March 2008 - Haskell-Cafe

Re: [Haskell-cafe] intersection of sets
by Tim Chevalier 18 Mar '08

18 Mar '08

[ccing to list] On 3/18/08, Walt Potter <potter(a)southwestern.edu> wrote: > Tim, > > We're using -O2 to optimize, but not changing the heap size. We're > trying to get the profiling running on our file. > The sets have about 25 million elements each. Each element is a sequence > of length 16. > > We'll want to compute larger examples! > > I do not know how intersection works in Data.Set. I'm worried that we're > having problems with stack overflow due to recursion. We'll try to > remove the lazy stuff. > Run profiling before you "try to remove the lazy stuff". There's no point in trying to optimize code until you know what's actually taking up space. If you have problems with profiling, then ask on the glasgow-haskell-users mailing list. Cheers, Tim -- Tim Chevalier * http://cs.pdx.edu/~tjc * Often in error, never in doubt "Aw, honey, you can keep what's in my pockets, but send me back my pants." --Greg Brown

1 0

intersection of sets
by Walt Potter 18 Mar '08

18 Mar '08

All, We're running out of space when we intersect two large sets. We've imported Data.Set. Any suggestions? Thanks, Walt

2 1

[Haskell-cafe] type families, fun deps, lattices imposed on/by types
by Isto Aho 18 Mar '08

18 Mar '08

Hi all, I was trying to solve a problem and by chance found texts on functional dependencies and then on type families (papers, ghc docs). Please, consider the example 03 of "Understanding functional dependencies via Constraint Handling rules" by Sulzmann, Duck, Peyton-Jones and Stuckey. There we are defining a multiplication over different numeric types: class Mul a b c | a b -> c where (*) :: a -> b -> c instance Mul Int Int Int where ... instance Mul Int Float Float where ... Ok, we could add instance Mul Float Int Float where ... but if we want to make everything work nicely together, Integer, Complex, Rational etc, there will be a lot of instances, especially if we have to give both "Float Int" and "Int Float" instances. And now the question on "lattices of types" (didn't know any better name to refer to the following): Is the following possible? Or can something similar achieved with type families (easily)? type lattice T a b -- Each of the following gives a "<"-relation between types, -- "upper" gives a method, how we get the larger (2nd) from -- the smaller (1st). lattice instance T Int Integer where upper = fromIntegral lattice instance T Int Float where upper = fromIntegral lattice instance T Integer (Ratio Integer) where upper = ... lattice instance T (Ratio Integer) Double where ... lattice instance T Float Double ... lattice instance T Double (Complex Double) ... -- e.g. Now the compiler can check that the top type is -- unique and that there is a path from every type to the top type. -- If the compiler founds this not to be the case, an error is output. -- But otherwise there can be types that are not comparable but -- the common top is quaranteed to exist. class Mul a b c where lattice T (*) :: a -> b -> c (*) x y = (upper x y x) * (upper x y y) -- Now there is no need for the instances like. instance Mul Int Float Float where ... instance Mul Float (Ratio Integer) Double where ... The compiler can generate those instances, because we have given the "upper"-methods. There is always the top available. Function (*) x y = (upper x y x) * (upper x y y) might could be replaced with (*) x y = x * y because of the presence of lattice T and thus, the "upper"-function. If this were possible, the benefits would be: - No need to state "Int Float" and "Float Int" instances on cases where the operands do commute. - No need to write a large number of instances if you have several "types on lattice" (e.g. some more or less well defined relation). - If the relation between types is not clear, we could define another lattice T2 to our Mul2 class for the other uses. Continuing the idea: we could override the default instances generated by the compiler with our own instances when needed. Ok, so, is it possible to write something like that already? (I just wouldn't like write a lot of instances...) If not, would it be possible to extend, say ghc, to work that way or are there too much inconsistencies in the above presentation? (Yes, where is the bottom. Why top and not bottom? Should it be possible to tell, which one to use, if both present? e.g.) Thanks for reading this far! -- br, Isto

