bevy/crates/bevy_gizmos/src/arrows.rs

233 lines
7.6 KiB
Rust
Raw Normal View History

Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
//! Additional [`GizmoBuffer`] Functions -- Arrows
//!
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
//! Includes the implementation of [`GizmoBuffer::arrow`] and [`GizmoBuffer::arrow_2d`],
//! and assorted support items.
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
use crate::{gizmos::GizmoBuffer, prelude::GizmoConfigGroup};
Add coordinate axes gizmo (#12211) # Objective - We introduce a gizmo that displays coordinate axes relative to a Transform*, primarily for debugging purposes. - See #9400 ## Solution A new method, `Gizmos::axes`, takes a `Transform`* as input and displays the standard coordinate axes, transformed according to it; its signature looks like this: ````rust pub fn axes(&mut self, transform: into TransformPoint, base_length: f32) { //... } ```` If my carefully placed asterisks hadn't already tipped you off, the argument here is not actually a `Transform` but instead anything which implements `TransformPoint`, which allows it to work also with `GlobalTransform` (and also `Mat4` and `Affine3A`, if the user happens to be hand-rolling transformations in some way). The `base_length` parameter is a scaling factor applied to the coordinate vectors before the transformation takes place; in other words, the caller can use this to help size the coordinate axes appropriately for the entity that they are attached to. An example invocation of this method looks something like this: ````rust fn draw_axes_system( mut gizmos: Gizmos, query: Query<&Transform, With<MyMarkerComponent>>, ) { for &transform in &query { gizmos.axes(transform, 2.); } } ```` The result is the three coordinate axes, X, Y, Z (colored red, green, and blue, respectively), drawn onto the entity: <img width="206" alt="Screenshot 2024-02-29 at 2 41 45 PM" src="https://github.com/bevyengine/bevy/assets/2975848/789d1703-29ae-4295-80ab-b87459cf8037"> Note that, if scaling was applied as part of the given transformation, it shows up in scaling on the axes as well: <img width="377" alt="Screenshot 2024-02-29 at 2 43 53 PM" src="https://github.com/bevyengine/bevy/assets/2975848/6dc1caf4-8b3e-47f7-a86a-8906d870fa72"> --- ## Changelog - Added `Gizmos::axes` in bevy_gizmos/src/arrows.rs - Fixed a minor issue with `ArrowBuilder::with_tip_length` not correctly implementing builder style (no external impact) --- ## Discussion ### Design considerations I feel pretty strongly that having no default length scale is for the best, at least for the time being, since it's very easy for the length scale to be too small, leading to the axes being hidden inside the body of the object they are associated with. That is, if the API instead looked like this: ````rust gizmos.axes(transform); // or gizmos.axes(transform).with_length_scale(3.0); ```` then I think it's a reasonable expectation that the first thing would "just work" for most applications, and it wouldn't, which would be kind of a footgun. ### Future steps There are a few directions that this might expand in the future: 1. Introduce additional options via the standard builder pattern; i.e. introducing `AxesBuilder<T: TransformPoint>` so that people can configure the axis colors, normalize all axes to a fixed length independent of scale deformations, etc. 2. Fold this functionality into a plugin (like AabbGizmoPlugin) so that the functionality becomes more-or-less automatic based on certain fixed marker components. This wouldn't be very hard to implement, and it has the benefit of making the axes more frictionless to use. Furthermore, if we coupled this to the AABB functionality we already have, we could also ensure that the plugin automatically sizes the axes (by coupling their size to the dimensions of the AABB, for example). 3. Implement something similar for 2d. Honestly, I have no idea if this is desired/useful, but I could probably just implement it in this PR if that's the case.
2024-02-29 23:52:05 +00:00
use bevy_color::{
palettes::basic::{BLUE, GREEN, RED},
Color,
};
use bevy_math::{Quat, Vec2, Vec3, Vec3Swizzles};
Add coordinate axes gizmo (#12211) # Objective - We introduce a gizmo that displays coordinate axes relative to a Transform*, primarily for debugging purposes. - See #9400 ## Solution A new method, `Gizmos::axes`, takes a `Transform`* as input and displays the standard coordinate axes, transformed according to it; its signature looks like this: ````rust pub fn axes(&mut self, transform: into TransformPoint, base_length: f32) { //... } ```` If my carefully placed asterisks hadn't already tipped you off, the argument here is not actually a `Transform` but instead anything which implements `TransformPoint`, which allows it to work also with `GlobalTransform` (and also `Mat4` and `Affine3A`, if the user happens to be hand-rolling transformations in some way). The `base_length` parameter is a scaling factor applied to the coordinate vectors before the transformation takes place; in other words, the caller can use this to help size the coordinate axes appropriately for the entity that they are attached to. An example invocation of this method looks something like this: ````rust fn draw_axes_system( mut gizmos: Gizmos, query: Query<&Transform, With<MyMarkerComponent>>, ) { for &transform in &query { gizmos.axes(transform, 2.); } } ```` The result is the three coordinate axes, X, Y, Z (colored red, green, and blue, respectively), drawn onto the entity: <img width="206" alt="Screenshot 2024-02-29 at 2 41 45 PM" src="https://github.com/bevyengine/bevy/assets/2975848/789d1703-29ae-4295-80ab-b87459cf8037"> Note that, if scaling was applied as part of the given transformation, it shows up in scaling on the axes as well: <img width="377" alt="Screenshot 2024-02-29 at 2 43 53 PM" src="https://github.com/bevyengine/bevy/assets/2975848/6dc1caf4-8b3e-47f7-a86a-8906d870fa72"> --- ## Changelog - Added `Gizmos::axes` in bevy_gizmos/src/arrows.rs - Fixed a minor issue with `ArrowBuilder::with_tip_length` not correctly implementing builder style (no external impact) --- ## Discussion ### Design considerations I feel pretty strongly that having no default length scale is for the best, at least for the time being, since it's very easy for the length scale to be too small, leading to the axes being hidden inside the body of the object they are associated with. That is, if the API instead looked like this: ````rust gizmos.axes(transform); // or gizmos.axes(transform).with_length_scale(3.0); ```` then I think it's a reasonable expectation that the first thing would "just work" for most applications, and it wouldn't, which would be kind of a footgun. ### Future steps There are a few directions that this might expand in the future: 1. Introduce additional options via the standard builder pattern; i.e. introducing `AxesBuilder<T: TransformPoint>` so that people can configure the axis colors, normalize all axes to a fixed length independent of scale deformations, etc. 2. Fold this functionality into a plugin (like AabbGizmoPlugin) so that the functionality becomes more-or-less automatic based on certain fixed marker components. This wouldn't be very hard to implement, and it has the benefit of making the axes more frictionless to use. Furthermore, if we coupled this to the AABB functionality we already have, we could also ensure that the plugin automatically sizes the axes (by coupling their size to the dimensions of the AABB, for example). 3. Implement something similar for 2d. Honestly, I have no idea if this is desired/useful, but I could probably just implement it in this PR if that's the case.
