Hacker News Reader

Monadic futures are a fairly effective solution to the problem.

No callback hell, just a clean linear flow of imperative looking non-blocking code that runs concurrently but still guarantees order of execution.

Haskell is nothing but indirect action (lazy evaluation...), which is why referential transparency is so important to the language.

In my experience referential transparency increases simplicity and readability in most languages.

Yes, but so does keeping your code first order, avoiding callback functions even without the hazard of side effects.