May 2010 - Haskell-Cafe

are forkIO threads event-driven?
by Aran Donohue 03 May '10

03 May '10

Hi Cafe, In GHC, if a thread spawned by forkIO blocks on some network or disk IO, is the threading system smart enough not to wake the thread until an IO event occurs on its input/output? The Control.Concurrent documentation doesn't specify, and the previous discussions I could find on this topic are out-of-date. There is a years-old GHC ticket, too, recently revived[2]. Put another way, is it possible yet to use forkIO for making a server to handle tens of thousands of concurrent network connections? If not, what is the best current Haskell/GHC way? Thanks, Aran [1] http://www.monkey.org/~provos/libevent/ [2] http://hackage.haskell.org/trac/ghc/ticket/635

8 14

Leibniz equality can be made injective
by oleg＠okmij.org 02 May '10

02 May '10

We show that type families let us write Leibniz equality witnesses of the injectivity of type constructors. We may use Leibniz equality not only for introduction but also for elimination. The paper on ``Typing Dynamic Typing'' (Baars and Swierstra, ICFP 2002) demonstrated the first implementation and application of Leibniz equality witnesses in Haskell: > newtype EQU a b = EQU{subst:: forall c. c a -> c b} A term (eq :: EQU a b) witnesses the equality of the types a and b: in any context c, the type a can be substituted with the type b. The context is represented by the type constructor c. The witness gives the constructive prove of substitutability: given a term of the type c a, we can always obtain a term of the type c b. Since we don't know anything about the context c, such a substitution can only happen if the two types are indeed equal. Hence the sole non-bottom witness is > refl :: EQU a a > refl = EQU id testifying that equality is reflexive. Leibniz equality became quite popular since it lets us represent GADTs to a large extent. The power of Leibniz equality comes from the fact that the context c may indeed be anything we want. The following example illustrates that power, by concisely proving that the equality is transitive. The proof of transitivity is very short > tran :: EQU a u -> EQU u b -> EQU a b > tran au ub = subst ub au We treat the type EQU a u as (EQU a) u, that is, as an application of the `type constructor' (EQU a) to the type 'u'. We then use the witness EQU u b to replace the type u with the type b in the context (EQU a), giving us the desired EQU a b. When writing type checkers and type inferencers in Haskell along the lines of typing dynamic typing, we also need to know that two arrow types are equal if and only if their components are equal. In one direction, > eq_arr :: EQU a1 a2 -> EQU b1 b2 -> EQU (a1->b1) (a2->b2) if the argument and result types of two arrow types are equal, the arrow types themselves are equal. Fortunately, for this, commonly used direction, the proof is easy, after we define suitable type-level combinators F1 and F2 to place the types being equated in the right context: > newtype F1 t b a = F1{unF1:: EQU t (a->b)} > newtype F2 t a b = F2{unF2:: EQU t (a->b)} > eq_arr :: EQU a1 a2 -> EQU b1 b2 -> EQU (a1->b1) (a2->b2) > eq_arr a1a2 b1b2 = > unF2 . subst b1b2 . F2 . unF1 . subst a1a2 . F1 $ refl In the type EQU (a1->b1) (a2->b2), the type `constructor' (F1 (a1->b1) b2) represents the context of the type a2, and the type constructor (F2 (a1->b1) a2) is the context of the type b2. Leibniz equality only applies if the context of a type is represented as a type constructor; therefore, we sometimes have to define auxiliary newtypes to put the contexts in the required form. In the reverse direction, we have to prove that if two arrow types are equal, their components (e.g., the argument types) are equal as well. The proof will witness the injectivity of the arrow type constructor. It is thought that injectivity cannot be proven with Leibniz equality at all. For example, the paper ``Implementing Cut Elimination: A Case Study of Simulating Dependent Types in Haskell'' by Chen, Zhu and Xi, PADL'04, needed the reverse, elimination direction and found it impossible to obtain with Leibniz equality witnesses. We demonstrate that type families help. Type families let us define ``subtractive contexts'', so that we can view a type 'a' as the type (a->b) placed into the context that removes (->b). Again we use an auxiliary newtype to make the subtractive context appear as a type constructor. > type family Fst a :: * > type instance Fst (x->y) = x > type instance Fst (x,y) = x > -- etc. > newtype FstA a b = FstA{unFstA :: EQU (Fst a) (Fst b)} > eq_arr1 :: EQU (a1->b1) (a2->b2) -> EQU a1 a2 > eq_arr1 eq = unFstA . subst eq $ ra > where > ra :: FstA (a1->b1) (a1->b1) > ra = FstA refl The complete code is available at http://okmij.org/ftp/Haskell/LeibnizInjective.hs The code includes a simple test of the injectivity.