2024-02-29 23:52:05 +00:00
use bevy_transform::TransformPoint;
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
/// A builder returned by [`GizmoBuffer::arrow`] and [`GizmoBuffer::arrow_2d`]
pub struct ArrowBuilder<'a, Config, Clear>
Contextually clearing gizmos (#10973) # Objective Allow `Gizmos` to work in `FixedUpdate` without any changes needed. This changes `Gizmos` from being a purely immediate mode api, but allows the user to use it as if it were an immediate mode API regardless of schedule context. Also allows for extending by other custom schedules by adding their own `GizmoStorage<Clear>` and the requisite systems: - `propagate_gizmos::<Clear>` before `update_gizmo_meshes` - `stash_default_gizmos` when starting a clear context - `pop_default_gizmos` when ending a clear context - `collect_default_gizmos` when grabbing the requested gizmos - `clear_gizmos` for clearing the context's gizmos ## Solution Adds a generic to `Gizmos` that defaults to `Update` (the current way gizmos works). When entering a new clear context the default `Gizmos` gets swapped out for that context's duration so the context can collect the gizmos requested. Prior work: https://github.com/bevyengine/bevy/pull/9153 ## To do - [x] `FixedUpdate` should probably get its own First, Pre, Update, Post, Last system sets for this. Otherwise users will need to make sure to order their systems before `clear_gizmos`. This could alternatively be fixed by moving the setup of this to `bevy_time::fixed`? PR to fix this issue: https://github.com/bevyengine/bevy/pull/10977 - [x] use mem::take internally for the swaps? - [x] Better name for the `Context` generic on gizmos? `Clear`? --- ## Changelog - Gizmos drawn in `FixedMain` now last until the next `FixedMain` iteration runs.
2024-04-23 00:16:12 +00:00
where
Config: GizmoConfigGroup,
Clear: 'static + Send + Sync,
{
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
gizmos: &'a mut GizmoBuffer<Config, Clear>,
start: Vec3,
end: Vec3,
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
color: Color,
double_ended: bool,
tip_length: f32,
}
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
impl<Config, Clear> ArrowBuilder<'_, Config, Clear>
Contextually clearing gizmos (#10973) # Objective Allow `Gizmos` to work in `FixedUpdate` without any changes needed. This changes `Gizmos` from being a purely immediate mode api, but allows the user to use it as if it were an immediate mode API regardless of schedule context. Also allows for extending by other custom schedules by adding their own `GizmoStorage<Clear>` and the requisite systems: - `propagate_gizmos::<Clear>` before `update_gizmo_meshes` - `stash_default_gizmos` when starting a clear context - `pop_default_gizmos` when ending a clear context - `collect_default_gizmos` when grabbing the requested gizmos - `clear_gizmos` for clearing the context's gizmos ## Solution Adds a generic to `Gizmos` that defaults to `Update` (the current way gizmos works). When entering a new clear context the default `Gizmos` gets swapped out for that context's duration so the context can collect the gizmos requested. Prior work: https://github.com/bevyengine/bevy/pull/9153 ## To do - [x] `FixedUpdate` should probably get its own First, Pre, Update, Post, Last system sets for this. Otherwise users will need to make sure to order their systems before `clear_gizmos`. This could alternatively be fixed by moving the setup of this to `bevy_time::fixed`? PR to fix this issue: https://github.com/bevyengine/bevy/pull/10977 - [x] use mem::take internally for the swaps? - [x] Better name for the `Context` generic on gizmos? `Clear`? --- ## Changelog - Gizmos drawn in `FixedMain` now last until the next `FixedMain` iteration runs.
2024-04-23 00:16:12 +00:00
where
Config: GizmoConfigGroup,
Clear: 'static + Send + Sync,
{
/// Change the length of the tips to be `length`.
/// The default tip length is [length of the arrow]/10.
///
/// # Example
/// ```
/// # use bevy_gizmos::prelude::*;
/// # use bevy_math::prelude::*;
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
/// # use bevy_color::palettes::basic::GREEN;
/// fn system(mut gizmos: Gizmos) {
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
/// gizmos.arrow(Vec3::ZERO, Vec3::ONE, GREEN)
/// .with_tip_length(3.);
/// }
/// # bevy_ecs::system::assert_is_system(system);
/// ```
#[doc(alias = "arrow_head_length")]
Add coordinate axes gizmo (#12211) # Objective - We introduce a gizmo that displays coordinate axes relative to a Transform*, primarily for debugging purposes. - See #9400 ## Solution A new method, `Gizmos::axes`, takes a `Transform`* as input and displays the standard coordinate axes, transformed according to it; its signature looks like this: ````rust pub fn axes(&mut self, transform: into TransformPoint, base_length: f32) { //... } ```` If my carefully placed asterisks hadn't already tipped you off, the argument here is not actually a `Transform` but instead anything which implements `TransformPoint`, which allows it to work also with `GlobalTransform` (and also `Mat4` and `Affine3A`, if the user happens to be hand-rolling transformations in some way). The `base_length` parameter is a scaling factor applied to the coordinate vectors before the transformation takes place; in other words, the caller can use this to help size the coordinate axes appropriately for the entity that they are attached to. An example invocation of this method looks something like this: ````rust fn draw_axes_system( mut gizmos: Gizmos, query: Query<&Transform, With<MyMarkerComponent>>, ) { for &transform in &query { gizmos.axes(transform, 2.); } } ```` The result is the three coordinate axes, X, Y, Z (colored red, green, and blue, respectively), drawn onto the entity: <img width="206" alt="Screenshot 2024-02-29 at 2 41 45 PM" src="https://github.com/bevyengine/bevy/assets/2975848/789d1703-29ae-4295-80ab-b87459cf8037"> Note that, if scaling was applied as part of the given transformation, it shows up in scaling on the axes as well: <img width="377" alt="Screenshot 2024-02-29 at 2 43 53 PM" src="https://github.com/bevyengine/bevy/assets/2975848/6dc1caf4-8b3e-47f7-a86a-8906d870fa72"> --- ## Changelog - Added `Gizmos::axes` in bevy_gizmos/src/arrows.rs - Fixed a minor issue with `ArrowBuilder::with_tip_length` not correctly implementing builder style (no external impact) --- ## Discussion ### Design considerations I feel pretty strongly that having no default length scale is for the best, at least for the time being, since it's very easy for the length scale to be too small, leading to the axes being hidden inside the body of the object they are associated with. That is, if the API instead looked like this: ````rust gizmos.axes(transform); // or gizmos.axes(transform).with_length_scale(3.0); ```` then I think it's a reasonable expectation that the first thing would "just work" for most applications, and it wouldn't, which would be kind of a footgun. ### Future steps There are a few directions that this might expand in the future: 1. Introduce additional options via the standard builder pattern; i.e. introducing `AxesBuilder<T: TransformPoint>` so that people can configure the axis colors, normalize all axes to a fixed length independent of scale deformations, etc. 2. Fold this functionality into a plugin (like AabbGizmoPlugin) so that the functionality becomes more-or-less automatic based on certain fixed marker components. This wouldn't be very hard to implement, and it has the benefit of making the axes more frictionless to use. Furthermore, if we coupled this to the AABB functionality we already have, we could also ensure that the plugin automatically sizes the axes (by coupling their size to the dimensions of the AABB, for example). 3. Implement something similar for 2d. Honestly, I have no idea if this is desired/useful, but I could probably just implement it in this PR if that's the case.
