June 2006 - Haskell-Cafe

Haskell I/O inside: down the rabbit's hole
by Bulat Ziganshin 30 Jun '06

30 Jun '06

Hello Haskell, I'm just wrote exhaustive introduction to Haskell I/O titled HASKELL I/O INSIDE: DOWN THE RABBIT'S HOLE Can you please help me (and Haskell community) by proof-reading this text? Even if you will correct paragraph or two, it will advance the work further. Also feel free to add more explanations, correct me and ask (here) about new topics you need to be explained. The page is http://haskell.org/haskellwiki/IO_inside -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

2 2

Re[2]: [Haskell-cafe] Haskell I/O inside: down the rabbit's hole
by Bulat Ziganshin 30 Jun '06

30 Jun '06

Hello Donald, Friday, June 30, 2006, 1:20:03 PM, you wrote: >> http://haskell.org/haskellwiki/IO_inside > Perhaps mention > http://research.microsoft.com/~simonpj/Papers/marktoberdorf > "Tackling the awkward squad: monadic input/output, concurrency, > exceptions, and foreign-language calls in Haskell" > for people wanting to go further into the area? of course. feel free to add anything you think is useful. like 'Arrays' page, it is collective work, and like this page it can go much further from it's initial state. a have added a section, but it needs further editing -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

1 0

Re: Speed of ByteString.Lazy
by Chad Scherrer 30 Jun '06

30 Jun '06

Then I should be thanking Duncan as well (thanks!). I had seen the array fusion idea before in the NDP work, but I hadn't thought of applying to this area. I wonder where else the concept might apply? Is there a typeclass to be built? -- Chad Scherrer "Time flies like an arrow; fruit flies like a banana" -- Groucho Marx Cheers! Duncan Coutts and I have put in many long hours in improving the bytestring code over the last 3 months, in particularly the Lazy bytestring library, and array fusion on bytestrings. You can see here, http://www/~dons/images/commits/fps-commits.png<http://www/%7Edons/images/commits/fps-commits.png>, some 200 patches a month :} Its gratifying to find out that people really use the code :) -- Don -- Chad Scherrer "Time flies like an arrow; fruit flies like a banana" -- Groucho Marx

1 0

A question about stack overflow
by Huazhi (Hank) Gong 29 Jun '06

29 Jun '06

Hi, all I'm just a newbie for Haskell and functional programming world. The idea I currently read is quite different and interesting. I have one general question about the recursively looping style. For example: myMax [ ] = error "empty list" myMax [x] = x myMax [x:xs] = if x>= (myMax xs) then x else (myMax xs) I just list out this kind of very simple program. However, if the list size if a big number such as 10000000, the Winhug will report that the stack is overflow. Does it mean that the functional programming is lacking of scalability? I do know that we can manually change the stack size for it. But that's not a good solution according to my opinion. Yours, Hank

6 6

how to write an haskell binding
by minh thu 28 Jun '06

28 Jun '06

hi, hopefully (well, i liked the discussions) this message will not spawn so much messages :) i'd like to know if there exist some general guidelines/information about writing an haskell (or is it *a* haskell ?) binding for a c or a c++ library. i worry about : * implementation tools : use tools like c2hs or others, or do that bare hands * haskell idioms/usage : make good use of monads, type classes and other haskell related things * for c++, is it better to first write a c api for the c++ code before writing the binding * maybe advice about the c api knowing there will be haskell binding * any thing i forgot thanks a lot, thu

8 14

Deducing Show for GADTs
by Joel Björnson 28 Jun '06

28 Jun '06

Hi. I came a cross the following phenomena which, at least to me, occurs kind of awkward. The code below : data MyData a where DC1 :: (Show a ) => a -> MyData a instance Show (MyData a) where show (DC1 a ) = show a yields the ghci error : 'Could not deduce (Show a) from the context (Show (MyData a))' Adding a Show restriction for the instantiation as in instance Show a => Show (MyData a ) where show (DC1 a ) = show a makes the type checker happy. However, this means that all parametrised values over MyData must have a Show type which isn't necessarily what one wants. I would also like to point out that adding a 'wrapper type' as in data Wrap a = Wrap a data MyData a where DC1 :: (Show a ) => a -> MyData (Wrap a) instance Show (MyData a ) where show (DC1 a ) = show a works fine. Even though 'Wrap' does not derive Show. So, if anyone can give me some hints about the reason for this, I will appreciate it :) Thanks /Joel

