Re: [Caml-list] Purity and lazyness

[Alain Frisch <alain@frisch.fr>]

On 01/07/2011 04:35 PM, Dario Teixeira wrote:
> that is, I don't see why purity would require lazyness.

I'd say that while purity certainly does not require laziness, they are 
a natural match: purity can benefit from laziness as an alternative to 
mutability in some situations.

For instance, you cannot do memoization of unary functions like you 
would do in OCaml, using a mutable datastructure as a cache. You can 
pass this cache around or use a monad to maintain it, but using laziness 
gives a natural alternative: extend the object data structure itself 
with an extra slot to hold the result of the function to be memoized and 
you want this extra slot to be computed lazily.



Alain