Re: [Haskell-cafe] Probably type checker error.
Not just a proposal any more. :-) GHC 7.0 does not generalize local let bindings in some situations. See here for information: http://hackage.haskell.org/trac/ghc/blog/LetGeneralisationInGhc7 There is a proposal (from Big Simon) to remove let-generalization: http://research.microsoft.com/en-us/um/people/simonpj/papers/constraints/let-gen.pdf On 19 Jun 2011, at 18:26, Serguey Zefirov wrote: Right now I write a quite heavy transformation of Haskell source code and found some strange behaviour of typechecker. Some prerequisites: -- dummy class. My own class is much bigger, but I -- could reproduce that behaviour with that class. class ToWires a -- a type with phantom type arguments. data E ins outs = E -- a function that relates E and its inputs. projectInsType :: E ins outs - ins projectInsType = error projectInsType gets evaluated. -- register function. register :: a - a - a register def a = def -- a simple addition. add2 :: (ToWires a, Num a) = (a,a) - a add2 (a,b) = a+b First I have a function: func :: (ToWires a, Num a) = Maybe a - a func mbA = currentSum where x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum It typechecks and works just fine after some transformation. The transformation I work on transform code into something like that: func_E :: (ToWires a, Num a) = E (Maybe a) a func_E = r where r = E -- here we relate mbA and r. mbA = projectInsType r x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum Note the absence of input of func in transformed func_E. I relate mbA with it's proper type using binding mbA = projectInsType r. Then suddently ghc loses all of the context associated with mbA. And find type error at the calling of add2. If I drop ToWires from contexts of func_E and add2 types, all works fine. If I change add2 to simple addition (x + currentSum), all works fine. Full source code is in attachment. I found it using ghc 6.12.1. I asked colleagues, they told me that the same error manifests itself in ghc 7.0.3. Should I fill a bug report or maybe I misunderstood something? a.hs___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Conditional IO ?
Hi, What is right way to do conditional IO? For example, I need to write to file errors only in case they exist, otherwise my function should do nothing: handleParseErrors errors | (not . null) errors = writeFile parse-errors.txt (show errors) | otherwise = ? What should be an 'otherwise' case ? Thanks! ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Conditional IO ?
| otherwise = return () or: handleParseErrors errors = when (not . null $ errors) $ writeFile . 2011/6/20 Dmitri O.Kondratiev doko...@gmail.com: Hi, What is right way to do conditional IO? For example, I need to write to file errors only in case they exist, otherwise my function should do nothing: handleParseErrors errors | (not . null) errors = writeFile parse-errors.txt (show errors) | otherwise = ? What should be an 'otherwise' case ? Thanks! ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Eugene Kirpichov Principal Engineer, Mirantis Inc. http://www.mirantis.com/ Editor, http://fprog.ru/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Conditional IO ?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 20/06/11 18:00, Dmitri O.Kondratiev wrote: Hi, What is right way to do conditional IO? For example, I need to write to file errors only in case they exist, otherwise my function should do nothing: handleParseErrors errors | (not . null) errors = writeFile parse-errors.txt (show errors) | otherwise = ? What should be an 'otherwise' case ? How about `return ()'? -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQIcBAEBAgAGBQJN/v/hAAoJEDiWqExGnQ/Q7VQP/js4XZc92bHeM8f/ErQeQ4LT UwINtt8WFtTWevQtwfO6wNc96CVcjnybb2SwPqnQMISnb4EPc9RvfFz24FTBneQ9 d7MD44uNnW1wAgY8VKaQQJ2v5VVrUd8Gq/m0toUKuefaT9yKMVIG/EeEPaLEUe3y tJUyykCXcUpTzllDyk9kjb/1Og9CNWD+y4rdrWSldWCwgIPMHE2nwNOvs6RquPDQ Om5eFdk2AFwAgDl1FIR2EiuWdQ3jUVPaDsHC0L24fvCeC/TA6AoHmC0panCYCchc 8nEruqwmiA/nU+Uh1Z6SWEicBeuWcXSk6iPuRYM7xU9dDLKFNEHAvCTLYdg2+zwR sovOOYrOEnQ8FDwyZGXC7atxrNGl/nk09Q2RNkXryGo/nNdkUG0AuvoXI5gn8Q6v 7mZiXJr2AbUFqcDA4qXOj8IwddtU5hvwqkGj7xSqCck9sHWWXjZLrJqFH+mW8oMU 8rHVkSpMEFH0vbfvvgRMqzdcpnuDKa4/t+o8wvJy/lr9pdhP2o2azpxvb6NFlDpP dgDwnOe/v8c9Mz7tqs41k94SYTVk2lNpsoXShVoB9tg0EttHKjDNM2ykGJl90Hcf 6vJUGEyNqE78s6qMxiAauxatlQdkw40j2ifcriElj5U7cogpZUGxnb/RODsl0rqn BHPJVkHC/iEQ+t/4lfaT =HGbC -END PGP SIGNATURE- ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Conditional IO ?