4 3

Cabal and linking with static libs (.a files)
by Ketil Malde 28 Jun '06

28 Jun '06

(Reposted to café - my -libraries mail seems to have gotten lost along the way) Hi, I'm building an interface to a C library, which comes in the form of two .a files. I can't seem to get Cabal to link statically with these, so that the resulting package (libHSfoo-v.v.a) is self contained. I've tried fiddling with Cabal's ld-options and extra-libraries, but so far with no success - loading my package either results in GHC looking for (and failing to find) .so files, or "unknown symbol". Help? -k PS: did --user stop working (with GHC 20060607)? I'm adding --user to both 'configure' and 'install', but Cabal still wants to put things in /usr/local. -- If I haven't seen further, it is by standing in the footprints of giants

2 6

Are FunPtr's stable? (was: how to write an haskell binding)
by Brian Hulley 28 Jun '06

28 Jun '06

Bulat Ziganshin wrote: > Hello Brian, > > Tuesday, June 27, 2006, 2:43:15 AM, you wrote: > >> achieve a goal. One other thing to bear in mind is that foreign >> calls are extremely slow, so for example it is much faster to use the >> Foreign.Marshal.Array and Foreign.C.String functions to allocate and >> populate a temporary array with the contents of a list, and send the >> pointer to this array to C with one foreign call, than to send each >> element of the list with multiple foreign calls (eg to paste only >> 19K of text from the Windows clipboard to my app took over 1 >> minute!!! > > he-he-he :) just add "unsafe" specifier: Thanks Bulat! It turns out nearly every foreign function I'm using doesn't make callbacks into Haskell so can be marked "unsafe" so your solution has saved me an enormous amount of refactoring... :-) On a related note, as I was reading the FFI specification again, I can't find any mention of whether or not FunPtr's are stable with respect to garbage collection. I'm assuming they are but am I correct? Thanks, Brian. -- Logic empowers us and Love gives us purpose. Yet still phantoms restless for eras long past, congealed in the present in unthought forms, strive mightily unseen to destroy us. http://www.metamilk.com

3 4

closures with side effects
by tpledger＠ihug.co.nz 28 Jun '06

28 Jun '06

dkarapet wrote: > I have been trying to understand closures > in haskell and how they relate > to side effects. I have been looking > around but all I find are trivial > examples with no side effects. Please let > me know if you know of any examples. The side effects occur in the context that causes the closure to be entered. Here's a nigh-trivial example. myClosure :: IO () myClosure = putStrLn "Hello, world." main = myClosure >> myClosure When myClosure is defined, the side effect doesn't occur yet. We just have a *definition* of an IO action that hasn't yet been bound into the program's sequence of actions. When main binds myClosure into the program's sequence of actions (twice), the side effect occurs (twice). Depending on the implementation of putStrLn, it may be faster the second time because the same closure has been entered before. HTH, Tom

1 0

Re: [Haskell] Google SoC: Software Transactional Memory for Parrot
by John Meacham 28 Jun '06

28 Jun '06

On Tue, Jun 27, 2006 at 08:29:21AM -0700, Jared Updike wrote: > >Am I the only one whose first instinct upon reading this is "EW!"? > > You are not the only one, judging from my own experience. I made my > own sort of algebraic datatypes / abstract datatypes in C# by using > Enums and Objects and runtime casts. It works, but the code itself is > hairy. I guess the good news is that once I built the ADT the---in my > case, ugly---implementation is hidden. When I first started using Haskell Sather was my language of choice. As I learned Haskell and all it's cool features I kept wanting to "backport" them to Sather... Algebraic data types was one of the things I proposed to the sather community, a wacky idea for an OO language of sathers heritage. but it was not long after then that I realized I should just be using Haskell to begin with and redirected my efforts here. :) John -- John Meacham - ⑆repetae.net⑆john⑈

1 0