- Glasgow-haskell-users

Case of "Language" pragma - bug?
by Robin Bate Boerop 02 Nov '07

02 Nov '07

I'm using GHC 6.6.1. The manual says: "Pragmas all take the form {-# word ... #-} where word indicates the type of pragma, and is followed optionally by information specific to that type of pragma. Case is ignored in word." However, when I use "Language CPP" instead of "LANGUAGE CPP" in the pragma, the pragma is ignored. Is this a documentation bug? To test the problem, one can use "ghc --make testpragma1.hs" and "ghc --make testpragma2.hs" on the following: -- Robin Bate Boerop

7 7

Incompatibility between Yi and GHC 6.8 (rc)
by Jean-Philippe Bernardy 25 Oct '07

25 Oct '07

Hello, I spent some time to make Yi work with 6.8, and failed up to now. Before I go into details, let's see what I'm trying to do with Yi. The idea is to have a fully dynamic application where the user is allowed to run arbitrary haskell code on the command line. This can be used to quickly test new code, etc. In particular, one needs to be able to "reload" the configuration in that manner. This is achieved basically by having everything running inside a specially made "GHC Session". The dynamic code is able to "reload itself" because it is passed a pointer to the GHC session. The GHC experts will see here that this cannot work :) Indeed, GHC has a number of "global variables", which, if accessed through code loaded in a GHC session, will be instanciated again. This means that accessing a "level 1" session with "level 2" code is normally not possible. However, I found a workaround: encapsulate all calls to the GHC API in a data structure in "level 1" code, which is then passed to "level 2". (A haskell version of C-style callbacks). While this worked beautifully with GHC 6.6, it fails with 6.8rc. with exception :: GhcException yi: panic! (the 'impossible' happened) (GHC version 6.8.0.20071019 for i386-unknown-linux): a static opt was looked at too early! This happens at a seemingly unrelated point in the code. (inside the bytecode interpreter?) I can't really report a bug here, because what I'm trying to do is probably fairly unsupported by GHC. However, I'd appreciate some sort of advice. Is my workaround the correct approach? How would GHC people implement a dynamic application? Should I drop the idea completely? ... Thanks in advance for your help :) -- JP Attached: a darcs patch to the Yi repo to (try to) support 6.8.

1 0

ANNOUNCE: GHC 6.8.1 Second Release Candidate
by Ian Lynagh 25 Oct '07

25 Oct '07

We are pleased to announce the Second Release Candidate phase for GHC 6.8.1. The 6.8.1 release is very close now. We only intend to look at the bugs listed here: http://hackage.haskell.org/trac/ghc/query?status=new&status=assigned&status… so if you think there is something important that is not listed there then this is the time to shout! You can get 6.8.1 snapshots to test from here: http://www.haskell.org/ghc/dist/stable/dist/ Right now we have the source bundles: http://www.haskell.org/ghc/dist/stable/dist/ghc-6.8.0.20071015-src.tar.bz2 http://www.haskell.org/ghc/dist/stable/dist/ghc-6.8.0.20071015-src-extralib… Only the first of these is necessary. The "extralibs" package contains various extra packages that are sometimes supplied with GHC - unpack the extralibs tarball on top of the source tree to add them, they will be included in the build automatically. There are also binary distributions for x86_64/Linux and i386/Linux, and an installer for Windows. Please test as much as possible, bugs are much cheaper if we find them before the release! We expect to release later this month. Thanks Ian, on behalf of the GHC team

6 10

kind error
by Brammert Ottens 24 Oct '07

24 Oct '07

Hello, I am trying to recompile some haskell code. I have previously compiled it with ghc 6.4 and that went fine. The lines that give the trouble are import qualified Data.Edison.Coll.UnbalancedSet as US import qualified Data.Edison.Coll.MinHeap as MH data DecoratedFormula = DF{formula :: Formula, iformula :: IFormula} type FastClause = MH.Min US.Set DecoratedFormula When compiling with ghc 6.6.1 and edison 1.2.1 I get the following error message: FastClause.hs:71:25: `US.Set' is not applied to enough type arguments Expected kind `*', but `US.Set' has kind `* -> *' In the type synonym declaration for `FastClause' I have no idea how to solve this error, since as far as I can see the type declarations of both MinHeap and Set do not seem to have changed. Any suggestions are more then welcome. Regards, Brammert

