joshsharp / braid Goto Github PK

Disclaimer: I come here in peace, in an effort to understand your thought process, not to proselytize or ask you to change direction.

Looking at your goals, it seems to me that writing a Go backend for ocamlc (e.g. it already has a JS backend using Bucklescript, a bytecode backend, and a native code backend) fulfills all the goals.

• Pair an OCaml-like language with the benefits of the Go platform which are:
  • speed
  • concurrency
  • static binaries
  • a healthy ecosystem
• Bring powerful FP concepts to Go
• Get around Go's lack of generics
• Interop with Go code
• Ability to use Go stdlib

I'm also listing some other pros and cons here.

Pros:

You benefit from

1. ReasonML's front-end which is actively maintained.
2. OCaml's type-checker which is actively maintained.
3. Having alternate backends to test your backend for correctness.

Unsure if Pros or Cons:

1. You'll end up writing the code in OCaml instead of Go.
2. A lot of the type system is already in place, however it may not be exactly what you're looking for.

Cons:

1. Different licensing/copyright setup compared to what you currently have.
2. You need to dive into a large, unfamiliar codebase instead of starting from a clean slate.

Of course, there are lots of more considerations that I haven't listed here.

I'd love to hear your opinion on this idea/why you didn't choose to go this route?