1 0

Happy: PATH issue
by Limestraël 02 May '10

02 May '10

Hello Café, When I was trying to cabal-install haskell-src, I came up with: cabal: The program happy is required but it could not be found However, the happy package was actually installed and the 'happy' executable was in ~/.cabal/bin (which was in my PATH) I had to link ~/.cabal/bin/happy to /usr/bin/happy so that cabal-install may finally find it.

5 13

Re: LLVM backend building segfault (Louis Wasserman)
by Jonathan Fischoff 02 May '10

02 May '10

I got this too, on windows. On linux I got a similiar issue, but I think the message was different. I ran the ghc stage2 with valgrind and it complained about there being an invalid instruction. I wonder if it is a 64bit thing. Anyone been able to build the LLVM backend with a 64 bit machine? -Jonathan Fischoff > Hokay, so I've installed LLVM 2.7, > gotten ghc-HEAD-2009-10-23-ghc- > corelibs.tar.bz2, run ./darcs-all pull -a, > run ./boot, applied > add-new-llvm-code-generator-to-ghc_.dpatch > ghc-llvmbackend-20100328.dpatch > and written the following in mk/build.mk: > > GhcWithLlvmCodeGen = YES > GhcEnableTablesNextToCode = NO > SRC_HC_OPTS = -H64m -O0 -fasm > GhcStage1HcOpts = -O -fasm > GhcStage2HcOpts = -O0 -fasm > GhcLibHcOpts = -O -fasm > SplitObjs = NO > HADDOCK_DOCS = NO > > And when I run make, > > make -r --no-print-directory -f ghc.mk all > "inplace/bin/ghc-stage2" -H64m -O0 -fasm -i -iutils/ghctags/. > -iutils/ghctags/dist/build -iutils/ghctags/dist/build/autogen > -Iutils/ghctags/dist/build -Iutils/ghctags/dist/build/autogen > -package ghc -no-user-package-conf -rtsopts -odir > utils/ghctags/dist/build -hidir utils/ghctags/dist/build -stubdir > utils/ghctags/dist/build -hisuf hi -osuf o -hcsuf hc -c > utils/ghctags/./Main.hs -o utils/ghctags/dist/build/Main.o > "inplace/bin/ghc-stage2" -H64m -O0 -fasm -package-name dph-base-0.4.0 > -hide-all-packages -i -ilibraries/dph/dph-base/. > -ilibraries/dph/dph-base/dist-install/build > -ilibraries/dph/dph-base/dist-install/build/autogen > -Ilibraries/dph/dph-base/dist-install/build > -Ilibraries/dph/dph-base/dist-install/build/autogen > -Ilibraries/dph/dph-base/include -optP-include > -optPlibraries/dph/dph-base/dist-install/build/autogen/cabal_macros.h > -package array-0.3.0.0 -package base-4.2.0.0 -package ghc-prim-0.2.0.0 > -package random-1.0.0.2 -Odph -funbox-strict-fields -haddock -fcpr-off > -XTypeFamilies -XGADTs -XRankNTypes -XBangPatterns -XMagicHash > -XUnboxedTuples -XTypeOperators -O -fasm -fno-warn-deprecated-flags -Wwarn > -odir libraries/dph/dph-base/dist-install/build -hidir > libraries/dph/dph-base/dist-install/build -stubdir > libraries/dph/dph-base/dist-install/build -hisuf hi -osuf o -hcsuf hc -c > libraries/dph/dph-base/./Data/Array/Parallel/Base/DTrace.hs -o > libraries/dph/dph-base/dist-install/build/Data/Array/Parallel/Base/DTrace.o > "inplace/bin/ghc-stage2" -H64m -O0 -fasm -package-name dph-base-0.4.0 > -hide-all-packages -i -ilibraries/dph/dph-base/. > -ilibraries/dph/dph-base/dist-install/build > -ilibraries/dph/dph-base/dist-install/build/autogen > -Ilibraries/dph/dph-base/dist-install/build > -Ilibraries/dph/dph-base/dist-install/build/autogen > -Ilibraries/dph/dph-base/include -optP-include > -optPlibraries/dph/dph-base/dist-install/build/autogen/cabal_macros.h > -package array-0.3.0.0 -package base-4.2.0.0 -package ghc-prim-0.2.0.0 > -package random-1.0.0.2 -Odph -funbox-strict-fields -haddock -fcpr-off > -XTypeFamilies -XGADTs -XRankNTypes -XBangPatterns -XMagicHash > -XUnboxedTuples -XTypeOperators -O -fasm -fno-warn-deprecated-flags -Wwarn > -odir libraries/dph/dph-base/dist-install/build -hidir > libraries/dph/dph-base/dist-install/build -stubdir > libraries/dph/dph-base/dist-install/build -hisuf hi -osuf o -hcsuf hc -c > libraries/dph/dph-base/./Data/Array/Parallel/Base/Hyperstrict.hs -o > > libraries/dph/dph-base/dist-install/build/Data/Array/Parallel/Base/Hyperstrict.o > Segmentation fault > make[1]: *** > > [libraries/dph/dph-base/dist-install/build/Data/Array/Parallel/Base/Hyperstrict.o] > Error 139 > make[1]: *** Waiting for unfinished jobs.... > Segmentation fault > make[1]: *** [utils/ghctags/dist/build/Main.o] Error 139 > Segmentation fault > make[1]: *** > > [libraries/dph/dph-base/dist-install/build/Data/Array/Parallel/Base/DTrace.o] > Error 139 > make: *** [all] Error 2 > > I get segfaults. Segfaults make me unhappy. Halp plz? > > Louis Wasserman > wasserman.louis(a)gmail.com > http://profiles.google.com/wasserman.louis >