2024-02-29 23:52:05 +00:00
pub fn with_tip_length(mut self, length: f32) -> Self {
self.tip_length = length;
Add coordinate axes gizmo (#12211) # Objective - We introduce a gizmo that displays coordinate axes relative to a Transform*, primarily for debugging purposes. - See #9400 ## Solution A new method, `Gizmos::axes`, takes a `Transform`* as input and displays the standard coordinate axes, transformed according to it; its signature looks like this: ````rust pub fn axes(&mut self, transform: into TransformPoint, base_length: f32) { //... } ```` If my carefully placed asterisks hadn't already tipped you off, the argument here is not actually a `Transform` but instead anything which implements `TransformPoint`, which allows it to work also with `GlobalTransform` (and also `Mat4` and `Affine3A`, if the user happens to be hand-rolling transformations in some way). The `base_length` parameter is a scaling factor applied to the coordinate vectors before the transformation takes place; in other words, the caller can use this to help size the coordinate axes appropriately for the entity that they are attached to. An example invocation of this method looks something like this: ````rust fn draw_axes_system( mut gizmos: Gizmos, query: Query<&Transform, With<MyMarkerComponent>>, ) { for &transform in &query { gizmos.axes(transform, 2.); } } ```` The result is the three coordinate axes, X, Y, Z (colored red, green, and blue, respectively), drawn onto the entity: <img width="206" alt="Screenshot 2024-02-29 at 2 41 45 PM" src="https://github.com/bevyengine/bevy/assets/2975848/789d1703-29ae-4295-80ab-b87459cf8037"> Note that, if scaling was applied as part of the given transformation, it shows up in scaling on the axes as well: <img width="377" alt="Screenshot 2024-02-29 at 2 43 53 PM" src="https://github.com/bevyengine/bevy/assets/2975848/6dc1caf4-8b3e-47f7-a86a-8906d870fa72"> --- ## Changelog - Added `Gizmos::axes` in bevy_gizmos/src/arrows.rs - Fixed a minor issue with `ArrowBuilder::with_tip_length` not correctly implementing builder style (no external impact) --- ## Discussion ### Design considerations I feel pretty strongly that having no default length scale is for the best, at least for the time being, since it's very easy for the length scale to be too small, leading to the axes being hidden inside the body of the object they are associated with. That is, if the API instead looked like this: ````rust gizmos.axes(transform); // or gizmos.axes(transform).with_length_scale(3.0); ```` then I think it's a reasonable expectation that the first thing would "just work" for most applications, and it wouldn't, which would be kind of a footgun. ### Future steps There are a few directions that this might expand in the future: 1. Introduce additional options via the standard builder pattern; i.e. introducing `AxesBuilder<T: TransformPoint>` so that people can configure the axis colors, normalize all axes to a fixed length independent of scale deformations, etc. 2. Fold this functionality into a plugin (like AabbGizmoPlugin) so that the functionality becomes more-or-less automatic based on certain fixed marker components. This wouldn't be very hard to implement, and it has the benefit of making the axes more frictionless to use. Furthermore, if we coupled this to the AABB functionality we already have, we could also ensure that the plugin automatically sizes the axes (by coupling their size to the dimensions of the AABB, for example). 3. Implement something similar for 2d. Honestly, I have no idea if this is desired/useful, but I could probably just implement it in this PR if that's the case.
2024-02-29 23:52:05 +00:00
self
}
/// Adds another tip to the arrow, appended in the start point.
/// the default is only one tip at the end point.
pub fn with_double_end(mut self) -> Self {
self.double_ended = true;
self
}
}
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
impl<Config, Clear> Drop for ArrowBuilder<'_, Config, Clear>
Contextually clearing gizmos (#10973) # Objective Allow `Gizmos` to work in `FixedUpdate` without any changes needed. This changes `Gizmos` from being a purely immediate mode api, but allows the user to use it as if it were an immediate mode API regardless of schedule context. Also allows for extending by other custom schedules by adding their own `GizmoStorage<Clear>` and the requisite systems: - `propagate_gizmos::<Clear>` before `update_gizmo_meshes` - `stash_default_gizmos` when starting a clear context - `pop_default_gizmos` when ending a clear context - `collect_default_gizmos` when grabbing the requested gizmos - `clear_gizmos` for clearing the context's gizmos ## Solution Adds a generic to `Gizmos` that defaults to `Update` (the current way gizmos works). When entering a new clear context the default `Gizmos` gets swapped out for that context's duration so the context can collect the gizmos requested. Prior work: https://github.com/bevyengine/bevy/pull/9153 ## To do - [x] `FixedUpdate` should probably get its own First, Pre, Update, Post, Last system sets for this. Otherwise users will need to make sure to order their systems before `clear_gizmos`. This could alternatively be fixed by moving the setup of this to `bevy_time::fixed`? PR to fix this issue: https://github.com/bevyengine/bevy/pull/10977 - [x] use mem::take internally for the swaps? - [x] Better name for the `Context` generic on gizmos? `Clear`? --- ## Changelog - Gizmos drawn in `FixedMain` now last until the next `FixedMain` iteration runs.
