4.6
Welp I will keep my question simple, does non uniform scale still matter for physics shapes, especially in my case the only interaction with physics I will have is raycasts, literally nothing more.
Especially now with Jolt, it just seems handy to use normal Node3D transforms instead of rescaling the shape itself.
IMHO, scaling is always bad, especially non-uniform. I would avoid it under all circumstances. There are some situations where it’s ok, when you know what you are doing.
I am mainly using it so I can rescale my Collision and MeshInstance at the same time, I know how inheriting transforms can mess stuff up and all but in this case it can come quite in handy if you set it up nicely.
I am curious what scenarios do you think would stuff get messed up and what should I be wary of?
Is there a reason to not use a csg node? It is effectively just a mesh and collision combined into one node.
It can mess up so many things. On the mesh side, auto LOD, z-sorting and parameters in shadow calculation (like bias) (not sure if this is the case in Godot).
In physics it can mess with your values like masses, collision margin etc.
CSG is only for gray (grey) boxing. I wouldn’t use it in the final game. And you cant texture it with UVs, and it’s slow. Build your CSGs and export them to blender as a reference.
Wow, I really didn’t know it extends that much, it just felt like a little bit of a hassle not to be able to use Node3D transforms but this makes a lot of sense now, thanks!