- Glasgow-haskell-users

RE: :kind feature request in ghci
by Simon Peyton-Jones 05 Apr '04

05 Apr '04

Done! In the HEAD anyway, and will therefore be in 6.4 Simon Prelude> :k Int Int :: * Prelude> :k Maybe Maybe :: * -> * Prelude> :k Maybe Int Maybe Int :: * Prelude> :q | -----Original Message----- | From: glasgow-haskell-users-bounces(a)haskell.org [mailto:glasgow-haskell-users- | bounces(a)haskell.org] On Behalf Of S. Alexander Jacobson | Sent: 27 March 2004 15:33 | To: GHC users | Subject: :kind feature request in ghci | | I keep running into kind errors (see my most | recent post on Haskell cafe). Would it be | possible for ghci to expose a :kind function | that operates on types the way :type operates on | values? | | Alternatively, is there another way to | figure out how to create an instance of | a class like this? | | class MyClass a b where emptyVal::a b | | Here are one of the thing that don't work: | | type MyType a = [a] | instance MyClass (MyType a) a where emptyVal=[] | | And I have no idea why... | | -Alex- | | _________________________________________________________________ | S. Alexander Jacobson mailto:me@alexjacobson.com | tel:917-770-6565 http://alexjacobson.com | _______________________________________________ | Glasgow-haskell-users mailing list | Glasgow-haskell-users(a)haskell.org | http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

1 0

RE: Subject: Generic compilation warnings...
by Simon Peyton-Jones 05 Apr '04

05 Apr '04

It should be ignored, but I've fixed it anyway. Thanks for mentioning it. Simon | -----Original Message----- | From: glasgow-haskell-users-bounces(a)haskell.org [mailto:glasgow-haskell-users- | bounces(a)haskell.org] On Behalf Of MR K P SCHUPKE | Sent: 17 February 2004 15:07 | To: glasgow-haskell-users(a)haskell.org | Subject: Subject: Generic compilation warnings... | | When deriving Data with all warnings enabled, for example compiling the | following line: | | data TestType = TestType Int deriving (Typeable,Data) | | ghc 6.2 gives the following warnings: | | Warning: Pattern match(es) are non-exhaustive | In a case alternative: | Patterns not matched: GHC.Base.I# #x with #x `notElem` [1#] | | Is there a problem with the 'derived' code, or should this warning be | ignored? | | regards, | Keean Schupke. | _______________________________________________ | Glasgow-haskell-users mailing list | Glasgow-haskell-users(a)haskell.org | http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

1 0

RE: [Haskell] Dynamically loading wxhaskell?
by Simon Marlow 03 Apr '04

03 Apr '04

> On Thu, Apr 01, 2004 at 10:00:23AM +0100, Simon Marlow wrote: > > > > > Has anyone succeeded in getting wxhaskell to work under > ghci on Linux? > > > On my system, I get an error message > > > > > > Loading package unix ... ghc-6.2: can't load .so/.DLL for: dl > > > (libdl.so: cannot open shared object file: No such file > or directory) > > > > > > This sounds like it has nothing to do with wxhaskell, but is a GHC > > > problem. Any pointers? > > > > Is libdl.so in the usual place? (/usr/lib on my system) > > Yes. > > (FYI, this is a Debian unstable system.) > > Everything works with GHC, by the way. Very strange. Is /usr/lib/libdl.so perhaps a symlink to a library that doesn't exist? That could happen if an upgrade had gone wrong, perhaps. Cheers, Simon

2 1

FAQ: How to Use Extensible Records in GHCi
by S. Alexander Jacobson 30 Mar '04

30 Mar '04

I feel like I have way too much code of the form: changeGpart (C a b c d e f g h i) g' = C a b c d e f g' h i So, I got excited reading Simon and Mark's "Lightweight Extensible Records For Haskell" paper as a way to get rid of this cruft. I immediately started GHCi, typed {a="Hello world"}, and got an error. Is the feature described in the paper just a proposal or is there a switch somethere to enable it? Alternatively, perhaps this code is a common beginner mistake due to ignorance of the proper idiom... -Alex- _________________________________________________________________ S. Alexander Jacobson mailto:me@alexjacobson.com tel:917-770-6565 http://alexjacobson.com