3 3

type families not advertised for 6.8
by Wolfgang Jeltsch 24 Oct '07

24 Oct '07

Hello, on <http://hackage.haskell.org/trac/ghc/ticket/1716#comment:2> I read: “we are not advertising type equalities for 6.8”. What does this mean? Is it possible that type family support will be removed from GHC at some point? Best wishes, Wolfgang

5 14

Can't seem to figure how to compile
by John Vogel 23 Oct '07

23 Oct '07

Running ghc 6.6.1 On Windows Vista I get this error when I run: ghc Main.hs gcc: installation problem, cannot exec `as': No such file or directory On this code: module Main where main = do putStrLn "Please enter your name: " name <- getLine putStrLn ("Hello, " ++ name ++ ", how are you?") What is my problem?

4 5

Bumping "extralib" library versions
by Duncan Coutts 23 Oct '07

23 Oct '07

All, There is unfortunately some synchronisation between ghc releases and the releases of various common libraries, the so-called "extralibs" collection. So I'd like to look briefly at what the version numbers of theses packages should be. It is important that we do increase the version numbers if there are changes since the last releases. In cases where the package's HEAD repo is maintained somewhere other than darcs.haskell.org/packages/$pkg that we do not break the version scheme used by the package maintainer. The current set of "extralibs" packages is: ALUT arrows cgi fgl GLUT haskell-src HGL html HUnit mtl network ObjectIO OpenAL OpenGL parallel parsec QuickCheck regex-base regex-compat regex-posix stm time X11 xhtml For each one I'd like to identify where the HEAD repo is, what the last released version was, what the current version number in the darcs version is, if there are any changes since that release and if so what the new release number ought to be. These version number bumps are based on existing versioning practise for these libraries rather than the proposed package versioning policy. http://haskell.org/haskellwiki/Package_versioning_policy#Version_numbers I have also listed for each package what the next version number would be under the proposed package versioning policy.They are marked as "PVP next version". If you think I've got any of these details wrong, please say so. ========== ALUT HEAD repo: http://darcs.haskell.org/packages/ALUT Released version: 2.0.1 (in ghc-6.6.1 extralibs tarball) Current version: 2.0.1 Changes since release: minor build system changes Suggested next version: 2.0.2 PVP next version: 2.0.1.1 ========== arrows HEAD repo: http://darcs.haskell.org/packages/arrows Released version: 0.3 (on hackage) Current version: 0.3 Changes since release: none Suggested next version: not needed PVP next version: not needed Notes: The latest version of this package depends on Stream which is not currently in the extralibs collection. This anomaly will need to be sorted out. ========== cgi HEAD repo: http://darcs.haskell.org/packages/cgi Released version: 3001.1.5 (on hackage) Current version: 3001.1.5 Changes since release: comments and build system changes Suggested next version: 3001.1.6 PVP next version: 3001.1.5.1 Notes: requires .cabal changes to build with older base versions ========== fgl HEAD repo: http://darcs.haskell.org/packages/fgl Released version: 5.4.1 (in ghc-6.6.1 extralibs tarball) Current version: 5.4.1 Changes since release: build system changes Suggested next version: 5.4.2 PVP next version: 5.4.1.1 Notes: requires .cabal changes to build with older base versions ========== GLUT HEAD repo: http://darcs.haskell.org/packages/GLUT Released version: 2.1.1 (in ghc-6.6.1 extralibs tarball) Current version: 2.1.1 Changes since release: build system changes Suggested next version: 2.1.2 PVP next version: 2.1.1.1 ========== haskell-src HEAD repo: http://darcs.haskell.org/packages/haskell-src Released version: 1.0.1 (in ghc-6.6.1 extralibs tarball) Current version: 1.0.1 Changes since release: build system and portability changes (same api) Suggested next version: 1.0.2 PVP next version: 1.0.1.1 Notes: requires .cabal changes to