Awesome
Unity.Mathematics.FixedPoint
A Fixed-point extension of Unity's C# math library based on FixedMath.Net and Unity.Mathematics. One of the main reasons for using this library instead of the built-in one is support for cross-platform determinism, until Unity officially supports it with Burst/DOTS.
The intention is to keep the API as close as possible to Unity.Mathematics. This should make it easy to convert your project to use floating point math if needed.
Usage
You can use this library in your Unity game by adding this repository to the package manifest file in your Unity project. PROJECT_ROOT/Packages/manifest.json
:
{
"dependencies": {
"com.danielmansson.mathematics.fixedpoint": "https://github.com/danielmansson/Unity.Mathematics.FixedPoint.git"
}
}
Testing and developing
The easiest way to iterate on this package is to clone this repository and the test project side-by-side and open the test project in Unity. This resolves the required dependencies (Unity.Mathematics) and makes the test runner available.
Example project
A simple example project can be found here. It contains two side-by-side simulations, one using a floating point implementation and the other using fixed point.
Missing features
Unity.Mathematics.FixedPoint is not feature complete yet. This is missing:
- quaternion
- fpmath.tanh
- fpmath.cosh
- fpmath.sinh
- fpmath.exp
- fpmath.log10
Method stubs are added to match the API, but they are marked as obsolete with a compile error.
Maintaining changes from Unity.Mathematics
Unity.Mathematics is using code generation to create their vector and matrix types. Most of the changes in this repository from Unity.Mathematics has been in VectorGenerator.cs
and fpmath.cs
.
The plan is to keep this repository almost up to date with Unity.Mathematics, so changes has been isolated as much as possible. However, VectorGenerator.cs
was not made with external extensibility in mind. When there are internal changes to Unity.Mathematics in upcoming versions, this will require manual patching.
fpmath.cs
contains the common fixed point math operations. The implementations are based on the floating point methods in Unity.Mathematics math.cs
.
Precision
More tests are needed to verify the precision of all fixed point operations. There might be intermediate calculations unsuitable for fixed point in the current implementation.
Licensing
This project is licensed under the MIT License (LICENSE.md)
Unity.Mathematics (Unity Companion License)
FixedMath.Net (Apache License, Version 2.0)