> You write enough system software in functional languages and you realize you just traded one set of headaches for a different but equally infuriating set, and

> You write enough system software in functional languages and you realize you just traded one set of headaches for a different but equally infuriating set, and that most headaches are sociological aspects of teams writing code, not just isolated gifted engineers separately writing perfectly idiomatic functional programming designs. I feel like I might know what you mean, but I'd be interested if you could elaborate. (Also, isn't OCaml multiparadigmatic?)