Your errors branch has the type writeFile parse-errors.txt (show errors) :: IO () This means that your otherwise branch should have the same type. You can use the return function that has the type return :: Monad m = a - m a specialised to m = IO in conjunction with the value () :: () giving return () :: IO () There is also the when function that eliminates the else case for conditional IO: http://hackage.haskell.org/packages/archive/base/latest/doc/html/Control-Monad.html#v:when On Mon, Jun 20, 2011 at 4:00 PM, Dmitri O.Kondratiev doko...@gmail.com wrote: Hi, What is right way to do conditional IO? For example, I need to write to file errors only in case they exist, otherwise my function should do nothing: handleParseErrors errors | (not . null) errors = writeFile parse-errors.txt (show errors) | otherwise = ? What should be an 'otherwise' case ? Thanks! ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Conditional IO ?
Thanks! Everything works, and 'when' is really nice. ( I still have only basic monad knowledge, need more time to spend on existing libraries) On Mon, Jun 20, 2011 at 12:14 PM, Lyndon Maydwell maydw...@gmail.comwrote: Your errors branch has the type writeFile parse-errors.txt (show errors) :: IO () This means that your otherwise branch should have the same type. You can use the return function that has the type return :: Monad m = a - m a specialised to m = IO in conjunction with the value () :: () giving return () :: IO () There is also the when function that eliminates the else case for conditional IO: http://hackage.haskell.org/packages/archive/base/latest/doc/html/Control-Monad.html#v:when On Mon, Jun 20, 2011 at 4:00 PM, Dmitri O.Kondratiev doko...@gmail.com wrote: Hi, What is right way to do conditional IO? For example, I need to write to file errors only in case they exist, otherwise my function should do nothing: handleParseErrors errors | (not . null) errors = writeFile parse-errors.txt (show errors) | otherwise = ? What should be an 'otherwise' case ? Thanks! ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Probably type checker error.
Yeah, seems to work too. Отправлено с iPhone Jun 20, 2011, в 10:55, Corey O'Connor coreyocon...@gmail.com написал(а): Not just a proposal any more. :-) GHC 7.0 does not generalize local let bindings in some situations. See here for information: http://hackage.haskell.org/trac/ghc/blog/LetGeneralisationInGhc7 There is a proposal (from Big Simon) to remove let-generalization: http://research.microsoft.com/en-us/um/people/simonpj/papers/constraints/let-gen.pdf On 19 Jun 2011, at 18:26, Serguey Zefirov wrote: Right now I write a quite heavy transformation of Haskell source code and found some strange behaviour of typechecker. Some prerequisites: -- dummy class. My own class is much bigger, but I -- could reproduce that behaviour with that class. class ToWires a -- a type with phantom type arguments. data E ins outs = E -- a function that relates E and its inputs. projectInsType :: E ins outs - ins projectInsType = error projectInsType gets evaluated. -- register function. register :: a - a - a register def a = def -- a simple addition. add2 :: (ToWires a, Num a) = (a,a) - a add2 (a,b) = a+b First I have a function: func :: (ToWires a, Num a) = Maybe a - a func mbA = currentSum where x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum It typechecks and works just fine after some transformation. The transformation I work on transform code into something like that: func_E :: (ToWires a, Num a) = E (Maybe a) a func_E = r where r = E -- here we relate mbA and r. mbA = projectInsType r x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum Note the absence of input of func in transformed func_E. I relate mbA with it's proper type using binding mbA = projectInsType r. Then suddently ghc loses all of the context associated with mbA. And find type error at the calling of add2. If I drop ToWires from contexts of func_E and add2 types, all works fine. If I change add2 to simple addition (x + currentSum), all works fine. Full source code is in attachment. I found it using ghc 6.12.1. I asked colleagues, they told me that the same error manifests itself in ghc 7.0.3. Should I fill a bug report or maybe I misunderstood something? a.hs___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Probably type checker error.