2024-04-23 00:16:12 +00:00
where
Config: GizmoConfigGroup,
Clear: 'static + Send + Sync,
{
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
/// Draws the arrow, by drawing lines with the stored [`GizmoBuffer`]
fn drop(&mut self) {
Multiple Configurations for Gizmos (#10342) # Objective This PR aims to implement multiple configs for gizmos as discussed in #9187. ## Solution Configs for the new `GizmoConfigGroup`s are stored in a `GizmoConfigStore` resource and can be accesses using a type based key or iterated over. This type based key doubles as a standardized location where plugin authors can put their own configuration not covered by the standard `GizmoConfig` struct. For example the `AabbGizmoGroup` has a default color and toggle to show all AABBs. New configs can be registered using `app.init_gizmo_group::<T>()` during startup. When requesting the `Gizmos<T>` system parameter the generic type determines which config is used. The config structs are available through the `Gizmos` system parameter allowing for easy access while drawing your gizmos. Internally, resources and systems used for rendering (up to an including the extract system) are generic over the type based key and inserted on registering a new config. ## Alternatives The configs could be stored as components on entities with markers which would make better use of the ECS. I also implemented this approach ([here](https://github.com/jeliag/bevy/tree/gizmo-multiconf-comp)) and believe that the ergonomic benefits of a central config store outweigh the decreased use of the ECS. ## Unsafe Code Implementing system parameter by hand is unsafe but seems to be required to access the config store once and not on every gizmo draw function call. This is critical for performance. ~Is there a better way to do this?~ ## Future Work New gizmos (such as #10038, and ideas from #9400) will require custom configuration structs. Should there be a new custom config for every gizmo type, or should we group them together in a common configuration? (for example `EditorGizmoConfig`, or something more fine-grained) ## Changelog - Added `GizmoConfigStore` resource and `GizmoConfigGroup` trait - Added `init_gizmo_group` to `App` - Added early returns to gizmo drawing increasing performance when gizmos are disabled - Changed `GizmoConfig` and aabb gizmos to use new `GizmoConfigStore` - Changed `Gizmos` system parameter to use type based key to retrieve config - Changed resources and systems used for gizmo rendering to be generic over type based key - Changed examples (3d_gizmos, 2d_gizmos) to showcase new API ## Migration Guide - `GizmoConfig` is no longer a resource and has to be accessed through `GizmoConfigStore` resource. The default config group is `DefaultGizmoGroup`, but consider using your own custom config group if applicable. --------- Co-authored-by: Nicola Papale <nicopap@users.noreply.github.com>
2024-01-18 15:52:50 +00:00
if !self.gizmos.enabled {
return;
}
// first, draw the body of the arrow
self.gizmos.line(self.start, self.end, self.color);
// now the hard part is to draw the head in a sensible way
// put us in a coordinate system where the arrow is pointing towards +x and ends at the origin
let pointing_end = (self.end - self.start).normalize();
let rotation_end = Quat::from_rotation_arc(Vec3::X, pointing_end);
let tips = [
Vec3::new(-1., 1., 0.),
Vec3::new(-1., 0., 1.),
Vec3::new(-1., -1., 0.),
Vec3::new(-1., 0., -1.),
];
// - extend the vectors so their length is `tip_length`
// - rotate the world so +x is facing in the same direction as the arrow
// - translate over to the tip of the arrow
let tips_end = tips.map(|v| rotation_end * (v.normalize() * self.tip_length) + self.end);
for v in tips_end {
// then actually draw the tips
self.gizmos.line(self.end, v, self.color);
}
if self.double_ended {
let pointing_start = (self.start - self.end).normalize();
let rotation_start = Quat::from_rotation_arc(Vec3::X, pointing_start);
let tips_start =
tips.map(|v| rotation_start * (v.normalize() * self.tip_length) + self.start);
for v in tips_start {
// draw the start points tips
self.gizmos.line(self.start, v, self.color);
}
}
}
}
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
impl<Config, Clear> GizmoBuffer<Config, Clear>
Contextually clearing gizmos (#10973) # Objective Allow `Gizmos` to work in `FixedUpdate` without any changes needed. This changes `Gizmos` from being a purely immediate mode api, but allows the user to use it as if it were an immediate mode API regardless of schedule context. Also allows for extending by other custom schedules by adding their own `GizmoStorage<Clear>` and the requisite systems: - `propagate_gizmos::<Clear>` before `update_gizmo_meshes` - `stash_default_gizmos` when starting a clear context - `pop_default_gizmos` when ending a clear context - `collect_default_gizmos` when grabbing the requested gizmos - `clear_gizmos` for clearing the context's gizmos ## Solution Adds a generic to `Gizmos` that defaults to `Update` (the current way gizmos works). When entering a new clear context the default `Gizmos` gets swapped out for that context's duration so the context can collect the gizmos requested. Prior work: https://github.com/bevyengine/bevy/pull/9153 ## To do - [x] `FixedUpdate` should probably get its own First, Pre, Update, Post, Last system sets for this. Otherwise users will need to make sure to order their systems before `clear_gizmos`. This could alternatively be fixed by moving the setup of this to `bevy_time::fixed`? PR to fix this issue: https://github.com/bevyengine/bevy/pull/10977 - [x] use mem::take internally for the swaps? - [x] Better name for the `Context` generic on gizmos? `Clear`? --- ## Changelog - Gizmos drawn in `FixedMain` now last until the next `FixedMain` iteration runs.
2024-04-23 00:16:12 +00:00
where
Config: GizmoConfigGroup,
Clear: 'static + Send + Sync,
{
Drawing Primitives with Gizmos (#11072) The PR is in a reviewable state now in the sense that the basic implementations are there. There are still some ToDos that I'm aware of: - [x] docs for all the new structs and traits - [x] implement `Default` and derive other useful traits for the new structs - [x] Take a look at the notes again (Do this after a first round of reviews) - [x] Take care of the repetition in the circle drawing functions --- # Objective - TLDR: This PR enables us to quickly draw all the newly added primitives from `bevy_math` in immediate mode with gizmos - Addresses #10571 ## Solution - This implements the first design idea I had that covered everything that was mentioned in the Issue https://github.com/bevyengine/bevy/issues/10571#issuecomment-1863646197 --- ## Caveats - I added the `Primitive(2/3)d` impls for `Direction(2/3)d` to make them work with the current solution. We could impose less strict requirements for the gizmoable objects and remove the impls afterwards if the community doesn't like the current approach. --- ## Changelog - implement capabilities to draw ellipses on the gizmo in general (this was required to have some code which is able to draw the ellipse primitive) - refactored circle drawing code to use the more general ellipse drawing code to keep code duplication low - implement `Primitive2d` for `Direction2d` and impl `Primitive3d` for `Direction3d` - implement trait to draw primitives with specialized details with gizmos - `GizmoPrimitive2d` for all the 2D primitives - `GizmoPrimitive3d` for all the 3D primitives - (question while writing this: Does it actually matter if we split this in 2D and 3D? I guess it could be useful in the future if we do something based on the main rendering mode even though atm it's kinda useless) --- --------- Co-authored-by: nothendev <borodinov.ilya@gmail.com>
2024-02-02 21:13:03 +00:00
/// Draw an arrow in 3D, from `start` to `end`. Has four tips for convenient viewing from any direction.
