Joshua Barretto

Thanks for the quick response! > For 1, you should be able to just use a regular type variable I thought this might be the case but I was unsure....

Apologies, I'm still a little confused. In the demo on this page: https://blog.polybdenum.com/2020/08/08/subtype-inference-by-example-part-6-numeric-types-and-operators.html I can type in the following expression without a type error. ``` if true then 42 else...

Ah, the code is implicitly generating sum types? In this case, I am, if anything, more confused as to how one extracts that information from the type checker core.

To clarify: In `cubiml-demo`, the type-checker simply enforces that the program is well-typed and then transpiles the code to a dynamically-typed language, JavaScript. If one wishes to actually compile the...

Thanks for the information! I think perhaps I'll take some inspiration from `cubiml` but attempt implementing some alternative algorithm. Once again, thanks for being patient with me!

This is not great and is likely to be an actual soundness bug that affects real-world code. The `DashMap` will reallocate as it grows, leaving a dangling pointer in place...

I've been trying to find the time to get round to this, but I'm not currently able to schedule any. I might be able to get this done during the...

   A few screenshots of dungeons generated using `vek`.

Perhaps there could be a convenience type like `struct MAdd(T);` that adds a `MulAdd` implementation in terms of `Mul` and `Add` to allow use with non-standard types?

Thanks for making this switch! The reason I opened the issue is that I'm working on a platform without any real acceleration hardware or intrinsics at all, to falling back...