[Haskell-cafe] Re: evaluation semantics of bind

[Achim Schneider]

Gregg Reynolds <dev at mobileink.com> wrote:

>     getChar >>= \x -> getChar
> 
> An optimizer can see that the result of the first getChar is discarded
>
It isn't discarded. The first getChar results in a value of type IO
Char, always and ever. Whether or not the Char inside the IO Char gets
evaluated or not is another question... >>= could not care less about
what you do to the values that are _inside_ the monad. 

If you look at the type
(>>=) :: (Monad m) => m a -> (a -> m b) -> m b
, you will find it hard to do much of significance to "a" and/or "b"[1]

Whether or not a monad behaves strictly or lazyily wrt. to "m" depends
on the definition of the functions that come with it: IO happens to be
strict, [] to be lazy. This is nothing that messes with semantics,
though, it celebrates them.


[1] Well, you can pass them to id, seq or par, and replace them with
undefined...

-- 
(c) this sig last receiving data processing entity. Inspect headers
for copyright history. All rights reserved. Copying, hiring, renting,
performance and/or quoting of this signature prohibited.