From HaskellWiki
Revision as of 10:36, 20 July 2007 by Piet Delport (talk | contribs) (over-costanding comonads)

Jump to: navigation, search
<Philippa> do we have a case of haskell faster than C on a platform where GHC
           compiles via C and doesn't screw with the output yet?
<jethr0> wouldn't that just be a blatant case of slow c benchmarking code? :)
<dons> the concurrency or binary tree benchmarks?
<jethr0> someone could put the haskell intermediate c code up as the c benchmark *g*
<musasabi> yes, 30000 lines of C? ;)
seen on comp.lang.functional:

 From: Ashley Yakeley <ashley@semantic.org>
 Subject: Re: Type advocacy
 Newsgroups: comp.lang.functional
 Date: Thu, 11 Oct 2001 21:16:20 -0700
 Organization: Myself

 In article <9pdvgc$u3d$1@news.fas.harvard.edu>, Ken Shan
 <ken@digitas.harvard.edu> wrote:

 > I am preparing a three-minute talk to tell incoming graduate students
 > at my school about types.

 Oh like that's going to work. You'd be better off selling T-shirts that
 say "WHAT PART OF" (and then the Hindley-Milner prinicipal-type
 algorithm) "DON'T YOU UNDERSTAND?".

 If anyone gives you any lip, ask them how to find the square-root of a
 string. Everything else follows on from that.

 > What pointers should I give?

 Safe ones.

 Ashley Yakeley, Seattle WA
<kaol> @src liftM
<lambdabot> liftM f m1 = do { x1 <- m1; return (f x1) }
<kaol> @src liftM2
<lambdabot> liftM2 f m1 m2 = do { x1 <- m1; x2 <- m2; return (f x1 x2) }
<osfameron> does that lift monads which are twice as heavy?
<LoganCapaldo> No, it just lifts a monad in each hand
<osfameron> harr!
<byorgey> you know what they say, a monad in each hand is better than two in
	  the bush... no wait...
<jfredett> if this were not a family chat channel, I might say something about that...
<DRMacIver> Hand me a long enough monad and I will move the world?
<cjeris> jfredett: yeah, well, the first time I tried to explain what was
	 different about Haskell to my wife, her response was "Monad?  Is that
	 when you only have one ball?"
* EvilTerran . o O ( is a comathematician a device for turning theorems into coffee? )
<oerjan> EvilTerran: no, it's for turning cotheorems into ffee.
<slava> oerjan: it's not clear that the category of cocoffee is isomorphic to ffee
<oerjan> slava: bah, it's mpletely coclear.
<roconnor> comonads are just as hard to understand as monads.
<Toxaris> you have to co-understand them?
<quicksilver> Toxaris: I believe you actually have to over-costand them