The fact is that (Num a) context works and (ToWires a, Num a) context doesn't. At least in 6.12.1. This still looks to me like a bug. 2011/6/19 Miguel Mitrofanov miguelim...@yandex.ru: Seems like let-generalization is at work here. Types of all values in the where section are inferred basically as if they are declared at the top level. Therefore, inheritance fails without NoMonomorphismRestriction. There is a proposal (from Big Simon) to remove let-generalization: http://research.microsoft.com/en-us/um/people/simonpj/papers/constraints/let-gen.pdf On 19 Jun 2011, at 18:26, Serguey Zefirov wrote: Right now I write a quite heavy transformation of Haskell source code and found some strange behaviour of typechecker. Some prerequisites: -- dummy class. My own class is much bigger, but I -- could reproduce that behaviour with that class. class ToWires a -- a type with phantom type arguments. data E ins outs = E -- a function that relates E and its inputs. projectInsType :: E ins outs - ins projectInsType = error projectInsType gets evaluated. -- register function. register :: a - a - a register def a = def -- a simple addition. add2 :: (ToWires a, Num a) = (a,a) - a add2 (a,b) = a+b First I have a function: func :: (ToWires a, Num a) = Maybe a - a func mbA = currentSum where x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum It typechecks and works just fine after some transformation. The transformation I work on transform code into something like that: func_E :: (ToWires a, Num a) = E (Maybe a) a func_E = r where r = E -- here we relate mbA and r. mbA = projectInsType r x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum Note the absence of input of func in transformed func_E. I relate mbA with it's proper type using binding mbA = projectInsType r. Then suddently ghc loses all of the context associated with mbA. And find type error at the calling of add2. If I drop ToWires from contexts of func_E and add2 types, all works fine. If I change add2 to simple addition (x + currentSum), all works fine. Full source code is in attachment. I found it using ghc 6.12.1. I asked colleagues, they told me that the same error manifests itself in ghc 7.0.3. Should I fill a bug report or maybe I misunderstood something? a.hs___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Data Flow Programming in FP
Hi all, I have recently become interested in Dataflow programming and how it related to functional languages. I am wondering if the community has any advice on reading matter or other directions to look at. So far I have been looking through the FRP libraries, using Haskell functions with lazy lists for co-routines and the Essence of Dataflow Programming by Uustalu and Vene where they propose using co-monads. It looks as though Iteratees are also relevant but I have not got round to looking at them in detail yet. Have I missed anything? Regards RS ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Text.CSV questions
Maybe you can directly distinguish if the parser returned an error (Left) or not (Right), instead of using lefts and rights: import Text.CSV import Data.Either import System import Data.List main = do [inpFileName] - getArgs putStrLn (Parsing ++inpFileName++...) result - parseCSVFromFile inpFileName case result of Left error - do print Parse error: print error Right csv - print csvL putStrLn All done. 2011/6/17 Dmitri O.Kondratiev doko...@gmail.com On Fri, Jun 17, 2011 at 1:04 PM, Vincent Hanquez t...@snarc.org wrote: On 06/17/2011 10:00 AM, Dmitri O.Kondratiev wrote: Hi, I try to parse csv file with Text.CSV, like this: import Text.CSV import System main = do [inpFileName] - getArgs putStrLn (Parsing ++inpFileName++...) let result = parseCSVFromFile inpFileName print result === As a result I get: No instance for (Show (IO (Either Text.Parsec.Error.ParseError CSV))) arising from a use of `print' Possible fix: add an instance declaration for (Show (IO (Either Text.Parsec.Error.ParseError CSV))) === Question: How to add a Show instance for (IO (Either Text.Parsec.Error.ParseError CSV))) ? Hi Dmitri, you don't add a show instance for IO, but you unwrap the IO Monad first and then show the result. let result = parseCSVFromFile inpFileName should be: result - parseCSVFromFile inpFileName -- Vincent Thanks everybody! This is how I solved it: import Text.CSV import Data.Either import System import Data.List main = do [inpFileName] - getArgs putStrLn (Parsing ++inpFileName++...) result - parseCSVFromFile inpFileName let errors = lefts [result] let csvL = rights [result] print CSV list print csvL print Errors print errors putStrLn All done. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Data Flow Programming in FP
On Mon, Jun 20, 2011 at 7:45 AM, Richard Senington sc06...@leeds.ac.ukwrote: I have recently become interested in Dataflow programming and how it related to functional languages. I am wondering if the community has any advice on reading matter or other directions to look at. So far I have been looking through the FRP libraries, using Haskell functions with lazy lists for co-routines and the Essence of Dataflow Programming by Uustalu and Vene where they propose using co-monads. It looks as though Iteratees are also relevant but I have not got round to looking at them in detail yet. Have I missed anything? Arrows are a useful model for dataflow programming. But several FRP models are arrowized, so you might already have observed this. Which FRP models have you looked at? (there are several) I'm developing a model for reactive dataflows in open distributed systems, called reactive demand programming (RDP). It's basically distributed FRP with carefully constrained side-effects and signals that model disruption. The effects model enforces spatial idempotence and commutativity, which allows developers to perform refactoring and abstraction similar to that in a pure functional model. That signals model disruption allows 'open' composition and extension (e.g. runtime plugins). RDP is more composable than FRP because client-server relationships can be captured as regular RDP behaviors. RDP isn't ready for release, yet, but you can read a bit more at my blog: [1] http://awelonblue.wordpress.com/2011/05/21/comparing-frp-to-rdp/ Regards, David Barbour ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Text.CSV questions