build with older base versions ========== HGL HEAD repo: http://darcs.haskell.org/packages/HGL Released version: 3.1.1 (in ghc-6.6.1 extralibs tarball) Current version: 3.1.1 Changes since: build system changes Suggested next version: 3.1.2 PVP next version: 3.1.1.1 Notes: requires .cabal changes to build with older base versions ========== html HEAD repo: http://darcs.haskell.org/packages/html Released version: 1.0.1 (in ghc-6.6.1 extralibs tarball) Current version: 1.0.1 Changes since: build system changes Suggested next version: 1.0.2 PVP next version: 1.0.1.1 ========== HUnit HEAD repo: http://darcs.haskell.org/packages/HUnit Released version: 1.1.1 (in ghc-6.6.1 extralibs tarball) Current version: 1.1.1 Changes since: some code changes, not obvious if api changed Suggested next version: 1.1.2 PVP next version: 1.1.2 ========== mtl HEAD repo: http://darcs.haskell.org/packages/mtl Released version: 1.0.1 (in ghc-6.6.1 extralibs tarball) Current version: 1.0.1 Changes since: added more module re-exports and docs Suggested next version: 1.0.2 PVP next version: 1.0.2 Note: does not build with ghc-6.4.x as base-1.0 ========== network HEAD repo: http://darcs.haskell.org/packages/network Released version: 2.0.1 (in ghc-6.6.1 extralibs tarball) Current version: 2.0.1 Changes since: IPV6 support added, api extended not changed Suggested next version: 2.0.2 PVP next version: 2.0.2 Note: needs a patch to build with ghc-6.4.x and base-1.0 ========== ObjectIO HEAD repo: http://darcs.haskell.org/packages/ObjectIO Released version: 1.0.1 (in ghc-6.6.1 extralibs tarball) Current version: 1.0.1 Changes since: Setup.hs changes only Suggested next version: 1.0.2 PVP next version: 1.0.1.1 ========== OpenAL HEAD repo: http://darcs.haskell.org/packages/OpenAL Released version: 1.3.1 (in ghc-6.6.1 extralibs tarball) Current version: 1.3.1 Changes since: build system changes Suggested next version: 1.3.2 PVP next version: 1.3.1.1 ========== OpenGL HEAD repo: http://darcs.haskell.org/packages/OpenGL Released version: 2.2.1 (in ghc-6.6.1 extralibs tarball) Current version: 2.2.1 Changes since: build system changes Suggested next version: 2.2.2 PVP next version: 2.2.1.1 ========== parallel HEAD repo: http://darcs.haskell.org/packages/parallel Released version: none Current version: 1.0 Changes since: n/a Suggested next version: 1.0 PVP next version: 1.0 Note: cannot build with older ghc/base versions as it exposes modules provided by base-2.x. ========== parsec HEAD repo: http://darcs.haskell.org/packages/parsec Released version: 2.0 (hackage and ghc-6.6.1 extralibs tarball) Current version: 2.0 Changes since: api additions (extended, not changed) Suggested next version: 2.0.1 PVP next version: 2.0.1 ========== QuickCheck HEAD repo: http://darcs.haskell.org/QuickCheck 1.x repo: http://darcs.haskell.org/packages/QuickCheck Released version: 1.0.1 (in ghc-6.6.1 extralibs tarball) Current version: 1.0.1 Changes since: build system Suggested next version: 1.0.2 PVP next version: 1.0.1.1 Note: we seem to be including QuickCheck 1.x rather than 2.x. ========== regex-base HEAD repo: http://darcs.haskell.org/packages/regex-unstable/regex-base 0.72 repo: http://darcs.haskell.org/packages/regex-base Released version: 0.72 (in ghc-6.6.1 extralibs tarball) 0.92 (on hackage) Current version: 0.72 (in ghc-6.8 extralibs) 0.92 (upstream) Changes since: build system changes Suggested next version: 0.72.1 PVP next version: 0.72.0.1 Note: we seem to be including regex-base 0.72 when the latest upstream version is 0.92. ========== regex-compat HEAD repo: http://darcs.haskell.org/packages/regex-unstable/regex-compat 0.71 repo: http://darcs.haskell.org/packages/regex-compat Released version: 0.71 (hackage and ghc-6.6.1 extralibs tarball) 0.90 (on hackage) Current version: 0.71 (in ghc-6.8 extralibs) 0.90 (upstream) Changes since: trivial code change and build system changes Suggested next version: 0.71.1 PVP next version: 0.71.0.1 Note: Setup.lhs needs fixing to be trivial. Again, we seem to be including regex-base 