Patricio Whittingslow

More complete interface: ```go type Transformer interface { Apply(Vec) Vec // Returns the inverse transform of Transformer Inv() Transformer } ```

So like `nthRoot:= quat.Number{Real:math.Pow(q.Real, 1/n), Imag: math.Pow(q.Imag, 1/n), Kmag:math.Pow(q.Kmag, 1/n),Jmag:math.Pow(q.Jmag, 1/n)}` ? I can't find anything on this algorithm with a cursory google search.

Awesome! So does adding a second type parameter `P` for the point type make sense to you?

Not entirely happy with how it turned out. Feel like there are some knobs that could be turned * kdtree algorithm: https://github.com/soypat/go-play3d/blob/main/kdtree/kdtree.go * implementation of kdtree interface: https://github.com/soypat/go-play3d/blob/main/triangle_kdtree.go Instantiation: `kdtree.New[kdPoint,...

@kortschak I've added methods `IsZero()` and `SetTranslate`. IsZero lets users check result of `Inv()`. I'm also on the fence about the `Add(v Vec)` signature. I feel like `Translate` fits in...

I suggest making a compromise- removing API which we are uncertain about quality and leaving in the bits which seem OK. I believe the most shaky bits are * Warp...

This seems out of line with what I feel is the purpose of the `drivers` namespace, to provide a hardware abstraction layer. This seems more natural-sciences or physics library oriented....

Would it be worth introducing a physics-like package, much like the [periph.io package](https://pkg.go.dev/periph.io/x/periph/conn/physic)?

I'll comment on the progress: I've managed to port parts of the arduino ENC28j60 library but I am having trouble reading/writing anything at all through the SPI connection. I think...

It just occurred to me the arguments to `Measurement()` could be _units_, so ```go func (d Dev) Temperature(unit units.Temperature) (T int32) { switch unit { case units.Kelvin: return d.K case...