>>106350826
>Type theory doesn't get useful until homotophy type theory
Nonsense, I use ocaml a lot and a lot of the reason why is because of the module system, which itself is basically just system f omega. It's probably the best metaprogramming experience I've ever had, like if sepples templates didn't suck dick and balls.
Dependent type theory is also obviously useful, but the undecidability of higher kinded unification requires proofs, and while that doesn't make it outright useless, it does mean you need serious manpower to do non-trivial things in it. I enjoyed playing around with Idris, but I'd need a more mature compiler and a team of good people to alleviate the burden of the proofs.

What do you like about homotopy type theory anon? Sell me on it.