News for hackers. Stuff that matters.
Re:What is the asymptotic cost of purely functional programm... by jerf on Wednesday April 11, 2012 @01:04PM (#3828448)
In addition to the great answer mentioned there, it should also be mentioned that every language referred to as functional still has the ability to use mutable arrays in some manner. Yes, even Haskell, it just requires you to carefully control where the mutation is used.
Re:What is the asymptotic cost of purely functional programm... by greiskul on Wednesday April 11, 2012 @02:44PM (#3828972)
Yes, I think the great lesson we can take from functional programming is not of being functional 100% of the time, but of not having mutable global state. Your inner loops, local variables, etc. can all be mutable and nobody will care, as long as your interface is functional.
Re:What is the asymptotic cost of purely functional programm... by fpgeek on Wednesday April 11, 2012 @10:50PM (#3830797)
Yes and no. It is true that what matters most of the time is a functional interface, not how that is achieved. That being said, it is much harder to successfully implement a functional interface if you are using side-effects willy-nilly inside.