On the unsafety of interleaved I/O

1 · Edward Kmett · July 22, 2015, 3:29 p.m.
One area where I'm at odds with the prevailing winds in Haskell is lazy I/O. It's often said that lazy I/O is evil, scary and confusing, and it breaks things like referential transparency. Having a soft spot for it, and not liking most of the alternatives, I end up on the opposite side when the [...]...