On Wed, Dec 07, 2005 at 09:11:44PM +0000, Robin Green wrote:
On Wednesday 07 December 2005 20:19, you wrote:
On Wed, Dec 07, 2005 at 07:47:46PM +0000, Robin Green wrote:
Some day you may thank for this verbosity, because it encourages you do program in a purely functional way making your program more friendly for SMP execution.
You are mistaken. The verbosity is necessary if you want "visual" referential transparency, but not necessary if you only want pure functional programming. Only the latter is helpful for optimisability. I am hoping to write a paper on this topic.
I am afraid I don't understand.
Let's say you want to write a function
seqPair :: (Monad m) => (m a, m b) -> m (a, b)
which returns a computation which does the left computation followed by the right computation (i.e. it's like the sequence function, but for pairs instead of lists).
In Haskell you could write this as:
seqPair mx my = do x <- mx y <- my return (x, y)
However, wouldn't it be nice if we could write something like (warning: hypothetical syntax)
I'm not sure. I wouldn't be the first one to jump with joy if it was allowed :-)
seqPair $= (,)
or (the slightly less cryptic version)
seqPair x y $= (x, y)
Wouldn't they be different, the first one forcing (,) to WHNF (NOP), and the second one forcing x and y to WHNF? Would you want seqPair written with $= to still have a type involving the Monad class?
This is not referentially transparent because it is not equivalent to
seqPair x y $= swap (y, x) where swap (a, b) = (b, a)
(can you see why not?)
Only guessing, because I am not sure what $= is supposed to do, but is it because y would be evaluated before x? I can't find anything wrong with it, besides the strange semantics of $=. Anyway, I didn't think my initial statement was so controversial. I even said "may", not "will". Best regards Tomasz -- I am searching for a programmer who is good at least in some of [Haskell, ML, C++, Linux, FreeBSD, math] for work in Warsaw, Poland