Conal Elliott

ARANK

This post describes an idea I’ve had in mind for years and have chatted about with friends but haven’t written down before.Functional programmers used to worry about how to solve “the I/O problem”. Functional programming (FP) eschews any notion of side-effect or order of execution, and instead has an order-independent notion of evaluation. In contrast, input & output (I/O) seems to be an inherently effectful, ordered notion.For a long time, I’ve been dissatisfied with the current Haskell solution to “the I/O problem”. Of course it’s very clever and effective. With “monadic IO” we can program both functionally and imperatively, and the imperative semantics doesn’t compromise the functional semantics.I look at monadic IO not as any solution to functional I/O, but rather as a (clever) way not to bring I/O into the functional world. There’s a ton of confusion buzzing around monadic IO, so let me try to restate more carefully: “monadic IO” is a way to generate imperative comp…