Aevyrie

Thanks @bitshifter, appreciate this review. I don't have your experience here, but I can see it being advantageous to lean on the type system at the game engine level, while...

> I think its worth while to generate bounding volumes(AABB) for shapes that aren't boxes/spheres. Most game engines will use this information to cull meshes using frustum culling. It would...

> So I think the `NormalizedDirection` wrapper type is the only thing that I think we _need_ to get right at this stage. Once that's done this has my seal...

I might have a way to do this in a zero-cost way using the typestate pattern. To absolutely minimize the use of `normalize` we could memoize the result _only when...

My take is that in general, bounding and collision implementation detail and discussion are out of scope. To reframe - the critical question I want to answer is: _**Are the...

> 1. What parts of this RFC are already covered by Parry, what would we have to implement additionally? The only overlap lies in the primitive types themselves. > 2....

@cart, this is ready for review. I'd like to get your feedback before making any further changes.

> > I've outlined [here](https://github.com/aevyrie/rfcs/blob/patch-1/rfcs/12-primitive-shapes.md#rationale-and-alternatives) why I think we need to have Bevy-owned primitive shape types. > > That paragraph is: > > Rationale and alternatives > > > >...

@alphamodder > I'm very confused by the code example in the Direction Normalization section, as `impl Trait` is not permitted or planned to be permitted (afaik) in `struct`s, and it's...

@alice-i-cecile with so much churn on this RFC, I'd like to update my implementation attempt, and we can collect concrete feedback there. I have some thoughts on further modifications to...