- Haskell - Haskell.org

ANNOUNCE: GCJNI - Java Native Interface for Haskell
by Antony Courtney 16 Dec '01

16 Dec '01

I am pleased to announce an experimental release of GCJNI. GCJNI is a library that allows Haskell to invoke Java code via the Java Native Interface (JNI). The implementation uses GreenCard to make the JNI (a C language interface) available to Haskell. GCJNI includes a few convenient features, such as: - Integration of the Haskell and Java garbage collectors, so that Java objects are garbage collected when they are no longer accessible from Haskell. - Type class based overloading, which makes it easy to pass common types (like Int, Float and String) to or from Java code. - A tool (GenBindings) which uses Java reflection on a set of compiled Java classes to generate a Haskell module with a simple, high-level, type-safe interface to the underlying Java code. GCJNI has been succesfully tested using both hugs98 and ghc under both Linux and Windows. Binary distributions are provided for all for of these (compiler, platform) combinations. More information (including pointers to the relevant distributions) is available from the GCJNI web page at: http://www.haskell.org/gcjni Bug reports and feedback should be sent to antony(a)apocalypse.org Enjoy! -- Antony Courtney Grad. Student, Dept. of Computer Science, Yale University antony(a)apocalypse.org http://www.apocalypse.org/pub/u/antony

1 0

gcd definition
by S.D.Mechveliani 16 Dec '01

16 Dec '01

Simon Peyton-Jones <simonpj(a)microsoft.com> writes > [..] > If someone could write a sentence or two to explain why gcd 0 0 = 0, > (ideally, brief ones I can put in the report by way of explanation), > I think that might help those of us who have not followed the details > of the discussion. Here it is. gcd n m :: Integer = if n == 0 && m == 0 then 0 else greatest integer that divides both n and m It is set so according to classic definition (by Pythago, Euclid ...) GCD x y = such g that divides both x and y and is a multiple of any g' that divides both x and y. In particular, GCD 0 0 :: Integer = 0 and can be nothing else. Because 0 divides 0 and divides nothing else, everything divides 0 (z*0 = 0). Other comments ------------------------------------------------- * People say, D.Knuth writes gcd 0 0 = 0 in his book. And he is a known mathematically reliable person. * Example. gcd 12 18 :: Integer = 6 or -6, because 6 divides 12 and 18, and any other such number divides 6. * The Haskell report says "greatest integer" for domain = Integer, and thus, chooses the sign `+' for the result. This choice is not a mistake and helps to write shorter programs. * All the above agrees with the modern generic theory of ideals (started in |XX by Kummer, Gauss, Lagrange) for any commutative domain R having the properties of (a /= 0, b /= 0 ==> a*b /= 0), existence of unique factorization to primes. According to it gcd x y = least by inclusion ideal of kind (g) that includes the ideal (x, y), where Ideal (a,b..c) =def= {x*a + y*b +..+ z*c | x,y..z <- R} - a subset in R. `includes` (as set) is a partial order on ideals, and it is true that (1) a divides b <==> (a) includes (b), (2) (a) includes (a). Specializing this definition to Integer, we obtain the source formula. ----------------- Serge Mechveliani mechvel(a)botik.ru

1 0

Integer Enum ?
by Frank Dellaert 15 Dec '01

15 Dec '01

Hi I'm not entirely clear how Integer can be an Enum instance. I thought Integer was arbitrary size, while for the Enum class you need to define a mapping from and to Int, which is bounded (in a machine dependent way, even ?). I'm probably missing something obvious... Cheers Frank Dellaert

2 1

Announce: Hugs Graphics Library
by Sigbjorn Finne 15 Dec '01

15 Dec '01

We are pleased to announce a new release of the Hugs Graphics Library, version 2.0.4. The Hugs Graphics Library gives the programmer access to the most interesting and portable parts of the Win32 and X11 library. The library is distributed as open source and is suitable for use in teaching and applications. This version is compatible with the December 2001 release of Hugs, but won't work with the February 2001 release. This version supports: * filled and unfilled 2-dimensional objects (text, lines, polygons, ellipses). * Bitmaps (Win32 version only, for now). * Control over text alignment, fonts, color. * Simple input events (keyboard, mouse, window resize) to support reactivity. * Timers and double-buffering to support simple animation. * Use of concurrency to avoid the usual inversion of the code associated with event-loop programming. * Multiple windows may be handled at one time. To keep the library simple and portable, the library makes no attempt to support: * User interface widgets (menus, toolbars, dialog boxes, etc.) * Palette manipulation and other advanced features. * Many kinds of input event. The library can be downloaded from: http://www.haskell.org/graphics/. Installation instructions are in graphics-2.0.4/Install. Bug reports should be sent to reid(a)cs.utah.edu Enjoy!

