On Wed, Jul 30, 2008 at 12:38:20PM -0700, Jason Dagit wrote:
> > Hmm, no! ByteString is 99.9% Haskell, with 7 lines of C for two
> > functions -- and those are obsolete.
> Oh, sorry for spreading misinformation! It's great to hear that the rock
> solid performance of ByteString is pretty much just Haskell.
Indeed, in fact, there is every reason to believe that low level unboxed
haskell code will actually compile to be _faster_ than C. The haskell
compiler has a lot more information about aliasing and is free to
rearrange data and play with calling conventions that are not possible
in C. I have found this quite fun to play with in jhc. It almost makes
me want to take jhcs back end and attach it to a pure unboxed monadic
language meant to replace C for systems programming... I'd call it
'undo' for the unboxed do monad. :)
Imagine being able to use type class based polymorphism, rewrite rules,
and monadic reasoning when writing low level code or device drivers. It
is enough to make me want to get back into operating systems design. (a
fancy of mine back in the day)
John
--
John Meacham - ⑆repetae.net⑆john⑈
_______________________________________________
darcs-users mailing list
[email protected]
http://lists.osuosl.org/mailman/listinfo/darcs-users