1 0

Problems with HStringTemplate UTF8 templates in ghc-6.12
by Dmitry Simonchik 02 May '10

02 May '10

Hi all, I need your help I have something i can't understand with hstringtemplate. I'm using this library in my project to render UTF-8 encoded templates. Everything was fine when I were using ghc-6.10 and hstringtemplate-0.6.1 But recently I upgraded to ghc-6.12 and hstringtemplate-0.6.2 and the output of my program now broken. I can see correctly only UTF8 values that I pass to template from database. It seems that toString method produces not UTF-8 encoded string, because if i use Codec.Binary.UTF8.String.encodeString i get correct output except for values from database. Any ideas how to solve the problem?

2 2

LLVM backend building segfault
by Louis Wasserman 02 May '10

02 May '10

Hokay, so I've installed LLVM 2.7, gotten ghc-HEAD-2009-10-23-ghc-corelibs.tar.bz2, run ./darcs-all pull -a, run ./boot, applied add-new-llvm-code-generator-to-ghc_.dpatch ghc-llvmbackend-20100328.dpatch and written the following in mk/build.mk: GhcWithLlvmCodeGen = YES GhcEnableTablesNextToCode = NO SRC_HC_OPTS = -H64m -O0 -fasm GhcStage1HcOpts = -O -fasm GhcStage2HcOpts = -O0 -fasm GhcLibHcOpts = -O -fasm SplitObjs = NO HADDOCK_DOCS = NO And when I run make, make -r --no-print-directory -f ghc.mk all "inplace/bin/ghc-stage2" -H64m -O0 -fasm -i -iutils/ghctags/. -iutils/ghctags/dist/build -iutils/ghctags/dist/build/autogen -Iutils/ghctags/dist/build -Iutils/ghctags/dist/build/autogen -package ghc -no-user-package-conf -rtsopts -odir utils/ghctags/dist/build -hidir utils/ghctags/dist/build -stubdir utils/ghctags/dist/build -hisuf hi -osuf o -hcsuf hc -c utils/ghctags/./Main.hs -o utils/ghctags/dist/build/Main.o "inplace/bin/ghc-stage2" -H64m -O0 -fasm -package-name dph-base-0.4.0 -hide-all-packages -i -ilibraries/dph/dph-base/. -ilibraries/dph/dph-base/dist-install/build -ilibraries/dph/dph-base/dist-install/build/autogen -Ilibraries/dph/dph-base/dist-install/build -Ilibraries/dph/dph-base/dist-install/build/autogen -Ilibraries/dph/dph-base/include -optP-include -optPlibraries/dph/dph-base/dist-install/build/autogen/cabal_macros.h -package array-0.3.0.0 -package base-4.2.0.0 -package ghc-prim-0.2.0.0 -package random-1.0.0.2 -Odph -funbox-strict-fields -haddock -fcpr-off -XTypeFamilies -XGADTs -XRankNTypes -XBangPatterns -XMagicHash -XUnboxedTuples -XTypeOperators -O -fasm -fno-warn-deprecated-flags -Wwarn -odir libraries/dph/dph-base/dist-install/build -hidir libraries/dph/dph-base/dist-install/build -stubdir libraries/dph/dph-base/dist-install/build -hisuf hi -osuf o -hcsuf hc -c libraries/dph/dph-base/./Data/Array/Parallel/Base/DTrace.hs -o libraries/dph/dph-base/dist-install/build/Data/Array/Parallel/Base/DTrace.o "inplace/bin/ghc-stage2" -H64m -O0 -fasm -package-name dph-base-0.4.0 -hide-all-packages -i -ilibraries/dph/dph-base/. -ilibraries/dph/dph-base/dist-install/build -ilibraries/dph/dph-base/dist-install/build/autogen -Ilibraries/dph/dph-base/dist-install/build -Ilibraries/dph/dph-base/dist-install/build/autogen -Ilibraries/dph/dph-base/include -optP-include -optPlibraries/dph/dph-base/dist-install/build/autogen/cabal_macros.h -package array-0.3.0.0 -package base-4.2.0.0 -package ghc-prim-0.2.0.0 -package random-1.0.0.2 -Odph -funbox-strict-fields -haddock -fcpr-off -XTypeFamilies -XGADTs -XRankNTypes -XBangPatterns -XMagicHash -XUnboxedTuples -XTypeOperators -O -fasm -fno-warn-deprecated-flags -Wwarn -odir libraries/dph/dph-base/dist-install/build -hidir libraries/dph/dph-base/dist-install/build -stubdir libraries/dph/dph-base/dist-install/build -hisuf hi -osuf o -hcsuf hc -c libraries/dph/dph-base/./Data/Array/Parallel/Base/Hyperstrict.hs -o libraries/dph/dph-base/dist-install/build/Data/Array/Parallel/Base/Hyperstrict.o Segmentation fault make[1]: *** [libraries/dph/dph-base/dist-install/build/Data/Array/Parallel/Base/Hyperstrict.o] Error 139 make[1]: *** Waiting for unfinished jobs.... Segmentation fault make[1]: *** [utils/ghctags/dist/build/Main.o] Error 139 Segmentation fault make[1]: *** [libraries/dph/dph-base/dist-install/build/Data/Array/Parallel/Base/DTrace.o] Error 139 make: *** [all] Error 2 I get segfaults. Segfaults make me unhappy. Halp plz? Louis Wasserman wasserman.louis(a)gmail.com http://profiles.google.com/wasserman.louis

1 0

WGP 2010 Second Call for Papers (Apologies for multiple copies)
by Bruno Oliveira 02 May '10