1 0

ANNOUNCE: Hugs98 Christmas release
by Johan Nordlander 15 Dec '01

15 Dec '01

------------------------------------------------------------------------------ __ __ __ __ ____ ___ _________________________________________ || || || || || || ||__ Hugs 98: Based on the Haskell 98 standard ||___|| ||__|| ||__|| __|| Copyright (c) 1994-2001 ||---|| ___|| World Wide Web: http://haskell.org/hugs || || Report bugs to: hugs-bugs(a)haskell.org || || Version: Dec 2001 _________________________________________ ------------------------------------------------------------------------------ We are pleased to announce a new release of Hugs98, a Haskell interpreter and programming environment for developing cool Haskell programs. Sources and binaries are freely available on the World-Wide Web. The most important features of this new release are: - The incompatibilities between Hugs and the Haskell Graphics Library have been fixed, and binaries for the HGL are now available on the Hugs download page. - The missing standard libraries Directory, CPUTime, Time and Locale have been added along with a complete implementation of Haskell98 IO. - Hugs is now delivered with most of the hslibs libraries installed in the lib/exts/ directory. The added modules cover the Edison, Parsec, HaXml, QuickCheck, concurrent, monad, and html subdirectories of hslibs. - The :set option now refuses the user to set a module search path that doesn't contain the Prelude. This is to protect users from accidentally rendering their Hugs setups unusable, esp. so on Windows machines where the options are persisted to the Registry. - MacOS X is now one of the supported unix ports, with pre-built binaries available on the download page. - Experimental support is provided for hierarchical module names, where a module name A.B.C is mapped onto the file path A/B/C{.hs,.lhs} and appended to each of the path prefixes in HUGSPATH until the name of a readable file is found. Most reported bugs have also been fixed; however, some issues still remain. See the heading "Hugs Bugs & Features" on the Hugs web site for further information. Send email to hugs-users-request(a)haskell.org to join the hugs-users mailing list. Bug reports should be sent to hugs-bugs(a)haskell.org. Send email to hugs-bugs-request(a)haskell.org to subscribe to the hugs-bugs list. The home page for Hugs is at http://www.haskell.org/hugs. ------------------------------------------------------------------------------ The Hugs 98 system is Copyright (c) Mark P Jones, Alastair Reid, the Yale Haskell Group, and the Oregon Graduate Institute of Science and Technology, 1994-2001, All rights reserved. It is distributed as free software under the license in the file "License", which is included in the distribution. ------------------------------------------------------------------------------

1 0

Confused about default
by Ian Lynagh 14 Dec '01

14 Dec '01

Hi all I am rather confused about default. In section 4.3.4 the report says "each ti must be a monotype for which Num ti holds" but according to grep this is the only place "monotype" appears in the report. If I have the module module TT (Foo(..)) where default (Foo, [Foo], Foo -> Int) data Foo = Foo deriving (Eq, Show) instance Num Foo where {} instance (Eq a, Show a) => Num [a] where {} instance Eq (a -> b) where {} instance Show (a -> b) where {} instance Num (a -> b) where {} then hugs accepts it and TT> 5 Program error: Undefined member: fromInteger ghc accepts it and TT> 5 5 nhc tells me Error when renaming:: Illegal type in default at 4:17 If I remove all the list and function stuff then hi says TT> 5 5 If I have a main = putStrLn $ show 5 then nhc gives me TT> main No default definition for class method fromInteger and ghci gives me Main> main *** Exception: TT.lhs:8: No instance nor default method for class operation PrelNum.fromInteger It seems to me that hugs is in the right with it's handling of 5, but I am not sure who is wrong with default ([Foo]) or (Foo -> Int). Finally, the context free grammar doesn't currently enforce the restriction that only one default declaration be given. Fixing it would make rather a mess, though. Ian

1 0

Re: Typesetting literate scripts in TeX
by Marcin 'Qrczak' Kowalczyk 14 Dec '01

14 Dec '01

> Recently I've hacked up a (yet another) TeX package for typesetting > literate scripts in TeX Cool! I will use it in my thesis. One bug: a line can be broken between an inline code and a comma which immediately follows it. -- __("< Marcin Kowalczyk * qrczak(a)knm.org.pl http://qrczak.ids.net.pl/ \__/ ^^ QRCZAK