///
/// This should be called for each frame the arrow needs to be rendered.
///
/// # Example
/// ```
/// # use bevy_gizmos::prelude::*;
/// # use bevy_math::prelude::*;
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
/// # use bevy_color::palettes::basic::GREEN;
/// fn system(mut gizmos: Gizmos) {
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
/// gizmos.arrow(Vec3::ZERO, Vec3::ONE, GREEN);
/// }
/// # bevy_ecs::system::assert_is_system(system);
/// ```
pub fn arrow(
&mut self,
start: Vec3,
end: Vec3,
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
color: impl Into<Color>,
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
) -> ArrowBuilder<'_, Config, Clear> {
let length = (end - start).length();
ArrowBuilder {
gizmos: self,
start,
end,
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
color: color.into(),
double_ended: false,
tip_length: length / 10.,
}
}
/// Draw an arrow in 2D (on the xy plane), from `start` to `end`.
///
/// This should be called for each frame the arrow needs to be rendered.
///
/// # Example
/// ```
/// # use bevy_gizmos::prelude::*;
/// # use bevy_math::prelude::*;
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
/// # use bevy_color::palettes::basic::GREEN;
/// fn system(mut gizmos: Gizmos) {
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
/// gizmos.arrow_2d(Vec2::ZERO, Vec2::X, GREEN);
/// }
/// # bevy_ecs::system::assert_is_system(system);
/// ```
Multiple Configurations for Gizmos (#10342) # Objective This PR aims to implement multiple configs for gizmos as discussed in #9187. ## Solution Configs for the new `GizmoConfigGroup`s are stored in a `GizmoConfigStore` resource and can be accesses using a type based key or iterated over. This type based key doubles as a standardized location where plugin authors can put their own configuration not covered by the standard `GizmoConfig` struct. For example the `AabbGizmoGroup` has a default color and toggle to show all AABBs. New configs can be registered using `app.init_gizmo_group::<T>()` during startup. When requesting the `Gizmos<T>` system parameter the generic type determines which config is used. The config structs are available through the `Gizmos` system parameter allowing for easy access while drawing your gizmos. Internally, resources and systems used for rendering (up to an including the extract system) are generic over the type based key and inserted on registering a new config. ## Alternatives The configs could be stored as components on entities with markers which would make better use of the ECS. I also implemented this approach ([here](https://github.com/jeliag/bevy/tree/gizmo-multiconf-comp)) and believe that the ergonomic benefits of a central config store outweigh the decreased use of the ECS. ## Unsafe Code Implementing system parameter by hand is unsafe but seems to be required to access the config store once and not on every gizmo draw function call. This is critical for performance. ~Is there a better way to do this?~ ## Future Work New gizmos (such as #10038, and ideas from #9400) will require custom configuration structs. Should there be a new custom config for every gizmo type, or should we group them together in a common configuration? (for example `EditorGizmoConfig`, or something more fine-grained) ## Changelog - Added `GizmoConfigStore` resource and `GizmoConfigGroup` trait - Added `init_gizmo_group` to `App` - Added early returns to gizmo drawing increasing performance when gizmos are disabled - Changed `GizmoConfig` and aabb gizmos to use new `GizmoConfigStore` - Changed `Gizmos` system parameter to use type based key to retrieve config - Changed resources and systems used for gizmo rendering to be generic over type based key - Changed examples (3d_gizmos, 2d_gizmos) to showcase new API ## Migration Guide - `GizmoConfig` is no longer a resource and has to be accessed through `GizmoConfigStore` resource. The default config group is `DefaultGizmoGroup`, but consider using your own custom config group if applicable. --------- Co-authored-by: Nicola Papale <nicopap@users.noreply.github.com>
2024-01-18 15:52:50 +00:00
pub fn arrow_2d(
&mut self,
start: Vec2,
end: Vec2,
Migrate from `LegacyColor` to `bevy_color::Color` (#12163) # Objective - As part of the migration process we need to a) see the end effect of the migration on user ergonomics b) check for serious perf regressions c) actually migrate the code - To accomplish this, I'm going to attempt to migrate all of the remaining user-facing usages of `LegacyColor` in one PR, being careful to keep a clean commit history. - Fixes #12056. ## Solution I've chosen to use the polymorphic `Color` type as our standard user-facing API. - [x] Migrate `bevy_gizmos`. - [x] Take `impl Into<Color>` in all `bevy_gizmos` APIs - [x] Migrate sprites - [x] Migrate UI - [x] Migrate `ColorMaterial` - [x] Migrate `MaterialMesh2D` - [x] Migrate fog - [x] Migrate lights - [x] Migrate StandardMaterial - [x] Migrate wireframes - [x] Migrate clear color - [x] Migrate text - [x] Migrate gltf loader - [x] Register color types for reflection - [x] Remove `LegacyColor` - [x] Make sure CI passes Incidental improvements to ease migration: - added `Color::srgba_u8`, `Color::srgba_from_array` and friends - added `set_alpha`, `is_fully_transparent` and `is_fully_opaque` to the `Alpha` trait - add and immediately deprecate (lol) `Color::rgb` and friends in favor of more explicit and consistent `Color::srgb` - standardized on white and black for most example text colors - added vector field traits to `LinearRgba`: ~~`Add`, `Sub`, `AddAssign`, `SubAssign`,~~ `Mul<f32>` and `Div<f32>`. Multiplications and divisions do not scale alpha. `Add` and `Sub` have been cut from this PR. - added `LinearRgba` and `Srgba` `RED/GREEN/BLUE` - added `LinearRgba_to_f32_array` and `LinearRgba::to_u32` ## Migration Guide Bevy's color types have changed! Wherever you used a `bevy::render::Color`, a `bevy::color::Color` is used instead. These are quite similar! Both are enums storing a color in a specific color space (or to be more precise, using a specific color model). However, each of the different color models now has its own type. TODO... - `Color::rgba`, `Color::rgb`, `Color::rbga_u8`, `Color::rgb_u8`, `Color::rgb_from_array` are now `Color::srgba`, `Color::srgb`, `Color::srgba_u8`, `Color::srgb_u8` and `Color::srgb_from_array`. - `Color::set_a` and `Color::a` is now `Color::set_alpha` and `Color::alpha`. These are part of the `Alpha` trait