On Mon, Jun 20, 2011 at 7:30 PM, Alejandro Serrano Mena trup...@gmail.comwrote: Maybe you can directly distinguish if the parser returned an error (Left) or not (Right), instead of using lefts and rights: import Text.CSV import Data.Either import System import Data.List main = do [inpFileName] - getArgs putStrLn (Parsing ++inpFileName++...) result - parseCSVFromFile inpFileName case result of Left error - do print Parse error: print error Right csv - print csvL putStrLn All done. Yes, you right, this is better, then printing empty error list. 2011/6/17 Dmitri O.Kondratiev doko...@gmail.com On Fri, Jun 17, 2011 at 1:04 PM, Vincent Hanquez t...@snarc.org wrote: On 06/17/2011 10:00 AM, Dmitri O.Kondratiev wrote: Hi, I try to parse csv file with Text.CSV, like this: import Text.CSV import System main = do [inpFileName] - getArgs putStrLn (Parsing ++inpFileName++...) let result = parseCSVFromFile inpFileName print result === As a result I get: No instance for (Show (IO (Either Text.Parsec.Error.ParseError CSV))) arising from a use of `print' Possible fix: add an instance declaration for (Show (IO (Either Text.Parsec.Error.ParseError CSV))) === Question: How to add a Show instance for (IO (Either Text.Parsec.Error.ParseError CSV))) ? Hi Dmitri, you don't add a show instance for IO, but you unwrap the IO Monad first and then show the result. let result = parseCSVFromFile inpFileName should be: result - parseCSVFromFile inpFileName -- Vincent Thanks everybody! This is how I solved it: import Text.CSV import Data.Either import System import Data.List main = do [inpFileName] - getArgs putStrLn (Parsing ++inpFileName++...) result - parseCSVFromFile inpFileName let errors = lefts [result] let csvL = rights [result] print CSV list print csvL print Errors print errors putStrLn All done. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Probably type checker error.
GHC 6.12 introduces MonoLocalBinds, which disables polymorphic values in let statements. Your original code works for me if I use -XNoMonoLocalBinds -XNoMonomorphismRestriction. On Mon, Jun 20, 2011 at 9:02 AM, Serguey Zefirov sergu...@gmail.com wrote: The fact is that (Num a) context works and (ToWires a, Num a) context doesn't. At least in 6.12.1. This still looks to me like a bug. 2011/6/19 Miguel Mitrofanov miguelim...@yandex.ru: Seems like let-generalization is at work here. Types of all values in the where section are inferred basically as if they are declared at the top level. Therefore, inheritance fails without NoMonomorphismRestriction. There is a proposal (from Big Simon) to remove let-generalization: http://research.microsoft.com/en-us/um/people/simonpj/papers/constraints/let-gen.pdf On 19 Jun 2011, at 18:26, Serguey Zefirov wrote: Right now I write a quite heavy transformation of Haskell source code and found some strange behaviour of typechecker. Some prerequisites: -- dummy class. My own class is much bigger, but I -- could reproduce that behaviour with that class. class ToWires a -- a type with phantom type arguments. data E ins outs = E -- a function that relates E and its inputs. projectInsType :: E ins outs - ins projectInsType = error projectInsType gets evaluated. -- register function. register :: a - a - a register def a = def -- a simple addition. add2 :: (ToWires a, Num a) = (a,a) - a add2 (a,b) = a+b First I have a function: func :: (ToWires a, Num a) = Maybe a - a func mbA = currentSum where x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum It typechecks and works just fine after some transformation. The transformation I work on transform code into something like that: func_E :: (ToWires a, Num a) = E (Maybe a) a func_E = r where r = E -- here we relate mbA and r. mbA = projectInsType r x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum Note the absence of input of func in transformed func_E. I relate mbA with it's proper type using binding mbA = projectInsType r. Then suddently ghc loses all of the context associated with mbA. And find type error at the calling of add2. If I drop ToWires from contexts of func_E and add2 types, all works fine. If I change add2 to simple addition (x + currentSum), all works fine. Full source code is in attachment. I found it using ghc 6.12.1. I asked colleagues, they told me that the same error manifests itself in ghc 7.0.3. Should I fill a bug report or maybe I misunderstood something? a.hs___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Dave Menendez d...@zednenem.com http://www.eyrie.org/~zednenem/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Software patents covered in GHC?
Hi, I would like to know if there are any software algorithms used in the Glasgow Haskell Compiler that have been patented? If yes, in which countries do they apply? Just curious to know. SK -- Shakthi Kannan http://www.shakthimaan.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Software patents covered in GHC?