1 0

gcd 0 0 = 0
by S.D.Mechveliani 14 Dec '01

14 Dec '01

People write on gcd 0 0 : Alan Bawden <Alan(a)LCS.MIT.EDU> > If you take the point-of-view that gcd is actually an operation on > ideals, then gcd(0, 0) is 0. I.e. define gcd(x, y) to be the smallest > z >= 0 such that {m*x + n*y | m, n in Z} = {n*z | n in Z}. This is > probably the most natural and general definition of gcd, and is, in fact, > what many mathematicians would expect. > > Also, it is nice to be able to know that gcd(x, 0) = x for -all- x. Frank Seaton Taylor <fstaylo(a)alpha.ncsc.mil> writes T> Let me chime in in favor of gcd(0,0) = 0 rather than undefined. T> T> On Tuesday, December 11, 2001, at 01:07 , Alan Bawden wrote: T> T> > Also, it is nice to be able to know that gcd(x, 0) = x for -all- x. T> T> Alan is in good company: T> T> 1. Steele in "Common Lisp: The Language" says the same thing. T> T> 2. Knuth in "The Art of Computer Programming" defines gcd(0,0)=0. T> classic algebra All right. Let us try to investigate this. In such cases one should consult mathematics, not programming. So, let us ignore "Common Lisp ...". And Knuth is all right, I think. Further, the definintion > gcd(x, y) to be the smallest > z >= 0 such that {m*x + n*y | m, n in Z} = {n*z | n in Z} is not natural. In particular, how does it generalize to gcd X Y for polynomials in X, Y with rational coefficients? It is 1. And how would one obtain 1 = f*X + g*Y ? According to classic algebra (Kummer, I guess), gcd x y in (such and such) domain R is the least by inclusion principal ideal (g) that includes the ideal (x, y). Here (a1..an) = set of all combinations y1*a1 + .. + yn*an, yi from R. According to this definition, gcd (X) (Y) = (1) in polynomials, all right, and gcd n m :: Integer is (0) Because (0) is the least by inclusion principal ideal containing (0). And this also contradicts the part "greatest integer that divides ..." in recent Haskell specification. Probably, the best specification would be gcd n m :: Integer = if n == 0 && m == 0 then 0 else greatest integer that divides both n and m ----------------- Serge Mechveliani mechvel(a)botik.ru

2 1

Typesetting literate scripts in TeX
by Patryk Zadarnowski 14 Dec '01

14 Dec '01

Recently I've hacked up a (yet another) TeX package for typesetting literate scripts in TeX, which, I think belongs in http://haskell.org/libraries/#tex". What's different about lambdaTeX is that it is not a cumbersome preprocessor, but a full TeX program (an almost-complete Haskell lexical analyzer written entirely in plain TeX ;-) which typesets Bird-style literate scripts _directly_. The output looks much like the code from Chris Okasaki's book "Purely Functional Data Structures", doing syntax highlighting and converting ASCII art such as "->" or "alpha" to proper mathematical symbols. It should work with both LaTeX and plain TeX, and it does its magic without any annotations, directly on the source code (the only thing that you will probably want to do is add \input lambdaTeX at the top of your source file, and manually typeset your literate comments so they look as good as the source code ;-) I've got a rudimentary web page for it at http://www.jantar.org/lambdaTeX/ with examples and a summary of features. If you have any comments or suggestions, I wouldn't mind hearing them. Enjoy. Pat. -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Patryk Zadarnowski University of New South Wales <pat(a)jantar.org> School of Computer Science and Engineering -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

1 0

Re: We need "Documentation" (Was: Re: Integer to String Conversion?)
by Tony Davie 13 Dec '01

13 Dec '01

>-----BEGIN PGP SIGNED MESSAGE----- >Hash: SHA1 > >On Tuesday 04 December 2001 02:51, you wrote: >> On Tue, 4 Dec 2001, Chris wrote: >> > is there a function that converts Integers to Strings and vice versa? >> >> Prelude> (reads "123 abc") :: [(Integer, String)] >> [(123," abc")] >> Prelude> show 123 > > "123" >> > ... >Why doesn't an author or two who have written Haskell books consider opening >their books for public use on the Internet? "The Craft Of ..." especially >seems great. I'm sure many coders would appreciate a non-dead-tree version of >that book or a shortened version of the book. > >The material we have is simply lacking. What we need is a good tutorial and a >comprehensive reference book. ... I suggest you look at http://www.haskell.org/bookshelf/, specifically the section on Papers available on the Web. This gives numerous tutorials and introductions. -- Tony Davie Computer Science (retired), St.Andrews University G.F. Handel Site: http://bruichladdich.dcs.st-and.ac.uk/HandelWWW/HandelCat.html Handel was half German, half Italian, and half English. He was very large. (from a music exam)

3 4