2 2

Two Lectureships in Nottingham
by Graham Hutton 18 Mar '08

18 Mar '08

Dear all, We are advertising two new Lectureships (Assistant Professorships) in Computer Science at the University of Nottingham in England. Applications in the area of functional programming are particularly welcome. Further details are available from: http://jobs.nottingham.ac.uk/vacancies.aspx?cat=160#j2797 The closing date for applications is 28th March 2008. Best wishes, Graham Hutton +-----------------------------------------------------------------+ | Dr Graham Hutton Email : gmh(a)cs.nott.ac.uk | | School of Computer Science | | University of Nottingham Web : www.cs.nott.ac.uk/~gmh | | Jubilee Campus, Wollaton Road | | Nottingham NG8 1BB, UK Phone : +44 (0)115 951 4220 | +-----------------------------------------------------------------+ This message has been checked for viruses but the contents of an attachment may still contain software viruses, which could damage your computer system: you are advised to perform your own checks. Email communications with the University of Nottingham may be monitored as permitted by UK legislation.

1 0

Re: ANN: Parsec 3.0.0
by John MacFarlane 18 Mar '08

18 Mar '08

> * A "compatibility" Text.ParserCombinators.Parsec tree for the old > Parsec. It's not perfect, but it should work with most Parsec 2 > code. A data point: I recompiled pandoc with the new Text.ParserCombinators.Parsec compatibility module, and performance is much worse than with parsec 2.1 (approximately twice as slow on my standard benchmark). That aside, this is a very welcome release! John

2 1

Converting image2PS function
by iliali16 18 Mar '08

18 Mar '08

Hi guys again just want to ask you I have this function ready -- Creates a Postscript file from an Image -- page shows -200 <= x <= 200, 300 <= y <= 300 -- boundaries, plus origin plotted on output image2PS:: Image -> String -> IO() image2PS (Image contents) filename = writeFile filename (header++contents++footer) where header = unlines ["%!PS-Adobe-3.1" ,"%%BoundingBox: 100 125 500 725" ,"%%PageOrder: Ascend" ,"%%Pages: 1" ,"%%BeginDocument" ,"%%Page: 1 1" ,"100 125 moveto 500 125 lineto 500 725 lineto 100 725 lineto closepath gsave stroke grestore clip" ,"275 425 moveto 50 0 rlineto stroke" ,"300 400 moveto 0 50 rlineto stroke" ] footer = unlines ["showpage" ,"%%EndDocument" ] but when I use it for example image2PS spiral iliali it tells me that the iliali is undefined. What do you thing how should I use it. Thanks in advance! PS: I need to know how this works so that I make a similar function that converts logo2PS since I have a Logo spiral and I want to see it in PostScript. Thanks again! -- View this message in context: http://www.nabble.com/Converting-image2PS-function-tp16122095p16122095.html Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

3 3

build recent(ish) ghc on macos 10.3.9 powerpc?
by Uwe Hollerbach 18 Mar '08

18 Mar '08

[Augh! gmail won't send! Apologies if this shows up more than once...] Hi, all, I have an old iMac G3, running OSX 10.3.9, to which I have a sentimental attachment. I'd like to get ghc running on it, but the pre-built binaries I can find are all for more-recent iMacs, so I thought I would try to build it myself. I believe I read somewhere that gcc 3.3.X didn't work quite right for recent ghc -- I'm trying for now to build ghc 6.6.1 -- so I started by upgrading gcc to 3.4.6. That's working. So, with that in place, I went to the "porting ghc to a new arch" page and started going through the steps. I'm using a laptop running linux as the "host" computer, so that's i386-unknown-linux, some Fedora core derivative. It's using gcc 3.4.4. I think I did all the initial stuff right, but in the section in the porting guide where it says to start diving into all the subdirectories and do a bunch of "make boot && make", I start getting errors, and by the time it says to do that in .../libraries, it just croaks. Did I miss something earlier? Is this a completely hopeless endeavor from the get-go? I'm not even sure what intermediate files or error messages I should post. Any hints appreciated!. If you want to tell me, "Uwe, you're a bozo", that's fine, too, and you could well be right, just tell me why... Thanks! Uwe