*sigh* CC'ing to the rest of haskell-cafe for completeness. I need to change 'reply all' to a default in my email I guess. On Mon, Jun 20, 2011 at 12:19 PM, austin seipp a...@hacks.yi.org wrote: Hello, Realistically, there probably is. Considering everything down to linked lists are patented by the USPO, I'd be willing to bet (at least a dollar or two) that there are many patents out there that apply to the code used in GHC. I don't know if you could say anything different about *any* nontrivial software, to be quite honest. IANAL however, so I won't attempt to claim in what countries these patents apply, nor could I cite you patent numbers (legalese being terse and software patents so seemingly open to interpretation sometimes, it would be hard to narrow down anyway I think.) On Mon, Jun 20, 2011 at 12:10 PM, Shakthi Kannan shakthim...@gmail.com wrote: Hi, I would like to know if there are any software algorithms used in the Glasgow Haskell Compiler that have been patented? If yes, in which countries do they apply? Just curious to know. SK -- Shakthi Kannan http://www.shakthimaan.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Regards, Austin -- Regards, Austin ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Conditional IO ?
On Mon, 20 Jun 2011 10:33:14 +0200, Dmitri O.Kondratiev doko...@gmail.com wrote: Thanks! Everything works, and 'when' is really nice. ( I still have only basic monad knowledge, need more time to spend on existing libraries) See A tour of the Haskell Monad functions[0]. Regards, Henk-Jan van Tuyl [0] http://members.chello.nl/hjgtuyl/tourdemonad.html -- http://Van.Tuyl.eu/ http://members.chello.nl/hjgtuyl/tourdemonad.html -- ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Text report tools?
Hi, I am looking for an easy way to generate text reports. For starters I need a very simple report that may contain: - Some headers - Lists of text strings where each string can include instances of basic Haskell types. Each string should be printed on a separate line (terminated with LF). Do I understand this right that for such simple reports (as above) Text.Printf is the main tool? As a next step, a simple report engine controlled by templates or DSL would help. Thanks! ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Text report tools?
Try http://hackage.haskell.org/package/HStringTemplate. G On Mon, Jun 20, 2011 at 2:27 PM, Dmitri O.Kondratiev doko...@gmail.com wrote: Hi, I am looking for an easy way to generate text reports. For starters I need a very simple report that may contain: - Some headers - Lists of text strings where each string can include instances of basic Haskell types. Each string should be printed on a separate line (terminated with LF). Do I understand this right that for such simple reports (as above) Text.Printf is the main tool? As a next step, a simple report engine controlled by templates or DSL would help. Thanks! ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Gregory Collins g...@gregorycollins.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Conditional IO ?
On Mon, Jun 20, 2011 at 1:33 AM, Dmitri O.Kondratiev doko...@gmail.comwrote:
Thanks! Everything works, and 'when' is really nice.
( I still have only basic monad knowledge, need more time to spend on
existing libraries)
I heavily use 'when' and 'unless'. ('unless b x' is the same as 'when (not
b) x'.)
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] ANNOUNCEMENT: TinyLaunchbury 1.0.1
Sankel Software is pleased to announce the release of TinyLaunchbury 1.0.1 [1], a small implementation of John Launchbury's natural semantics for call-by-need. We hope that is is useful in helping to understand how call-by-need works. There is also a new blog post[2], explaining the semantics, the code and why the semantics are useful to understand. Elliot Stern Sankel Software [1] http://hackage.haskell.org/package/TinyLaunchbury [2] http://netsuperbrain.com/blog/posts/haskells-evaluation-isnt-magic/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Probably type checker error.
Thank you very much. I'll try that too. 2011/6/20 David Menendez d...@zednenem.com: GHC 6.12 introduces MonoLocalBinds, which disables polymorphic values in let statements. Your original code works for me if I use -XNoMonoLocalBinds -XNoMonomorphismRestriction. On Mon, Jun 20, 2011 at 9:02 AM, Serguey Zefirov sergu...@gmail.com wrote: The fact is that (Num a) context works and (ToWires a, Num a) context doesn't. At least in 6.12.1. This still looks to me like a bug. 2011/6/19 Miguel Mitrofanov miguelim...@yandex.ru: Seems like let-generalization is at work here. Types of all values in the where section are inferred basically as if they are declared at the top level. Therefore, inheritance fails without NoMonomorphismRestriction. There is a proposal (from Big Simon) to remove let-generalization: http://research.microsoft.com/en-us/um/people/simonpj/papers/constraints/let-gen.pdf On 19 Jun 2011, at 18:26, Serguey Zefirov wrote: Right now I write a quite heavy transformation of Haskell source code and found some strange behaviour of typechecker. Some prerequisites: -- dummy class. My own class is much bigger, but I -- could reproduce that behaviour with that class. class ToWires a -- a type with phantom type arguments. data E ins outs = E -- a function that relates E and its inputs. projectInsType :: E ins outs - ins projectInsType = error projectInsType gets evaluated. -- register function. register :: a - a - a register def a = def -- a simple addition. add2 :: (ToWires a, Num a) = (a,a) - a add2 (a,b) = a+b First I have a function: func :: (ToWires a, Num a) = Maybe a - a func mbA = currentSum where x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum It typechecks and works just fine after some transformation. The transformation I work on transform code into something like that: func_E :: (ToWires a, Num a) = E (Maybe a) a func_E = r where r = E -- here we relate mbA and r. mbA = projectInsType r x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum Note the absence of input of func in transformed func_E. I relate mbA with it's proper type using binding mbA = projectInsType r. Then suddently ghc loses all of the context associated with mbA. And find type error at the calling of add2. If I drop ToWires from contexts of func_E and add2 types, all works fine. If I change add2 to simple addition (x + currentSum), all works fine. Full source code is in attachment. I found it using ghc 6.12.1. I asked colleagues, they told me that the same error manifests itself in ghc 7.0.3. Should I fill a bug report or maybe I misunderstood something? a.hs___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Dave Menendez d...@zednenem.com http://www.eyrie.org/~zednenem/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Probably type checker error.