0.71 when the latest upstream version is 0.90. ========== regex-posix HEAD repo: http://darcs.haskell.org/packages/regex-unstable/regex-posix 0.72 repo: http://darcs.haskell.org/packages/regex-posix Released version: 0.71 (hackage and ghc-6.6.1 extralibs tarball) 0.91 (on hackage) Current version: 0.72 (in ghc-6.8 extralibs) 0.91 (upstream) Changes since: compatible code changes and build system changes Suggested next version: 0.72.1 PVP next version: 0.72.0.1 Note: will need patching to build with base-2.0. Again, we seem to be including regex-base 0.72 when the latest upstream version is 0.91. ========== stm HEAD repo: http://darcs.haskell.org/packages/stm Released version: 2.1 (hackage and ghc-6.6.1 extralibs tarball) Current version: 2.1 Changes since: two new new functions and build system changes Suggested next version: 2.1.1 PVP next version: 2.1.1 Note: will not currently build with base-2.0. It's not clear if this is a problem, or if stm is tightly tied to the base/ghc version anyway. ========== time HEAD repo: http://darcs.haskell.org/packages/time Released version: 1.1.1 (in ghc-6.6.1 extralibs tarball) Current version: 1.1.1 Changes since: minor behaviour changes (bug fixes) Suggested next version: 1.1.2 PVP next version: 1.1.2 ========== X11 HEAD repo: http://darcs.haskell.org/packages/X11 Released version: 1.2.3 (on hackage) Current version: 1.2.3 Changes since: none Suggested next version: not needed PVP next version: not needed ========== xhtml HEAD repo: http://darcs.haskell.org/packages/xhtml Released version: 3000.0.2 (in ghc-6.6.1 extralibs tarball) Current version: 3000.0.2 Changes since: minor build system changes Suggested next version: 3000.0.3 PVP next version: 3000.0.2.1 Additionally, there are some core libs that could be usable with older versions of other libs, like base and so could be used with older releases of ghc or other implementations: bytestring Cabal filepath haskell98 readline unix Win32 ========== bytestring HEAD repo: http://darcs.haskell.org/bytestring Released version: 0.9 (on hackage) Current version: 0.9 Changes since: none Suggested next version: not needed PVP next version: not needed ========== Cabal HEAD repo: http://darcs.haskell.org/cabal Released version: 1.2.1 (on hackage) Current version: 1.2.1 Changes since: some minor changes Suggested next version: 1.2.2 PVP next version: 1.2.2 ========== filepath HEAD repo: http://darcs.haskell.org/packages/filepath 1.1 repo: http://darcs.haskell.org/ghc-6.8/packages/filepath Released version: 1.0 (on hackage) Current version: 1.1 Changes since: api additions, bug fixes Suggested next version: 1.1 PVP next version: 1.0.1 Note: filepath head branch is marked as version 1.0 while the version included with ghc-6.8 is marked as 1.1. ========== haskell98 HEAD repo: http://darcs.haskell.org/packages/haskell98 Released version: 1.0 (hackage and ghc-6.6.1 extralibs tarball) Current version: 1.0.1 Changes since: build system changes Suggested next version: ??? PVP next version: 1.0.0.1 Note: If the version number identifies the api to some degree, then the H98 api never changes so it's version number should never increment, except in the micro revision for build system changes. ========== readline HEAD repo: http://darcs.haskell.org/packages/readline Released version: 1.0 (ghc-6.6.1 extralibs tarball) Current version: 1.0.1 Changes since: api additions, build system changes Suggested next version: 1.0.1 PVP next version: 1.0.1 ========== unix HEAD repo: http://darcs.haskell.org/packages/unix Released version: 2.1 (ghc-6.6.1 extralibs tarball) Current version: 2.2 Changes since: many api additions Suggested next version: 2.2 PVP next version: 2.1.1 Note: will not currently build with base-2.0. ========== Win32 HEAD repo: http://darcs.haskell.org/packages/Win32 Released version: 2.1 (hackage and ghc-6.6.1 extralibs tarball) Current version: 2.1.1 Changes since: api additions Suggested next version: 2.1.1 PVP next version: 2.1.1 Note: will not currently build with base-2.0. Duncan