3 3

Small question about something easy
by iliali16 18 Mar '08

18 Mar '08

Hi guys I am a bit new to haskell but I am doing good till now. I have to write a function that takes 2 inputs and then reutns one composite output. Now my problem is that I have to make composition of that function meaning that I have to access in some way the output of the function before it is really computed. I will show you part of my code which is working prefectly: play :: Logo -> TurtleState -> (Image, TurtleState) play DoNothing (pen, (x,y), angle) = (emptyImage, (pen, (x,y), angle)) play PenDown (pen, (x,y), angle) = (emptyImage,(pen, (x,y), angle)) play PenUp (pen, (x,y), angle) = (emptyImage,(pen, (x,y), angle)) play (Forward n) (pen, (x,y), angle) |pen == True = ((line (x,y) (x+n,y+n)), (True, (x+n,y+n), angle)) |otherwise = (emptyImage,(pen, (x+n,y+n), angle)) play (Turn n) (pen, (x,y), angle) = (emptyImage, (pen, (x,y), (angle+n))) play (DoNothing :>: p2) (pen, (x,y), angle) = play p2 (pen, (x,y), angle) play (p1 :> DoNothing) (pen, (x,y), angle) = play p1 (pen, (x,y), angle) play (PenDown :>: PenUp) (pen, (x,y), angle) = (emptyImage,(pen, (x,y), angle)) play (PenDown :>: (Forward n)) (pen, (x,y), angle) = play (Forward n) (True, (x,y), angle) play (PenDown :>: PenDown) (pen, (x,y), angle) = (emptyImage,(pen, (x,y), angle)) play (PenDown :>: (Turn n)) (pen, (x,y), angle) = play (Turn n) (pen, (x,y), angle) play (PenUp :>: PenUp) (pen, (x,y), angle) = (emptyImage,(pen, (x,y), angle)) play (PenUp :>: (Forward n)) (pen, (x,y), angle) = play (Forward n) (False, (x,y), angle) play (PenUp :>: (Turn n)) (pen, (x,y), angle) = play (Turn n) (pen, (x,y), angle) play (PenUp :>: PenDown) (pen, (x,y), angle) = (emptyImage,(pen, (x,y), angle)) Now the problem comes here: play (p1 :>: p2) state |play p1 state == (i1,state1) && play p2 state1 == (i2,state2) = (i1+++i2,state2) I know that if I manage to do that function the one above with this sign :>: do not need to be impelmented since this one will cater for all the cases. Can you please help me? Thanks in advance! -- View this message in context: http://www.nabble.com/Small-question-about-something-easy-tp16119618p161196… Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

3 3

Need Cabal/library building help for windows
by Ryan Ingram 18 Mar '08

18 Mar '08

For reference, I'm using GHC6.8.1 on WinXP. I'm trying to package up some code into a library. But cabal fails to configure my project: >runhaskell setup.hs configure Configuring Prompt-1.0... setup.hs: ld is required but it could not be found. Prompt.cabal contains: name: Prompt version: 1.0 cabal-version: >= 1.2 build-type: Simple library exposed-modules: Control.Monad.Prompt Does GHC require additional tools in order to build libraries on Win32? If so, I hope this can be remedied; it builds executables without additional tools after all. What do I need to do next? -- ryan

4 5

haskell.org is accepted to GSoC
by Malcolm Wallace 18 Mar '08

18 Mar '08

Google announced last night that Haskell.org has been accepted as a mentoring organisation for this year's Summer of Code (amongst 175 Open Source organisations). http://code.google.com/soc/2008/ The game is on! Student applications open at 1900 UTC on Mon 24th March, and close at 2400 UTC on Mon 31st March. In the meantime, my _strong_ suggestion is for students to search for ideas amongst the many lists available on the web: http://hackage.haskell.org/trac/summer-of-code/ http://haskell.org/haskellwiki/Hac_2007_II/Projects http://haskell.org/haskellwiki/Hac4/Projects and when crafting your proposal: * post it to an appropriate mailing list for comments e.g. haskell-cafe, libraries, wxHaskell, cabal-devel (prefix your subject line with [GSoC] to make it easy to find) * use feedback from the list to improve your proposal. Potential mentors will rank your final submitted proposal based on your ability and commitment to engage with the wider Haskell community, in addition to technical merit. Proposals that will benefit the entire Haskell world will generally be ranked more highly than niche projects with only a few potential users. So things like Cabal-hacking, darcs-hacking, compiler-hacking, are good, as are new libraries to do tasks that are in high demand. (This is not an exhaustive list.) Regards, Malcolm

1 0