserves the credit for the baseline visibility inheritance work. This pr moves the inherited visibility field into `ComputedVisibility`, then does the additional work of porting everything to `ComputedVisibility`. See my [comments here](https://github.com/bevyengine/bevy/pull/5146#issuecomment-1182783452) for rationale.
## Follow up work
* Now that lights use ComputedVisibility, VisibleEntities now includes "visible lights" in the entity list. Functionally not a problem as we use queries to filter the list down in the desired context. But we should consider splitting this out into a separate`VisibleLights` collection for both clarity and performance reasons. And _maybe_ even consider scoping `VisibleEntities` down to `VisibleMeshes`?.
* Investigate alternative sprite rendering impls (in combination with visibility system tweaks) that avoid re-generating a per-view fixedbitset of visible entities every frame, then checking each ExtractedEntity. This is where most of the performance overhead lives. Ex: we could generate ExtractedEntities per-view using the VisibleEntities list, avoiding the need for the bitset.
* Should ComputedVisibility use bitflags under the hood? This would cut down on the size of the component, potentially speed up the `is_visible()` function, and allow us to cheaply expand ComputedVisibility with more data (ex: split out local visibility and parent visibility, add more culling classes, etc).
---
## Changelog
* ComputedVisibility now takes hierarchy visibility into account.
* 2D, UI and Light entities now use the ComputedVisibility component.
## Migration Guide
If you were previously reading `Visibility::is_visible` as the "actual visibility" for sprites or lights, use `ComputedVisibilty::is_visible()` instead:
```rust
// before (0.7)
fn system(query: Query<&Visibility>) {
for visibility in query.iter() {
if visibility.is_visible {
log!("found visible entity");
}
}
}
// after (0.8)
fn system(query: Query<&ComputedVisibility>) {
for visibility in query.iter() {
if visibility.is_visible() {
log!("found visible entity");
}
}
}
```
Co-authored-by: Carter Anderson <mcanders1@gmail.com>
# Objective
Rust's nightly builds semi-regularly break us (or our dependencies). This creates churn and angst when we're just trying to get our jobs done.
We do still want nightly builds for a variety of reasons:
* cargo-udeps requires nightly and likely always will.
* Helps us catch rust nightly bugs quickly. We're "good citizens" if we regularly report regressions.
* Lets us prepare for "actual expected breakage" ahead of stable releases so we avoid breaking main users.
## Solution
* This pr parameterizes the nightly toolchain, making it an easy one-liner to pin our builds to a specific nightly, when required.
* Put nightly jobs to check wasm and nightly on their own matrix. I also removed tests on nightly linux, just build check
* alternative to https://github.com/bevyengine/bevy/pull/5329
Co-authored-by: Carter Anderson <mcanders1@gmail.com>
Export and register a missing type from `glam`.
Reflect impls were already present, but not registered.
Co-authored-by: devil-ira <justthecooldude@gmail.com>
# Objective
- Showcase how to use a `Material` and `Mesh` to spawn 3d lines
## Solution
- Add an example using a simple `Material` and `Mesh` definition to draw a 3d line
- Shows how to use `LineList` and `LineStrip` in combination with a specialized `Material`
## Notes
This isn't just a primitive shape because it needs a special Material, but I think it's a good showcase of the power of the `Material` and `AsBindGroup` abstractions. All of this is easy to figure out when you know these options are a thing, but I think they are hard to discover which is why I think this should be an example and not shipped with bevy.
Co-authored-by: Charles <IceSentry@users.noreply.github.com>
# Objective
- There is a warning when building in release:
```
warning: unused import: `Local`
--> crates/bevy_render/src/extract_resource.rs:4:34
|
4 | use bevy_ecs::system::{Commands, Local, Res, ResMut, Resource};
| ^^^^^
|
= note: `#[warn(unused_imports)]` on by default
```
- It's used 814f8d1635/crates/bevy_render/src/extract_resource.rs (L45)
- Fix it
## Solution
- Gate the import
# Objective
- Closes#4954
- Reduce the complexity of the `{System, App, *}Label` APIs.
## Solution
For the sake of brevity I will only refer to `SystemLabel`, but everything applies to all of the other label types as well.
- Add `SystemLabelId`, a lightweight, `copy` struct.
- Convert custom types into `SystemLabelId` using the trait `SystemLabel`.
## Changelog
- String literals implement `SystemLabel` for now, but this should be changed with #4409 .
## Migration Guide
- Any previous use of `Box<dyn SystemLabel>` should be replaced with `SystemLabelId`.
- `AsSystemLabel` trait has been modified.
- No more output generics.