in `bevy_color`. - `Color::is_fully_transparent` is now part of the `Alpha` trait in `bevy_color` - `Color::r`, `Color::set_r`, `Color::with_r` and the equivalents for `g`, `b` `h`, `s` and `l` have been removed due to causing silent relatively expensive conversions. Convert your `Color` into the desired color space, perform your operations there, and then convert it back into a polymorphic `Color` enum. - `Color::hex` is now `Srgba::hex`. Call `.into` or construct a `Color::Srgba` variant manually to convert it. - `WireframeMaterial`, `ExtractedUiNode`, `ExtractedDirectionalLight`, `ExtractedPointLight`, `ExtractedSpotLight` and `ExtractedSprite` now store a `LinearRgba`, rather than a polymorphic `Color` - `Color::rgb_linear` and `Color::rgba_linear` are now `Color::linear_rgb` and `Color::linear_rgba` - The various CSS color constants are no longer stored directly on `Color`. Instead, they're defined in the `Srgba` color space, and accessed via `bevy::color::palettes::css`. Call `.into()` on them to convert them into a `Color` for quick debugging use, and consider using the much prettier `tailwind` palette for prototyping. - The `LIME_GREEN` color has been renamed to `LIMEGREEN` to comply with the standard naming. - Vector field arithmetic operations on `Color` (add, subtract, multiply and divide by a f32) have been removed. Instead, convert your colors into `LinearRgba` space, and perform your operations explicitly there. This is particularly relevant when working with emissive or HDR colors, whose color channel values are routinely outside of the ordinary 0 to 1 range. - `Color::as_linear_rgba_f32` has been removed. Call `LinearRgba::to_f32_array` instead, converting if needed. - `Color::as_linear_rgba_u32` has been removed. Call `LinearRgba::to_u32` instead, converting if needed. - Several other color conversion methods to transform LCH or HSL colors into float arrays or `Vec` types have been removed. Please reimplement these externally or open a PR to re-add them if you found them particularly useful. - Various methods on `Color` such as `rgb` or `hsl` to convert the color into a specific color space have been removed. Convert into `LinearRgba`, then to the color space of your choice. - Various implicitly-converting color value methods on `Color` such as `r`, `g`, `b` or `h` have been removed. Please convert it into the color space of your choice, then check these properties. - `Color` no longer implements `AsBindGroup`. Store a `LinearRgba` internally instead to avoid conversion costs. --------- Co-authored-by: Alice Cecile <alice.i.cecil@gmail.com> Co-authored-by: Afonso Lage <lage.afonso@gmail.com> Co-authored-by: Rob Parrett <robparrett@gmail.com> Co-authored-by: Zachary Harrold <zac@harrold.com.au>
2024-02-29 19:35:12 +00:00
color: impl Into<Color>,
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
) -> ArrowBuilder<'_, Config, Clear> {
self.arrow(start.extend(0.), end.extend(0.), color)
}
}
Add coordinate axes gizmo (#12211) # Objective - We introduce a gizmo that displays coordinate axes relative to a Transform*, primarily for debugging purposes. - See #9400 ## Solution A new method, `Gizmos::axes`, takes a `Transform`* as input and displays the standard coordinate axes, transformed according to it; its signature looks like this: ````rust pub fn axes(&mut self, transform: into TransformPoint, base_length: f32) { //... } ```` If my carefully placed asterisks hadn't already tipped you off, the argument here is not actually a `Transform` but instead anything which implements `TransformPoint`, which allows it to work also with `GlobalTransform` (and also `Mat4` and `Affine3A`, if the user happens to be hand-rolling transformations in some way). The `base_length` parameter is a scaling factor applied to the coordinate vectors before the transformation takes place; in other words, the caller can use this to help size the coordinate axes appropriately for the entity that they are attached to. An example invocation of this method looks something like this: ````rust fn draw_axes_system( mut gizmos: Gizmos, query: Query<&Transform, With<MyMarkerComponent>>, ) { for &transform in &query { gizmos.axes(transform, 2.); } } ```` The result is the three coordinate axes, X, Y, Z (colored red, green, and blue, respectively), drawn onto the entity: <img width="206" alt="Screenshot 2024-02-29 at 2 41 45 PM" src="https://github.com/bevyengine/bevy/assets/2975848/789d1703-29ae-4295-80ab-b87459cf8037"> Note that, if scaling was applied as part of the given transformation, it shows up in scaling on the axes as well: <img width="377" alt="Screenshot 2024-02-29 at 2 43 53 PM" src="https://github.com/bevyengine/bevy/assets/2975848/6dc1caf4-8b3e-47f7-a86a-8906d870fa72"> --- ## Changelog - Added `Gizmos::axes` in bevy_gizmos/src/arrows.rs - Fixed a minor issue with `ArrowBuilder::with_tip_length` not correctly implementing builder style (no external impact) --- ## Discussion ### Design considerations I feel pretty strongly that having no default length scale is for the best, at least for the time being, since it's very easy for the length scale to be too small, leading to the axes being hidden inside the body of the object they are associated with. That is, if the API instead looked like this: ````rust gizmos.axes(transform); // or gizmos.axes(transform).with_length_scale(3.0); ```` then I think it's a reasonable expectation that the first thing would "just work" for most applications, and it wouldn't, which would be kind of a footgun. ### Future steps There are a few directions that this might expand in the future: 1. Introduce additional options via the standard builder pattern; i.e. introducing `AxesBuilder<T: TransformPoint>` so that people can configure the axis colors, normalize all axes to a fixed length independent of scale deformations, etc. 2. Fold this functionality into a plugin (like AabbGizmoPlugin) so that the functionality becomes more-or-less automatic based on certain fixed marker components. This wouldn't be very hard to implement, and it has the benefit of making the axes more frictionless to use. Furthermore, if we coupled this to the AABB functionality we already have, we could also ensure that the plugin automatically sizes the axes (by coupling their size to the dimensions of the AABB, for example). 3. Implement something similar for 2d. Honestly, I have no idea if this is desired/useful, but I could probably just implement it in this PR if that's the case.