Defaulting. Inferred type of r is still polymorphic – r :: E ins outs – but type of x, nextSum and currentSum is defaulted to Integer. Try adding default () to your program, and you'll see the same error again. Or, if you enable NoMonomorphismRestriction (or MonoLocalBinds – I didn't know it works in 6.12 also), you'll see that the type of nextSum etc. is derived as Num a = a. On 20 Jun 2011, at 17:02, Serguey Zefirov wrote: The fact is that (Num a) context works and (ToWires a, Num a) context doesn't. At least in 6.12.1. This still looks to me like a bug. 2011/6/19 Miguel Mitrofanov miguelim...@yandex.ru: Seems like let-generalization is at work here. Types of all values in the where section are inferred basically as if they are declared at the top level. Therefore, inheritance fails without NoMonomorphismRestriction. There is a proposal (from Big Simon) to remove let-generalization: http://research.microsoft.com/en-us/um/people/simonpj/papers/constraints/let-gen.pdf On 19 Jun 2011, at 18:26, Serguey Zefirov wrote: Right now I write a quite heavy transformation of Haskell source code and found some strange behaviour of typechecker. Some prerequisites: -- dummy class. My own class is much bigger, but I -- could reproduce that behaviour with that class. class ToWires a -- a type with phantom type arguments. data E ins outs = E -- a function that relates E and its inputs. projectInsType :: E ins outs - ins projectInsType = error projectInsType gets evaluated. -- register function. register :: a - a - a register def a = def -- a simple addition. add2 :: (ToWires a, Num a) = (a,a) - a add2 (a,b) = a+b First I have a function: func :: (ToWires a, Num a) = Maybe a - a func mbA = currentSum where x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum It typechecks and works just fine after some transformation. The transformation I work on transform code into something like that: func_E :: (ToWires a, Num a) = E (Maybe a) a func_E = r where r = E -- here we relate mbA and r. mbA = projectInsType r x = case mbA of Just a - a Nothing - 0 nextSum = add2 (x,currentSum) currentSum = register 0 nextSum Note the absence of input of func in transformed func_E. I relate mbA with it's proper type using binding mbA = projectInsType r. Then suddently ghc loses all of the context associated with mbA. And find type error at the calling of add2. If I drop ToWires from contexts of func_E and add2 types, all works fine. If I change add2 to simple addition (x + currentSum), all works fine. Full source code is in attachment. I found it using ghc 6.12.1. I asked colleagues, they told me that the same error manifests itself in ghc 7.0.3. Should I fill a bug report or maybe I misunderstood something? a.hs___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Job position for haskell/clojure dev. in San Dimas, CA
Anyone interested in full time employment working with haskell and clojure in San Dimas, CA (local job only, NO telecommute) please let me know. Regards, Vagif Verdi ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Opaque types vs automated instance deriving
Hi all, I encountered a problem when trying to derive makeBinary ''DiffTime with help of *derive* package. The error was: Not in scope: data constructor `MkDiffTime' Which makes a sense, since it's not exported in Data.Time.Clock. I bypassed the problem (yes, I'm too lazy to write instances by hands) with instance Binary UTCTime where put = putGeneric get = getGeneric But it must be less efficient (and more verbose) than compile-time deriving. If there was such a module like Data.Time.Clock.Internal, I could import it to get hidden constructors (and maybe I'll have to fork the library for this purpose). OTOH, is it possible to change the derive TH function so it can bypass module encapsulation mechanism and access un-exported things? ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Text report tools?
