Ben Visness

It's true that a manual update process means breaking changes are less of an issue, but I still want to preserve backwards compatibility to the fullest extent we can. And...

Well this is supposed to be for a breaking change, so a little bit is to be expected. Maybe we can have only the explicit versions out of the box,...

@StrangeZak Yeah this is a 2.0 change, so breaking things is fine and expected (to a point). I wasn't really considering the macros as a means of achieving backwards compatibility;...

And you know, the magic of C/C++ is that if you really want to use `HMM_Orthographic` without the suffix, you can just define it yourself. 🙂

If we want to make it really robust, I think we should add mat2 and mat3 with all the same features as mat4. It shouldn't be too difficult to do,...

[Comment from the Handmade network:](https://handmade-math.handmade.network/blogs/p/2740-handmade_math_is_back#13531) > Glad to hear you are looking to support matrix inverses. Currently working with 3D picking and this would be extremely helpful. And since this...

I'm working on the SpiderMonkey test generator @eqrion mentioned, and a sizable chunk of our program's runtime is spent in [`Span::linecol_in`](https://docs.rs/wast/48.0.0/wast/token/struct.Span.html#method.linecol_in). This is essentially quadratic behavior; only the `offset` is...

I'll give it a try tomorrow. I doubt the memory impact will be significant enough to warrant a side data structure that would take up space on its own, plus...

HMM 2.0 preserves the C++ overloads.

Fixed in HMM 2.0.