2024-02-29 23:52:05 +00:00
Retained `Gizmo`s (#15473) # Objective Add a way to use the gizmo API in a retained manner, for increased performance. ## Solution - Move gizmo API from `Gizmos` to `GizmoBuffer`, ~ab~using `Deref` to keep usage the same as before. - Merge non-strip and strip variant of `LineGizmo` into one, storing the data in a `GizmoBuffer` to have the same API for retained `LineGizmo`s. ### Review guide - The meat of the changes are in `lib.rs`, `retained.rs`, `gizmos.rs`, `pipeline_3d.rs` and `pipeline_2d.rs` - The other files contain almost exclusively the churn from moving the gizmo API from `Gizmos` to `GizmoBuffer` ## Testing ### Performance Performance compared to the immediate mode API is from 65 to 80 times better for static lines. ``` 7900 XTX, 3700X 1707.9k lines/ms: gizmos_retained (21.3ms) 3488.5k lines/ms: gizmos_retained_continuous_polyline (31.3ms) 0.5k lines/ms: gizmos_retained_separate (97.7ms) 3054.9k lines/ms: bevy_polyline_retained_nan (16.8ms) 3596.3k lines/ms: bevy_polyline_retained_continuous_polyline (14.2ms) 0.6k lines/ms: bevy_polyline_retained_separate (78.9ms) 26.9k lines/ms: gizmos_immediate (14.9ms) 43.8k lines/ms: gizmos_immediate_continuous_polyline (18.3ms) ``` Looks like performance is good enough, being close to par with `bevy_polyline`. Benchmarks can be found here: This branch: https://github.com/tim-blackbird/line_racing/tree/retained-gizmos Bevy 0.14: https://github.com/DGriffin91/line_racing ## Showcase ```rust fn setup( mut commands: Commands, mut gizmo_assets: ResMut<Assets<GizmoAsset>> ) { let mut gizmo = GizmoAsset::default(); // A sphere made out of one million lines! gizmo .sphere(default(), 1., CRIMSON) .resolution(1_000_000 / 3); commands.spawn(Gizmo { handle: gizmo_assets.add(gizmo), ..default() }); } ``` ## Follow-up work - Port over to the retained rendering world proper - Calculate visibility and cull `Gizmo`s
2024-12-04 21:21:06 +00:00
impl<Config, Clear> GizmoBuffer<Config, Clear>
Contextually clearing gizmos (#10973) # Objective Allow `Gizmos` to work in `FixedUpdate` without any changes needed. This changes `Gizmos` from being a purely immediate mode api, but allows the user to use it as if it were an immediate mode API regardless of schedule context. Also allows for extending by other custom schedules by adding their own `GizmoStorage<Clear>` and the requisite systems: - `propagate_gizmos::<Clear>` before `update_gizmo_meshes` - `stash_default_gizmos` when starting a clear context - `pop_default_gizmos` when ending a clear context - `collect_default_gizmos` when grabbing the requested gizmos - `clear_gizmos` for clearing the context's gizmos ## Solution Adds a generic to `Gizmos` that defaults to `Update` (the current way gizmos works). When entering a new clear context the default `Gizmos` gets swapped out for that context's duration so the context can collect the gizmos requested. Prior work: https://github.com/bevyengine/bevy/pull/9153 ## To do - [x] `FixedUpdate` should probably get its own First, Pre, Update, Post, Last system sets for this. Otherwise users will need to make sure to order their systems before `clear_gizmos`. This could alternatively be fixed by moving the setup of this to `bevy_time::fixed`? PR to fix this issue: https://github.com/bevyengine/bevy/pull/10977 - [x] use mem::take internally for the swaps? - [x] Better name for the `Context` generic on gizmos? `Clear`? --- ## Changelog - Gizmos drawn in `FixedMain` now last until the next `FixedMain` iteration runs.
2024-04-23 00:16:12 +00:00
where
Config: GizmoConfigGroup,
Clear: 'static + Send + Sync,
{
Add coordinate axes gizmo (#12211) # Objective - We introduce a gizmo that displays coordinate axes relative to a Transform*, primarily for debugging purposes. - See #9400 ## Solution A new method, `Gizmos::axes`, takes a `Transform`* as input and displays the standard coordinate axes, transformed according to it; its signature looks like this: ````rust pub fn axes(&mut self, transform: into TransformPoint, base_length: f32) { //... } ```` If my carefully placed asterisks hadn't already tipped you off, the argument here is not actually a `Transform` but instead anything which implements `TransformPoint`, which allows it to work also with `GlobalTransform` (and also `Mat4` and `Affine3A`, if the user happens to be hand-rolling transformations in some way). The `base_length` parameter is a scaling factor applied to the coordinate vectors before the transformation takes place; in other words, the caller can use this to help size the coordinate axes appropriately for the entity that they are attached to. An example invocation of this method looks something like this: ````rust fn draw_axes_system( mut gizmos: Gizmos, query: Query<&Transform, With<MyMarkerComponent>>, ) { for &transform in &query { gizmos.axes(transform, 2.); } } ```` The result is the three coordinate axes, X, Y, Z (colored red, green, and blue, respectively), drawn onto the entity: <img width="206" alt="Screenshot 2024-02-29 at 2 41 45 PM" src="https://github.com/bevyengine/bevy/assets/2975848/789d1703-29ae-4295-80ab-b87459cf8037"> Note that, if scaling was applied as part of the given transformation, it shows up in scaling on the axes as well: <img width="377" alt="Screenshot 2024-02-29 at 2 43 53 PM" src="https://github.com/bevyengine/bevy/assets/2975848/6dc1caf4-8b3e-47f7-a86a-8906d870fa72"> --- ## Changelog - Added `Gizmos::axes` in bevy_gizmos/src/arrows.rs - Fixed a minor issue with `ArrowBuilder::with_tip_length` not correctly implementing builder style (no external impact) --- ## Discussion ### Design considerations I feel pretty strongly that having no default length scale is for the best, at least for the time being, since it's very easy for the length scale to be too small, leading to the axes being hidden inside the body of the object they are associated with. That is, if the API instead looked like this: ````rust gizmos.axes(transform); // or gizmos.axes(transform).with_length_scale(3.0); ```` then I think it's a reasonable expectation that the first thing would "just work" for most applications, and it wouldn't, which would be kind of a footgun. ### Future steps There are a few directions that this might expand in the future: 1. Introduce additional options via the standard builder pattern; i.e. introducing `AxesBuilder<T: TransformPoint>` so that people can configure the axis colors, normalize all axes to a fixed length independent of scale deformations, etc. 2. Fold this functionality into a plugin (like AabbGizmoPlugin) so that the functionality becomes more-or-less automatic based on certain fixed marker components. This wouldn't be very hard to implement, and it has the benefit of making the axes more frictionless to use. Furthermore, if we coupled this to the AABB functionality we already have, we could also ensure that the plugin automatically sizes the axes (by coupling their size to the dimensions of the AABB, for example). 3. Implement something similar for 2d. Honestly, I have no idea if this is desired/useful, but I could probably just implement it in this PR if that's the case.