On 21 June 2011 04:27, Dmitri O.Kondratiev doko...@gmail.com wrote: Hi, I am looking for an easy way to generate text reports. For starters I need a very simple report that may contain: - Some headers - Lists of text strings where each string can include instances of basic Haskell types. Each string should be printed on a separate line (terminated with LF). Do I understand this right that for such simple reports (as above) Text.Printf is the main tool? As a next step, a simple report engine controlled by templates or DSL would help. If you are wanting to output to HTML or something, I suggest using pandoc as a library or else one of the HTML templating libraries. If it's more of a specific format, maybe use a pretty-printing library directly? -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] AusHac2011 - The Australasian Haskell Hackathon - July 8-10
It's almost that time of year again! Next month is the second annual AusHac event, at UNSW in Sydney, and you are invited! But if you want to come, we need to hear from you soon, as we need to organise access to the university network for you. If you're an Australian Haskell hacker, enthusiast, newbie or professional, we would love to see you there. Below is the announcement, if you have any questions, please just hit reply, and Ivan or I will get back to you. We would absolutely love to see you there if you can make it, we'd love for this year to be even more awesome than last year. -- Alex Mason -- Attention all Australasian Haskellers! After last year’s fantastic turnout for the first Australian Haskell Hackathon - AusHac 2010 - we’ve decided to organise another. As it’s rather boring to have a Hackathon with only two people, we encourage and welcome anyone interested in Haskell in Australia, New Zealand and surrounding areas (or further afield if you want!) to come and join us. Last year we had more than 25 Haskellers come from Brisbane, Canberra, Melbourne, Sydney and even two from as far away as New Zealand. AusHac 2011 will be held from Friday July 8th until Sunday July 10th, again at UNSW’s Computer Science building in Sydney. There’s no need to come for the entire weekend, just for whenever you’re able to. Last year, people worked on several projects including: • Accelerate: a high performance library for array computations using various backends (CUDA, LLVM, etc.) being written at UNSW. • Leksah: the Haskell IDE written in Haskell • DDC: the compiler for the Haskell-like language Disciple, which has strict evaluation by default, region, effect and closure typing and other interesting language features. • Hubris: the Haskell/Ruby bridge for calling Haskell from Ruby and vice-versa. • haskell-mpi: bindings to MPI, the message passing interface used frequently on high performance super-computers to allow programs to run on many nodes in parallel with easy communication. • There was even some work on LLVM to produce HTML output showing register allocations (as an aid to those writing new code generators etc.), which has now been included upstream. This work did not use Haskell, but it did happen at AusHac 2010. • And much more! This is a great chance to collaborate with others and get feedback on your work, while having heaps of fun and meeting other Haskellers, to verify they do actually exist! And don’t worry if you can’t think of a project to work on; just come along and pitch in with whatever strikes your fancy! If you’re interested in coming and have a project you’d like to work on, check out the wikipage http://www.haskell.org/haskellwiki/AusHac2011 where you can add your project ideas, or just take a look and see what others will be working on that you can help out with. People of all skill levels are welcome, from complete beginners to Oleg-level sage masters. We’re more than happy to help you learn Haskell if you’re just starting out or help you solve that little niggly bit in your code. If you’re interested in coming along, fill out our registration form http://axman6.wufoo.com/forms/aushac-2011-sign-up/. Registration is required so we know numbers and for you to get access to the university Wifi. Even if you’re not sure, fill it in to express your interest, obligation free. We’d rather know we have room for too many people than not enough. So come along! And if for some reason you don’t enjoy yourself, registration comes with a money-back guarantee! Hope to see you there, -- Alex Mason and Ivan Miljenovic, AusHac organisers -- ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Job position for haskell/clojure dev. in San Dimas, CA
Sorry, forgot to mention, and i already got questions about it. No worker visa sponsorship, no relocation from abroad. US only. On Jun 20, 2:39 pm, vagif.ve...@gmail.com wrote: Anyone interested in full time employment working with haskell and clojure in San Dimas, CA (local job only, NO telecommute) please let me know. Regards, Vagif Verdi ___ Haskell-Cafe mailing list Haskell-C...@haskell.orghttp://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Software patents covered in GHC?