2 1

RE: How to Use Extensible Records in GHCi
by Simon Peyton-Jones 30 Mar '04

30 Mar '04

No, extensible records are not implemented in GHC. Hugs (or some variant of it) implements TRex, another extensible record variant. The problem is that the record design space is large, and seems to lack local optima. And all reasonable variants break backward compatibility. As a result, nothing much happens. But for the particular code you write, H98 does the job fine: data C = C { a :: Int, b:: Bool, c::Int, ... i::String} Then if there's an expression e::C, you can write e { b=True } to change the b component to True. Check the H98 report SImon | -----Original Message----- | From: glasgow-haskell-users-bounces(a)haskell.org [mailto:glasgow-haskell-users- | bounces(a)haskell.org] On Behalf Of S. Alexander Jacobson | Sent: 30 March 2004 00:14 | To: GHC users | Subject: FAQ: How to Use Extensible Records in GHCi | | I feel like I have way too much code of the form: | | changeGpart (C a b c d e f g h i) g' = | C a b c d e f g' h i | | So, I got excited reading Simon and Mark's | "Lightweight Extensible Records For Haskell" paper | as a way to get rid of this cruft. I immediately | started GHCi, typed {a="Hello world"}, and got an | error. | | Is the feature described in the paper just a | proposal or is there a switch somethere to enable | it? | | Alternatively, perhaps this code is a common | beginner mistake due to ignorance of the proper | idiom... | | -Alex- | | | _________________________________________________________________ | S. Alexander Jacobson mailto:me@alexjacobson.com | tel:917-770-6565 http://alexjacobson.com | _______________________________________________ | Glasgow-haskell-users mailing list | Glasgow-haskell-users(a)haskell.org | http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

1 0

Re: :kind feature request in ghci
by Benjamin Franksen 28 Mar '04

28 Mar '04

On Saturday 27 March 2004 16:33, S. Alexander Jacobson wrote: > Alternatively, is there another way to > figure out how to create an instance of > a class like this? > > class MyClass a b where emptyVal::a b > > Here are one of the thing that don't work: > > type MyType a = [a] > instance MyClass (MyType a) a where emptyVal=[] > > And I have no idea why... Well, you've been told why. Anyway, why do you insist on using a two parameter type class? The more general class MyClass a where emptyVal::a would do exactly the same. To make type constructor B applied to type C an instance you just write instance MyClass (B C) where ... for instance instance MyClass [a] where emptyVal=[] If you enable overlapping instances you could even add instance MyClass [Int] where emptyVal=[0] The only difference to your 2-parameter class is that now you can declare instances for simple types like Integer too. Is that a problem? Ben

1 0

:kind feature request in ghci
by S. Alexander Jacobson 27 Mar '04

27 Mar '04

I keep running into kind errors (see my most recent post on Haskell cafe). Would it be possible for ghci to expose a :kind function that operates on types the way :type operates on values? Alternatively, is there another way to figure out how to create an instance of a class like this? class MyClass a b where emptyVal::a b Here are one of the thing that don't work: type MyType a = [a] instance MyClass (MyType a) a where emptyVal=[] And I have no idea why... -Alex- _________________________________________________________________ S. Alexander Jacobson mailto:me@alexjacobson.com tel:917-770-6565 http://alexjacobson.com

1 0

GHC 6.2.1 under darwinports
by Gregory Wright 26 Mar '04

26 Mar '04

Hi, The Glasgow Haskell Compiler supported under darwinports has been bumped to version 6.2.1. By default, GHC now builds with OpenGL support. The build should work for both OS X 10.2.x and 10.3.x. This lets Mac OS X jaguar users to get the latest version of GHC, as Wolfgang Thaller's binary release is only tested with panther. It can also provide amusement for those who just like to build from source. For information on getting darwinports, see http://darwinports.opendarwin.org. Best Wishes, Greg Wright Gregory Wright Antiope Associates LLC gwright(a)antiope.com

1 0

RE: [Haskell] deriving with newtypes
by Simon Peyton-Jones 25 Mar '04

25 Mar '04

[Redirecting to GHC users.] This seems like a very reasonable point. Not possible today, but it should be. It'll go on my to-do list. Meanwhile, I guess you can always do it by steam. Simon | -----Original Message----- | From: haskell-bounces(a)haskell.org [mailto:haskell-bounces@haskell.org] On Behalf Of Wolfgang | Jeltsch | Sent: 21 March 2004 17:55 | To: The Haskell Mailing List | Subject: [Haskell] deriving with newtypes | | Hello, | | I'm trying to use GHC's deriving mechanism for newtypes in the following way: | class C a b | instance C [a] Char | newtype T = T Char deriving C [a] | Unfortunately, this isn't possible. Is there a reason for this? Can I | circumvent this restriction? | | Wolfgang | | _______________________________________________ | Haskell mailing list | Haskell(a)haskell.org | http://www.haskell.org/mailman/listinfo/haskell

1 0

RE: Bug report.
by Simon Marlow 25 Mar '04

25 Mar '04

> Started getting: > > ghc-6.2.1.20040313: panic! (the `impossible' happened, GHC > version 6.2.1.20040313): > mkTyVar a1 {- tv a75q -} > > Please report it as a compiler bug to > glasgow-haskell-bugs(a)haskell.org, > or http://sourceforge.net/projects/ghc/. > > today... definitely a but to do with class inference, as it > started appearing > aftyer changing some instance constraints and fundeps on classes. > > Can't really be a lot more specific as code is big. Let me > know if you want > me to produce a small code extract which triggers it. Could you send us enough code to reproduce the bug? The smaller the better, but don't worry if you can't reduce it. Cheers, Simon

1 0