2024-02-29 23:52:05 +00:00
/// Draw a set of axes local to the given transform (`transform`), with length scaled by a factor
/// of `base_length`.
///
/// This should be called for each frame the axes need to be rendered.
///
/// # Example
/// ```
/// # use bevy_gizmos::prelude::*;
/// # use bevy_ecs::prelude::*;
/// # use bevy_transform::components::Transform;
/// # #[derive(Component)]
/// # struct MyComponent;
/// fn draw_axes(
/// mut gizmos: Gizmos,
/// query: Query<&Transform, With<MyComponent>>,
/// ) {
/// for &transform in &query {
/// gizmos.axes(transform, 1.);
/// }
/// }
/// # bevy_ecs::system::assert_is_system(draw_axes);
/// ```
pub fn axes(&mut self, transform: impl TransformPoint, base_length: f32) {
let start = transform.transform_point(Vec3::ZERO);
let end_x = transform.transform_point(base_length * Vec3::X);
let end_y = transform.transform_point(base_length * Vec3::Y);
let end_z = transform.transform_point(base_length * Vec3::Z);
self.arrow(start, end_x, RED);
self.arrow(start, end_y, GREEN);
self.arrow(start, end_z, BLUE);
}
/// Draw a set of axes local to the given transform (`transform`), with length scaled by a factor
/// of `base_length`.
///
/// This should be called for each frame the axes need to be rendered.
///
/// # Example
/// ```
/// # use bevy_gizmos::prelude::*;
/// # use bevy_ecs::prelude::*;
/// # use bevy_transform::components::Transform;
/// # #[derive(Component)]
/// # struct AxesComponent;
/// fn draw_axes_2d(
/// mut gizmos: Gizmos,
/// query: Query<&Transform, With<AxesComponent>>,
/// ) {
/// for &transform in &query {
/// gizmos.axes_2d(transform, 1.);
/// }
/// }
/// # bevy_ecs::system::assert_is_system(draw_axes_2d);
/// ```
pub fn axes_2d(&mut self, transform: impl TransformPoint, base_length: f32) {
let start = transform.transform_point(Vec3::ZERO);
let end_x = transform.transform_point(base_length * Vec3::X);
let end_y = transform.transform_point(base_length * Vec3::Y);
self.arrow_2d(start.xy(), end_x.xy(), RED);
self.arrow_2d(start.xy(), end_y.xy(), GREEN);
}
Add coordinate axes gizmo (#12211) # Objective - We introduce a gizmo that displays coordinate axes relative to a Transform*, primarily for debugging purposes. - See #9400 ## Solution A new method, `Gizmos::axes`, takes a `Transform`* as input and displays the standard coordinate axes, transformed according to it; its signature looks like this: ````rust pub fn axes(&mut self, transform: into TransformPoint, base_length: f32) { //... } ```` If my carefully placed asterisks hadn't already tipped you off, the argument here is not actually a `Transform` but instead anything which implements `TransformPoint`, which allows it to work also with `GlobalTransform` (and also `Mat4` and `Affine3A`, if the user happens to be hand-rolling transformations in some way). The `base_length` parameter is a scaling factor applied to the coordinate vectors before the transformation takes place; in other words, the caller can use this to help size the coordinate axes appropriately for the entity that they are attached to. An example invocation of this method looks something like this: ````rust fn draw_axes_system( mut gizmos: Gizmos, query: Query<&Transform, With<MyMarkerComponent>>, ) { for &transform in &query { gizmos.axes(transform, 2.); } } ```` The result is the three coordinate axes, X, Y, Z (colored red, green, and blue, respectively), drawn onto the entity: <img width="206" alt="Screenshot 2024-02-29 at 2 41 45 PM" src="https://github.com/bevyengine/bevy/assets/2975848/789d1703-29ae-4295-80ab-b87459cf8037"> Note that, if scaling was applied as part of the given transformation, it shows up in scaling on the axes as well: <img width="377" alt="Screenshot 2024-02-29 at 2 43 53 PM" src="https://github.com/bevyengine/bevy/assets/2975848/6dc1caf4-8b3e-47f7-a86a-8906d870fa72"> --- ## Changelog - Added `Gizmos::axes` in bevy_gizmos/src/arrows.rs - Fixed a minor issue with `ArrowBuilder::with_tip_length` not correctly implementing builder style (no external impact) --- ## Discussion ### Design considerations I feel pretty strongly that having no default length scale is for the best, at least for the time being, since it's very easy for the length scale to be too small, leading to the axes being hidden inside the body of the object they are associated with. That is, if the API instead looked like this: ````rust gizmos.axes(transform); // or gizmos.axes(transform).with_length_scale(3.0); ```` then I think it's a reasonable expectation that the first thing would "just work" for most applications, and it wouldn't, which would be kind of a footgun. ### Future steps There are a few directions that this might expand in the future: 1. Introduce additional options via the standard builder pattern; i.e. introducing `AxesBuilder<T: TransformPoint>` so that people can configure the axis colors, normalize all axes to a fixed length independent of scale deformations, etc. 2. Fold this functionality into a plugin (like AabbGizmoPlugin) so that the functionality becomes more-or-less automatic based on certain fixed marker components. This wouldn't be very hard to implement, and it has the benefit of making the axes more frictionless to use. Furthermore, if we coupled this to the AABB functionality we already have, we could also ensure that the plugin automatically sizes the axes (by coupling their size to the dimensions of the AABB, for example). 3. Implement something similar for 2d. Honestly, I have no idea if this is desired/useful, but I could probably just implement it in this PR if that's the case.
2024-02-29 23:52:05 +00:00
}