- Method `as_system_label` now returns `SystemLabelId`, removing an unnecessary level of indirection.
- If you *need* a label that is determined at runtime, you can use `Box::leak`. Not recommended.
## Questions for later
* Should we generate a `Debug` impl along with `#[derive(*Label)]`?
* Should we rename `as_str()`?
* Should we remove the extra derives (such as `Hash`) from builtin `*Label` types?
* Should we automatically derive types like `Clone, Copy, PartialEq, Eq`?
* More-ergonomic comparisons between `Label` and `LabelId`.
* Move `Dyn{Eq, Hash,Clone}` somewhere else.
* Some API to make interning dynamic labels easier.
* Optimize string representation
* Empty string for unit structs -- no debug info but faster comparisons
* Don't show enum types -- same tradeoffs as asbove.
This replaces `rand` with `fastrand` as the source of randomness for `HandleId::new()` in `bevy_asset`. This was the only crate with a dependency on `rand`, and now the dependency exists only as a dev-dependency.
`fastrand` was already in the dependency tree, thanks to `futures-lite`, `async-executor`, and `tempfile` to name a few.
## Changelog
Removed `rand` from dependencies in `bevy_asset` in favor of existing in-tree `fast-rand`
Add compile time check for if a system is an exclusive system. Resolves#4788
Co-authored-by: Daniel Liu <mr.picklepinosaur@gmail.com>
Co-authored-by: Daniel Liu <danieliu3120@gmail.com>
Following https://github.com/bevyengine/bevy/pull/5124 I decided to add the `ExactSizeIterator` impl for `QueryCombinationIter`.
Also:
- Clean up the tests for `size_hint` and `len` for both the normal `QueryIter` and `QueryCombinationIter`.
- Add tests to `QueryCombinationIter` when it shouldn't be `ExactSizeIterator`
---
## Changelog
- Added `ExactSizeIterator` implementation for `QueryCombinatonIter`
# Objective
- `.iter_combinations_*()` cannot be used on custom derived `WorldQuery`, so this fixes that
- Fixes#5284
## Solution
- `#[derive(Clone)]` on the `Fetch` of the proc macro derive.
- `#[derive(Clone)]` for `AnyOf` to satisfy tests.
Someone noted that the `rotate_around` method did not give the results they expected: [discord thread](https://discord.com/channels/691052431525675048/996497295325544479)
I tested `rotate_around` and their workaround and it seems like it was indeed incorrect.
Here is a scene with some cubes at different angles all being rotated around the center on the Y axis.
https://user-images.githubusercontent.com/29694403/178598432-407d7e80-1caf-4b17-b69b-66d9156c81e1.mp4
Interestingly, the middle cube rotates as you might expect. This threw me for a bit of a loop before I added the other cubes to the test haha.
Here is the same scene with the order multiplication of the quaternions flipped in `rotate_around`.
https://user-images.githubusercontent.com/29694403/178598446-a98026f3-524c-448b-8437-4d0d3175c6ca.mp4
That looks better :)
## Changelog
* Fixed `rotate_around` rotating the wrong way around
* Added `translate_around`. - Split out the translation code from `rotate_around`.
* Simplified/optimized `rotate_local_*` methods. - Yep, That works somehow.
<sup>Quaternions sure are wacky. Do not ask me how this works exactly, haha.</sup>
Co-authored-by: devil-ira <justthecooldude@gmail.com>
Generally a good idea.
I ran into this because I wanted to store `Gamepads` in a wrapper struct in https://github.com/Leafwing-Studios/leafwing-input-manager/pull/168.
This PR allows the `Debug` derive used there to continue working. I could workaround this with a custom impl, but a PR upstream seemed like the right fix.
# Objective
Improve documentation, information users of the limitations in bevy's idiomatic patterns, and suggesting alternatives for when those limitations are encountered.
## Solution
* Add documentation to `Commands` informing the user of the option of writing one-shot commands with closures.
* Add documentation to `EventWriter` regarding the limitations of event types, and suggesting alternatives using commands.
# Objective
Fixes#5304
## Solution
Instead of using a simple utility function for loading, which uses a default allocation limit of 512MB, we use a Reader object which can be configured ad hoc.
## Changelog
> This section is optional. If this was a trivial fix, or has no externally-visible impact, you can delete this section.
- Allows loading of textures larger than 512MB
# Objective
- Wireframes are currently not rendering on main because they aren't being extracted correctly
## Solution
- Extract the wireframes correctly
# Objective
When someone searches in rustdoc for `world_to_screen`, they now will
find `world_to_viewport`. The method was renamed in 0.8, it would be
nice to allow users to find the new name more easily.