austin seipp: *sigh* CC'ing to the rest of haskell-cafe for completeness. I need to change 'reply all' to a default in my email I guess. On Mon, Jun 20, 2011 at 12:19 PM, austin seipp a...@hacks.yi.org wrote: Hello, Realistically, there probably is. Considering everything down to linked lists are patented by the USPO, I'd be willing to bet (at least a dollar or two) that there are many patents out there that apply to the code used in GHC. I don't know if you could say anything different about *any* nontrivial software, to be quite honest. IANAL however, so I won't attempt to claim in what countries these patents apply, nor could I cite you patent numbers (legalese being terse and software patents so seemingly open to interpretation sometimes, it would be hard to narrow down anyway I think.) In fact, you are better of not to know. Given that GHC (like all non-trivial software) surely infringes on some patents, the damages that a patent holder can sue you for are less if you do not know about the patents you are infringing on. IIRC, a plaintiff can triple the charges if they can reasonably show that you have been aware of the patents that you are infringing on. Nasty business! Manuel On Mon, Jun 20, 2011 at 12:10 PM, Shakthi Kannan shakthim...@gmail.com wrote: Hi, I would like to know if there are any software algorithms used in the Glasgow Haskell Compiler that have been patented? If yes, in which countries do they apply? Just curious to know. SK -- Shakthi Kannan http://www.shakthimaan.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Regards, Austin -- Regards, Austin ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Haskell *interpreter* on iPad? (Scheme and Ocaml are there)
And while we are dreaming, in an iOS port of GHCi (meaning GHCi runs on iOS and doesn't just generate code for it), it would be great to make bytecode persistent — ie, the bytecode that GHCi currently generates internally to interpret programs should be serialized to save and load it. (Note that iOS apps are not allowed to generate binary code, so we wouldn't need the gcc tool chain etc, but just have a stripped down GHCi that generates bytecode only. All pre-compiled libraries that are part of the GHCi app can of course still be native code and load and interact with the interpreter as usual.) Manuel Don Stewart: Oh, Scheme is trivial to implement, when compared with Haskell. So people write it from scratch as a tutorial exercise. Haskell isn't trivial to implement from scratch, so instead we port existing implementations mostly. That means really, porting Hugs or GHC. And you've been pointed at examples. I think people are clearly keen for this, now it is a small matter of programming talent and will. -- Don On Sat, Jun 18, 2011 at 3:03 PM, John Velman vel...@cox.net wrote: On Sat, Jun 18, 2011 at 10:44:01PM +0400, MigMit wrote: Well, this is my point. THERE ARE 3 SCHEME INTERPRETERS in the iPad app store. They run on factory iPads, not jailbroken. The GUI for the gambitREPL (Read, Evaluate, Print, Loop) is just like a console. Input a scheme expression. CR. Answer appears, new prompt. In haskell we need to allow for some way to input layout. I don't recall how Hugs handles this, if at all. There are probably 5 or 10 people out there who want to learn functional programming, and they are studying Scheme on their iPads. Or Ocaml. I don't forsee doing production programming ON THE IPAD, but experimenting, testing some functions, and, by the way, learning Haskell. While I'm fantasizing, something like Hugs or ghci with SOE would really be neat. Sorry for shouting :-) John Velman Well, Haskell is fun, isn't it? And that's what iPhone is perfect for: fun. Back when I had iPod Touch 1G (jailbroken, of course), I used to run Hugs on it. Now I would love to see a Haskell interpreter in the App Store -- which, by the way, is possible; as there are Scheme interpreters there, why not Haskell? Отправлено с iPhone Jun 18, 2011, в 22:27, Jack Henahan jhena...@uvm.edu написал(а): I suppose you could make a GUI, by why? Given that you'll have to be working on a jailbroken device, anyway, one could just as well use one of the numerous terminal emulators now floating around for jailbroken iOS. That said, the idea of people writing Haskell on phones and iPads and so on makes me just a little bit grinny. On Jun 18, 2011, at 2:17 PM, Alexander Solla wrote: On Sat, Jun 18, 2011 at 10:46 AM, John Velman vel...@cox.net wrote: To further emphasize, I'd like to type in (or paste in) Haskell code and have it executed on the iPad. To reiterate: Something like Hugs, or ghci on the iPad. Since the iPhone OS is pretty much OS X for ARM, and GHC apparently now supports cross-compilation, you can compile GHC for iOS. I guess you could cross compile Hugs with GCC. Doing so probably isn't trivial, but it should be straightforward. I bet you could even use Xcode to make a graphical user interface to GHCi. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Opaque types vs automated instance deriving
On Mon, Jun 20, 2011 at 5:07 PM, Alexey Karakulov ankaraku...@gmail.com wrote: Hi all, I encountered a problem when trying to derive makeBinary ''DiffTime with help of derive package. The error was: Not in scope: data constructor `MkDiffTime' Which makes a sense, since it's not exported in Data.Time.Clock. I bypassed the problem (yes, I'm too lazy to write instances by hands) with instance Binary UTCTime where put = putGeneric get = getGeneric But it must be less efficient (and more verbose) than compile-time deriving. If there was such a module like Data.Time.Clock.Internal, I could import it to get hidden constructors (and maybe I'll have to fork the library for this purpose). OTOH, is it possible to change the derive TH function so it can bypass module encapsulation mechanism and access un-exported things? I would benchmark putGeneric/getGeneric against: put = put . toRational get = fromRational $ get It looks like that's what the safecopy package on Hackage uses. Antoine ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