02 May '10

====================================================================== CALL FOR PAPERS WGP 2010 6th ACM SIGPLAN Workshop on Generic Programming Baltimore, Maryland, US Sunday, September 26th, 2010 http://osl.iu.edu/wgp2010 Collocated with the International Conference on Functional Programming (ICFP 2010) ====================================================================== Goals of the workshop --------------------- Generic programming is about making programs more adaptable by making them more general. Generic programs often embody non-traditional kinds of polymorphism; ordinary programs are obtained from them by suitably instantiating their parameters. In contrast with normal programs, the parameters of a generic program are often quite rich in structure; for example they may be other programs, types or type constructors, class hierarchies, or even programming paradigms. Generic programming techniques have always been of interest, both to practitioners and to theoreticians, and, for at least 20 years, generic programming techniques have been a specific focus of research in the functional and object-oriented programming communities. Generic programming has gradually spread to more and more mainstream languages, and today is widely used in industry. This workshop brings together leading researchers and practitioners in generic programming from around the world, and features papers capturing the state of the art in this important area. We welcome contributions on all aspects, theoretical as well as practical, of * polytypic programming, * programming with dependent types, * programming with type classes, * programming with (C++) concepts, * generic programming, * programming with modules, * meta-programming, * adaptive object-oriented programming, * component-based programming, * strategic programming, * aspect-oriented programming, * family polymorphism, * object-oriented generic programming, * and so on. Organizers ---------- Co-Chair Bruno C. d. S. Oliveira, Seoul National University Co-Chair Marcin Zalewski, Indiana University Programme Committee ------------------- Alley Stoughton, Kansas State University Andrei Alexandrescu, Facebook Bruno C. d. S. Oliveira (Co-Chair), Seoul National University Doug Gregor, Apple Gilad Bracha, I am a Computational Theologist Emeritus Magne Haveraaen, Universitetet i Bergen Marcin Zalewski (Co-Chair), Indiana University Neil Mitchell, Standard Chartered Ralf Lämmel, University of Koblenz-Landau Shin-Cheng Mu, Academia Sinica Thorsten Altenkirch, University of Nottingham Ulf Norell, Chalmers University Important Information --------------------- We plan to have formal proceedings, published by the ACM. Submission details Deadline for submission: Sunday 2010-06-13 Notification of acceptance: Monday 2010-07-12 Final submission due: Tuesday 2010-07-27 Workshop: Sunday 2010-09-26 Authors should submit papers, in postscript or PDF format, formatted for A4 paper, to the WGP09 EasyChair instance by 13th of June 2010. The length should be restricted to 12 pages in standard (two-column, 9pt) ACM format. Accepted papers are published by the ACM and will additionally appear in the ACM digital library. History of the Workshop on Generic Programming ---------------------------------------------- This year: * Baltimore, Maryland, US 