I think as a computer scientist you should really have at least some familiarity with both Haskell and one of the ML languages, just because it'll expand your t

I think as a computer scientist you should really have at least some familiarity with both Haskell and one of the ML languages, just because it'll expand your thinking. Whether to invest deeply in one of them is a matter of whether you have the time (especially if it's just as a hobby) and how well one or the other matches your preferences. I don't generally evangelize OCaml – because as a language it can be a bit of a mixed bag – but end up using it a lot because in the end it's a pragmatic language and a really solid workhorse. The niche that it fills for me is that of a statically typed language with native compilation and a GC, a niche that has historically not been served all that well. Things I like: * The language itself is not too opinionated. It's a multi-paradigm, functional-first language, that allows me to use imperative or OO styles when that's the best fit. * The compiler is fast. * While the ecosystem has some gaps, it's very solid in those areas that I need (such as systems programming or language tooling). * The language does not prioritize speed at the expense of correctness; this includes the compiler. There are languages where, let's say, I have trust issues. Code quality is still pretty darn good. * I really like ML functors for parametric polymorphism. While more verbose than other approaches, they are simple and powerful. * Having a time-traveling debugger can occasionally be very helpful. * For all their warts, the language and the compiler are mature.