June 2011 - Beginners

Example needed
by Zhi-Qiang Lei 15 Jun '11

15 Jun '11

Dear all, I wonder why haskell do not take foldl' as default foldl if foldl is not used in production often. I was told that because foldl is lazy evaluated. I know what the advantage of lazy evaluation: do not evaluate the expression which is not needed to, and can work on infinite list. Can you give me a example about how foldl can work on an infinite list? Thanks. Best regards, Zhi-Qiang Lei zhiqiang.lei(a)gmail.com

4 3

GHCI: C Library Load Path
by Christopher Howard 15 Jun '11

15 Jun '11

Is it possible to adjust from where GHCI loads C libraries? And the order of where it looks? I have this weird situation where I need GHCI to prefer one directory of .so files over another. Specifically, I need this in interactive mode (GHCI), not regular compiling (GHC). Also, I'm not writing code that uses C libraries, just using a module that loads one, so I'm not knowledgeable in the former subject. I could go into more details as to why I need this, but it would probably make this e-mail about four times longer, and I don't want to bore anyone unnecessarily. -- frigidcode.com theologia.indicium.us

1 0

Profiling install error
by blackcat＠pro-ns.net 14 Jun '11

14 Jun '11

I'm trying to install the profiling version of the base libraries on Fedora 15. I set the profiling options in ~/.cabal/config then tried to reinstall but I get %cabal install base --reinstall Resolving dependencies... cabal: internal error: impossible Is there anything I can do now?

3 3

[Haskell-Beginners] Just
by John Sampson 14 Jun '11

14 Jun '11

What does the word "Just" mean in Haskell code? What is its purpose? Regards _John Sampson_

7 7

Rechnung
by Ertugrul Soeylemez 14 Jun '11

14 Jun '11

Und nochmal. -- nightmare = unsafePerformIO (getWrongWife >>= sex) http://ertes.de/

1 1

(no subject)
by Fernando Henrique Sanches 13 Jun '11

13 Jun '11

http://maipenarai.com/lindex02.html

1 2

Does Haskell have a function that creates a list corresponding to the nodes in a tree, from the leaf node to the root node?
by Costello, Roger L. 13 Jun '11

13 Jun '11

Hi Folks, I am implementing the following function in another language. I want to see if Haskell already has a function that does this; if so, I want to use the Haskell function name rather than inventing my own. Here is the function that I am implementing: consider a tree data structure. I pass to the function one of its leaf nodes. The function returns a list containing that leaf node as the first element of the list, the leaf node's parent as the second element of the list, ..., the tree's root node as the n'th element of the list. (The list consists of all the nodes from the leaf node to the root node) Does Haskell have a function that does this? If not, and you were to define a function for this, what name would you give to the function? /Roger

7 6

Graph algorithms
by . 13 Jun '11

13 Jun '11

Hello Haskellers, I would like to implement a graph based algorithm, the sum-product algorithm, in Haskell. I have looked at the FGL package, but I am not sure whether this is really what I need. For example, I need to know whether I am at a leaf (assuming the first version I implement is the "original" algorithm which works for trees). I would also need to store numerical values as well as functions at nodes and edges. The FGL provides for "labels" at nodes and edges, but these appear to have to be instances of classes such as Eq or Ord, which does not work well for functions. I could of course store node values and edge functions externally, e.g. in a Map. All this seems a bit awkward, but I may still just not have enough experience in Haskell. Can anyone give me a hint, or is there maybe even already a factor graph implementation out there which I have missed? Thanks, Christian

2 1

Memory usage patterns for server-like processes
by Sean Charles 13 Jun '11

13 Jun '11

Does anybody know of any reading material that I can ingest that will give me a good feel for how Haskell memory allocation may (or may not) spike at run-time. I want to write a server process (yes, another web server) as part of my continued learning and I just wondered how lazy evaluation might affect a long running server process under a heavy "slashdot" type of load. I know there are lazy and strict alternatives to a lot of functions, I understand 9at least in principle) that you can use strict forms in Data.Map et al. to ensure that the contents aren't thunked, and I know that you can use $! to enforce a return result to be evaluated rather than thunked but are there still issues to be wary of or am I needlessly scaring myself? Has anybody got any long-term use cases with Yesod, HappStack, Snapp and such like. It would be interesting to know! :) Thanks, Sean Charles.

1 0

[Updated] Summary of Chapter 1 of Pearls of Functional Algorithm Design
by Costello, Roger L. 12 Jun '11

12 Jun '11

Hi Folks, I completed my Powerpoint slides which summarize the material in chapter 1 of "Pearls of Functional Algorithm Design": http://www.xfront.com/Pearls-of-Functional-Algorithm-Design/Chapter1/Find-t… /Roger

4 3