5 11

Little help with length bounded lists (Addendum)
by Philip K.F. Hölzenspies 23 Oct '07

23 Oct '07

PS. This program requires: {-# OPTIONS -fglasgow-exts -fallow-undecidable-instances #-} PPS. Sorry for the extra mail

1 0

Little help with length bounded lists
by Philip K.F. Hölzenspies 23 Oct '07

23 Oct '07

Dear GHC-ers, I'm trying to write a program using list-length codings in types. Based on the material from the many "faking it" papers I've found, I wrote up the following class, where 'l' is the list type, 'a' is the type of the elements in the list and 'n' is the number of elements in the list: class BoundList l a n | l -> a n where asList :: l -> [a] llength :: l -> n mkList :: [a] -> (l,[a]) Now I made two obvious list-constructors WITH the corresponding type constructors: data Nil a = Nil data Cons a n tl = Cons a tl Empty lists are typed with the parameter of the type constructor. Non-empty lists are typed, but by using only the definition above, the list may be heterogeneous (there are no restrictions on tl). By making these types instances of the BoundList class, lists are homogeneous (see below). Lengths are coded linearly as: data Z -- zero data S x -- successor Giving the means to restrict the length of lists as well. The instance declarations look like this: instance BoundList (Nil a) a Z where asList _ = [] llength _ = (undefined :: Z) mkList as = (Nil, as) instance BoundList l a n => BoundList (Cons a (S n) l) a (S n) where asList (Cons a as) = a : asList as llength _ = (undefined :: S n) mkList (a:as) = (Cons a tl, rs) where (tl,rs) = mkList as Based on Uktaad B'mal's "Beginning Faking It" paper, I constructed reflection and reification of the length coding, viz: class Reflect s where intValue :: s -> Int instance Reflect Z where intValue _ = 0 instance Reflect x => Reflect (S x) where intValue _ = intValue (undefined :: x) + 1 reifyInt :: Int -> (forall s . Reflect s => s -> w) -> w reifyInt 0 k = k (undefined :: Z) reifyInt (n+1) k = reifyInt n (\(_::s) -> k (undefined :: S s)) What I can't bend my mind around, is how I should construct a function that will partition a normal list into a list of length bounded lists. Basically, I want to have a function that, given an Int n and a list list xs, that chops up xs into chunks of size n and actually constructs Cons/Nil lists for me. I understand that I can't "bring the type out," as in: partition :: BoundList l a n => Int -> [a] -> l but I would have hoped for a way to apply functions to my constructed bounded lists, like, maybe: partition :: Int -> [a] -> (forall l . BoundList l a n => l -> w) -> [w] Like I said, my mind doesn't bend this way. Does anybody have a suggestion that can help me out? Also, have I blatantly missed a library that will neatly do this for me? Eventually, I want to use this library so that I can define a block size in my function types. This could (hopefully) be used to do SDF-like run-time scheduling of threads, or - when distributed over multiple processors - processes. Regards, Philip

1 0

Status of exception handling
by Florian Weimer 22 Oct '07

22 Oct '07

What's the status of exception handling? A couple of years ago, I think it wasn't clear how to do that in the Haskell context. Are there any GHC extensions which provide means to retry on temporary errors, without passing too much status around? I've seen on the Wiki that backtraces are still an issue. I understand that the execution environment doesn't really provide them at reasonable cost, but I wonder if there are any tools to figure out, based on a core dump, why the exception was raised. I don't really need a traditional backtrace for that, I think. Information where the program would go next would be equally helpful, as long as it's possible to get to a piece of use-written code that is somewhat near the place where the actual bug lies. Even something as terse as Perl's error messages would better than nothing. (Perl's error reporting optionally uses the location of the first cross-module caller to pinpoint the error message; this is not necessarily close to the error, but better than pointing to an internal location.)

2 1