2010 (affiliated with ICFP10) Earlier Workshops on Generic Programming have been held in * Edinburgh, UK 2009 (affiliated with ICFP09) * Victoria, BC, Canada 2008 (affiliated with ICFP), * Portland 2006 (affiliated with ICFP), * Ponte de Lima 2000 (affiliated with MPC), * Marstrand 1998 (affiliated with MPC). Furthermore, there were a few informal workshops * Utrecht 2005 (informal workshop), * Dagstuhl 2002 (IFIP WG2.1 Working Conference), * Nottingham 2001 (informal workshop), There were also (closely related) DGP workshops in Oxford (June 3-4 2004), and a Spring School on DGP in Nottingham (April 24-27 2006, which had a half-day workshop attached). Additional information: The WGP steering committee consists of J Gibbons, R Hinze, P Jansson, J Jarvi, J Jeuring, B Oliveira, S Schupp and M Zalewski

1 0

Haskell Jaskell Interface?
by John Creighton 02 May '10

02 May '10

I was wondering if a good approach to java interoperability might be a Haskell Jaskell Interface. Jaskell allows some haskell programs to run on the java virtual machine and dealt with some of the issues with regards to interfacing haskell with java on the virtual machine side of things. Therefore if their was say a GHC bridge between haskell and jaskell their would be two similar languages talking to each other. Code ment to be more closely coupled with java could be coded in jaskell and code that requires say GHC extensions could be written in haskell and perhaps some io interface between the two languages.

1 0

FGL instance constraint
by Kevin Quick 02 May '10

02 May '10

I need help understanding how to express the following: > data (Cls a) => B a = B [a] > data GrB a b = GrB (B a) > instance Graph GrB where ... In the methods for the instance specification, I need to perform Cls a operations on a. * As shown, the compiler complains that it cannot deduce (Cls a) from the context () on those methods. * I can't redefine the Graph methods to introduce the (Cls a) constraint [reasonable] * If I try to express the constraint as part of the Graph instance: "instance (Cls a) => Graph GrB where ..." then it says it's an ambiguous constraint because 'a' isn't mentioned. * I've tried specifying a functional constraint: "instance (Cls a) => Graph GrB | GrB -> a where ..." but that's not valid for an instance declaration. * I can't include a in the GrB instance: "instance (Cls a) => Graph (GrB a b) where ..." because that's a kind conflict. Suggestions/solutions are appreciated. Thanks! -- -KQ

6 22

Build problems (hsp, trhsx, ultimately Happstack)
by Alexander Solla 01 May '10

01 May '10

Consider the following bash session: [ ajs@kizaru:~/ ]$ which trhsx /home/ajs/.cabal/bin/trhsx [ ajs@kizaru:~/ ]$ trhsx Usage: trhsx <infile> [<outfile>] [ ajs@kizaru:~/ ]$ cabal install hsp Resolving dependencies... Configuring hsp-0.4.5... Preprocessing library hsp-0.4.5... Building hsp-0.4.5... ghc: could not execute: trhsx cabal: Error: some packages failed to install: hsp-0.4.5 failed during the building phase. The exception was: ExitFailure 1 Does anybody have any suggestions? I'm trying to install Happstack on Arch Linux (GHC 6.12.1). ~/.cabal/bin/ is in my path and trhsx runs. I've had this problem before, and I think I ran cabal as root for the HSP package. That's not working this time. I get the same failure.

7 19