ra’s Twitter Archive—№ 43,950

@alexnixon_uk @lexi_lambda @brendanzab @puffnfresh @typedrat I remember Rich asserting that a complete correct Haskell type for transducers would be difficult if not impossible to derive? it's in the original talk. I mean beyond a handwavy rhetorical device, something that would pass the type checker while maintaining transducer semantics.