News for hackers. Stuff that matters.

A comment on Can functional programming be liberated from the von Neumann paradigm? (2010). Parent comment

Re:Can functional programming be liberated from the von Neum... by jdiez17 on Thursday January 01, 2015 @09:43AM (#8823033)
But that's an optimisation. Which can be switched off to get the benefits from parallelism. This is a tradeoff that programmers would traditionally have to consider, but with functional programming the compiler can figure it out for us.
[ Parent | on HN ]
Re:Can functional programming be liberated from the von Neum... by chongli on Thursday January 01, 2015 @10:19AM (#8823141)
That's assuming that "figuring it out for us" is decidable. Many of the things we wish the compiler could do for us are not.
[ Parent | on HN ]
Re:Can functional programming be liberated from the von Neum... by dllthomas on Thursday January 01, 2015 @02:46PM (#8824005)
"It's not decidable" is not a guarantee that humans can do better than a compiler. It means any approach will sometimes be wrong, but a compiler might well (in principle) be wrong less often than a human.

This is separate from the "a human has a lot of additional context that's hard to express" argument, which is valid but has nothing to do with decidability.

[ Parent | on HN ]