February 2007 - Haskell-Cafe

ANNOUNCE: nobench: Haskell implementaion benchmarks. GHC v Hugs v Yhc v NHC v ...
by dons＠cse.unsw.edu.au 20 Feb '07

20 Feb '07

Following recent discussion about a cross-implementation performance benchmark suite, based on nofib, I've gone and combined nofib with the great language shootout programs, and rewritten the build system to support cross implementation measurements. The result is: nobench http://www.cse.unsw.edu.au/~dons/nobench.html The benchmark suite runs regularly, and currently reports the speed of each program in the suite, running under each system. The results are quite interesting. The most recent run is available: http://www.cse.unsw.edu.au/~dons/nobench/bench.results http://www.cse.unsw.edu.au/~dons/nobench/bench.log The programs are a mixture of traditional nofib style Haskell, with more performance-tuned code from the shootout. More tweaking is required to help better support nhc and yhc (and jhc, and ...). The entire benchmark set and framework is available via darcs: darcs get http://www.cse.unsw.edu.au/~dons/code/nobench Currently todo are porting the rest of nofib, pretty graphs of the results (and html), and memory use measurements. Patches welcome! Cheers, Don

9 12

our worst unsafePerformIO nightmares are upon us!
by Nicolas Frisby 19 Feb '07

19 Feb '07

http://www.thinkgeek.com/geektoys/cubegoodies/86b8/ Now you can really show your coders why unsafePerformIO is to be avoided! -Nick ps - Please don't consider this post a product advertisement or endorsement of any kind.

2 1

Is lazyness make big difference?
by Nick 19 Feb '07

19 Feb '07

Hi all, (Another topic stolen from a Russian forum discussion). As everyone know, there are lot of strict languages, that have possibilities to "switch on" lazy evaluation when needed. But the examples that I saw on Haskell, there was not much use of lazy evaluation, often there were just several lazy points, and the rest could be done strictly without loss of generality. For example, in finding primes: main = print primes primes = 2:filter is_prime [3,5..] is_prime n = all (\p-> n `mod` p /= 0) (takeWhile (\p-> p*p<=n) primes) We can rewrite this in strict languages with lazy constructs. For example, in Scala (of course stream is not only lazily evaluated thing there) def main(args: Array[String]): Unit = { val n = Integer.parseInt(args(0)) System.out.println(primes(ints(2)) take n toList) } def primes(nums: Stream[Int]): Stream[Int] = Stream.cons(nums.head, primes ((nums tail) filter (x => x % nums.head != 0)) ) def ints(n: Int): Stream[Int] = Stream.cons(n, ints(n+1)) I think the Haskell solution is more compact due to syntactic sugar, curring and "parentheses-free-ness", *not* lazy evaluation. According to one guy's analogy: the Real World is strict - in order to drink tea, you have to put the cattle on the fire, wait until water boils, brew tea and then drink. Not the cattle is put on the fire, water boils and the tea is brewed when you take the empty cup to start drinking. :-) The question is the following: how big the gap between strict languages with lazy constructs and Haskell? Does the default lazyness have irrefutable advantage over default strictness? Thank you for the attention. With best regards, Nick.

12 23

ANNOUNCE: Derangement version 0.1.0
by dgriffi3＠uiuc.edu 19 Feb '07

19 Feb '07

This is the inital version of a very small library, mostly using the functional graph library, to find a derangement of a set, or rather a list. A derangement of a set is a permutation with no fixed points, like many constrained matching problems it is susceptible to solution via a Max-flow algorithm. The source and its darcs repository is available at http://www.acm.uiuc.edu/~dgriffi3/projects/derangement/ Of course questions, comments and patches are welcome. -- Dennis Griffith, ACM@UIUC Treasurer

1 0

Recursion in Haskell
by P. R. Stanley 19 Feb '07

19 Feb '07

Hi I understand the basic principle of recursion but have difficulty with the following: -- a recursive function -- for calculating the length of lists myLen [] = 0 myLen (x:xs) = 1 + myLen xs What's happening here? Top marks for a comprehensive jargon-free explanation. Thanks in advance Paul

9 15

Why do I have to specify (Monad m) here again?
by Marc Weber 19 Feb '07

19 Feb '07

Why do I have to specify (Monad m) here again? class (Monad m) => GetMV m a where ... instance GetMV m c where ... No instance for (Monad m) arising from the superclasses of an instance declaration possible fix: add (Monad m) to the instance declaration superclass context Marc

7 21

Lambada and connecting Haskell to a Weblogic server
by Joel Reymont 19 Feb '07

19 Feb '07

Folks, Where can I find Lambada these days and would it be of any use to me in trying to connect to a Weblogic server? To make the long story short, my broker's Java software connects to a remote Weblogic server and I would like to do the same. I suppose this would require me to implement Java RMI in Haskell which is non- trivial. I wonder if it would be easier for me to implement a Java server that talks to Weblogic and then connect to that server. What do you think? Thanks, Joel -- http://wagerlabs.com/

6 7

stateful walk through a tree?
by David Tolpin 19 Feb '07

19 Feb '07

Hi, I am looking for help in design of a stateful tree walker. I have a tree-like data structure (an AST for a small language) for which there is a general stateful walk procedure. That is, each node type is an instance of a walker class, and the walk function is monadic: class Walker a where walk :: a -> State Context b The context is used to store names in the scope, for example. Now, I'd like to use Walker as a general class for implementing several different transformations on the tree (like cross-referencing, code emission, tree visualisation). Those transformations will need expanded state. What is the proper design for that? In Common Lisp, for example, I would Context to contain an opaque slot (visitor), and would assign additional state information to it; a would also define a generic function that would dispatch on the visitor (and probably on walker if I end up having more than one walker). How would I do that in Haskell? I'd like to avoid using mutable variables for now (mostly for didactic puproses). David

2 2

A new code search engine
by Stephane Bortzmeyer 19 Feb '07

19 Feb '07

http://www.krugle.com/ Unlike Google, you can specify Haskell as a language. But the very useful (and, again, unlike Google) parsing service, which allows you to specify that you search "foldl" only in function calls, not function definitions, seems broken for Haskell ("map" is never found in function calls...)

4 3

GHC throws IOError on Win32 when there is no console
by John Ky 19 Feb '07

19 Feb '07

Hi, I noticed on Windows that when I use IO functions that write to stdout when the process is lacking a console, those functions throw an IOError. I'm not sure if this also occurs for stderr because I haven't tried it. Some classes of processes are created without a console because they never interact with the user and include System services. Crashing with IOError in this case is difficult to diagnose because because the only symptom is the process crashes with no visible output. I believe the most sensible behaviour should be for those functions to not throw, but instead do nothing. -John

7 13