---
# Objective
This PR aims to document the `bevy_asset` crate to complete coverage, while also trying to improve some bits of UX.
### Progress
- [x] Root items
- [x] `handle` module
- [x] `info` module
- [x] `path` module
- [x] `loader` module
- [x] `io` and `filesystem_watcher` module
- [x] `assets` module
- [x] `asset_server` module
- [x] `diagnostic` module
- [x] `debug_asset_server` module
- [x] Crate level documentation
- [x] Add `#![warn(missing_docs)]` lint
Coverage: 100%
## Migration Guide
- Rename `FileAssetIo::get_root_path` uses to `FileAssetIo::get_base_path`
`FileAssetIo::root_path()` is a getter for the `root_path` field, while `FileAssetIo::get_root_path` returned the parent directory of the asset root path, which was the executable's directory unless `CARGO_MANIFEST_DIR` was set. This change solves the ambiguity between the two methods.
# Objective
- Added a bunch of backticks to things that should have them, like equations, abstract variable names,
- Changed all small x, y, and z to capitals X, Y, Z.
This might be more annoying than helpful; Feel free to refuse this PR.
# Objective
- The time update is currently done in the wrong part of the schedule. For a single frame the current order of things is update input, update time (First stage), other stages, render stage (frame presentation). So when we update the time it includes the input processing of the current frame and the frame presentation of the previous frame. This is a problem when vsync is on. When input processing takes a longer amount of time for a frame, the vsync wait time gets shorter. So when these are not paired correctly we can potentially have a long input processing time added to the normal vsync wait time in the previous frame. This leads to inaccurate frame time reporting and more variance of the time than actually exists. For more details of why this is an issue see the linked issue below.
- Helps with https://github.com/bevyengine/bevy/issues/4669
- Supercedes https://github.com/bevyengine/bevy/pull/4728 and https://github.com/bevyengine/bevy/pull/4735. This PR should be less controversial than those because it doesn't add to the API surface.
## Solution
- The most accurate frame time would come from hardware. We currently don't have access to that for multiple reasons, so the next best thing we can do is measure the frame time as close to frame presentation as possible. This PR gets the Instant::now() for the time immediately after frame presentation in the render system and then sends that time to the app world through a channel.
- implements suggestion from @aevyrie from here https://github.com/bevyengine/bevy/pull/4728#discussion_r872010606
## Statistics
---
## Changelog
- Make frame time reporting more accurate.
## Migration Guide
`time.delta()` now reports zero for 2 frames on startup instead of 1 frame.
Remove unnecessary calls to `iter()`/`iter_mut()`.
Mainly updates the use of queries in our code, docs, and examples.
```rust
// From
for _ in list.iter() {
for _ in list.iter_mut() {
// To
for _ in &list {
for _ in &mut list {
```
We already enable the pedantic lint [clippy::explicit_iter_loop](https://rust-lang.github.io/rust-clippy/stable/) inside of Bevy. However, this only warns for a few known types from the standard library.
## Note for reviewers
As you can see the additions and deletions are exactly equal.
Maybe give it a quick skim to check I didn't sneak in a crypto miner, but you don't have to torture yourself by reading every line.
I already experienced enough pain making this PR :)
Co-authored-by: devil-ira <justthecooldude@gmail.com>
Resolves#5004. As suggested in the original issue, change tuple types to their corresponding vector type.
## migration guide
Changed the following fields
- `WindowCommand::SetWindowMode.resolution` from `(u32, u32)` to `UVec2`
- `WindowCommand::SetResolution.logical_resolution` from `(f32, f32)` to `Vec2`
Co-authored-by: Daniel Liu <mr.picklepinosaur@gmail.com>
# Objective
- Reduce confusion as the example opens a window and isn't truly "headless"
- Fixes https://github.com/bevyengine/bevy/issues/5260.
## Solution
- Rename the example and add to the docs that the window is expected.
# Objective
- Validate the format of the values with the expected attribute format.
- Currently, if you pass the wrong format, it will crash somewhere unrelated with a very cryptic error message, so it's really hard to debug for beginners.
## Solution
- Compare the format and panic when unexpected format is passed
## Note
- I used a separate `error!()` for a human friendly message because the panic message is very noisy and hard to parse for beginners. I don't mind changing this to only a panic if people prefer that.
- This could potentially be something that runs only in debug mode, but I don't think inserting attributes is done often enough for this to be an issue.
Co-authored-by: Charles <IceSentry@users.noreply.github.com>
Small optimization. `.collect()` from arrays generates very nice code without reallocations: https://rust.godbolt.org/z/6E6c595bq
Co-authored-by: Kornel <kornel@geekhood.net>
