August 2001 - Haskell

PLEAC - Programming Language Examples Alike Cookbook
by Simon Peyton-Jones 14 Aug '01

14 Aug '01

Folks, [copied from the Caml mailing list; thanks to Pixel] I came across the PLEAC project recently. There's a Haskell section in it, and it looks like something to which people on this list would be well qualified to contribute. Olaf, John: worth a link from haskell.org? Simon http://pleac.sourceforge.net/ http://pleac.sourceforge.net/pleac-faq.html http://freshmeat.net/projects/pleac The aim of the project is to rewrite Perl CookBook's code in other programming languages, to: - provide an easy resource for quick learning of some languages field (file IOs, string manipulation, etc) - better see the differences and strengths of the languages - provide a quick reference for language comparisons - and, erhm, advocate the graphical layout of a good Emacs configuration There is a web infrastructure enabling syntaxically colorized sourcecode, and the Perl code has been imported from what's given for free at oreilly (also, the copyright stuff has been addressed with oreilly). We would very much enjoy a OCaml version, which has not yet begun for the moment. As for now, the Merd version is around 30% complete ; the Ruby version is around 25% complete ; the Haskell and Python versions are around 10% complete ; and embryonic Nasm and Masd versions are also available. [...caml stuff...] You may subscribe to the ML if interested (see the web for details). cu Pixel.

1 0

inference with functional dependencies
by Avi Pfeffer 14 Aug '01

14 Aug '01

Inferring equality between types when there are functional dependencies seems to be less powerful than I expected. Here's a simple example: class Eq b => C a b | a -> b data T a = forall b . C a b => T b data U a = forall b . C a b => U b compare :: T a -> U a -> Bool compare (T x) (U y) = x == y I expected the compiler (GHC) would be able to deduce that the b type in the representation of T a and U a must be the same, since both stand in a C a b relationship, and a functionally determines b in that relationship. Instead I get the message: Inferred type is less polymorphic than expected Quantified type variable `b1' is unified with `b' When checking a pattern that binds x :: b y :: b1 In an equation for function `Test.compare': Test.compare (T x) (U y) = x == y Is this expected behavior? Is there a way for me to provide the necessary hints so that code like this could be accepted? Thanks, Avi Pfeffer -- Avi Pfeffer avi(a)eecs.harvard.edu www.eecs.harvard.edu/~avi Maxwell Dworkin 251 Division of Engineering and Applied Sciences Tel: (617) 496-1876 Harvard University Fax: (617) 496-1066

2 1

RE: Haskell Kernel
by Simon Peyton-Jones 14 Aug '01

14 Aug '01

It's not formally defined, I'm afraid. The Haskell report simply desribes how to translate some constructs into simpler ones. It does not formally define the set of simpler one. (This isn't something I propose to fix in the revised Haskell 98 report!) Simon | -----Original Message----- | From: Juan Carlos Arévalo Baeza [mailto:jcab@roningames.com] | Sent: 10 August 2001 23:13 | To: Haskell List | Subject: Haskell Kernel | | | | Is there any place where I could find detailed | documentation on the Haskell kernel? The Haskell Report | doesn't seem to have much on this. | | Salutaciones, | JCAB | email: jcab(a)roningames.com | ICQ: 101728263 | The Rumblings are back: http://www.JCABs-Rumblings.com | | | | _______________________________________________ | Haskell mailing list | Haskell(a)haskell.org http://www.haskell.org/mailman/listinfo/haskell |

1 0

RE: Unary minus (was: micro-rant)
by Simon Peyton-Jones 14 Aug '01

14 Aug '01

Folks I suggest we transfer this discussion about unary minus to haskell-cafe(a)haskell.org. As the original poster noted, it is not a subject on which the existing Haskell report is ambiguous or inconsistent, so I don't propose to make any change in Haskell 98. The next version of Haskell, maybe. Simon | -----Original Message----- | From: Jerzy Karczmarczuk [mailto:karczma@info.unicaen.fr] | Sent: 13 August 2001 16:12 | Cc: haskell(a)haskell.org | Subject: Unary minus (was: micro-rant) | | | matt hellige: | | > ... i think people would | > be pretty frustrated if "-56.2" worked but not "-x"... | | Weell, the Cleaners around have to live with it. -5 works. | For variables they write ~x. But "-" in front of a number is | not a lexical entity, but a syntactic one, so f -5 and f (-5) | are a bit different. | | I didn't follow this discussion, but please avoid this mess | in Haskell. | | Jerzy Karczmarczuk | | _______________________________________________ | Haskell mailing list | Haskell(a)haskell.org http://www.haskell.org/mailman/listinfo/haskell |

1 0

Re: micro-rant
by Marcin 'Qrczak' Kowalczyk 13 Aug '01

13 Aug '01

Tue, 7 Aug 2001 20:20:49 -0700, Memovich, Gary <GARY.MEMOVICH(a)kla-tencor.com> pisze: > As long as were trying to clean up a final version of the Haskell > 98 report, lets simplify it a little by getting rid of unary minus. I'm against removing it, even if compatibility was not an issue. Yes, it's an irregularity. But it's convenient, readable, and it allows using negative constants in patterns. -- __("< Marcin Kowalczyk * qrczak(a)knm.org.pl http://qrczak.ids.net.pl/ \__/ ^^ SYGNATURA ZASTÊPCZA QRCZAK

3 2

Haskell Kernel
by Juan Carlos Ar�valo Baeza 10 Aug '01

10 Aug '01

Is there any place where I could find detailed documentation on the Haskell kernel? The Haskell Report doesn't seem to have much on this. Salutaciones, JCAB email: jcab(a)roningames.com ICQ: 101728263 The Rumblings are back: http://www.JCABs-Rumblings.com

1 0

Contemporary Letter
by Justin Huntington 10 Aug '01

10 Aug '01

Hi! How are you? I send you this file in order to have your advice See you later. Thanks [Filename: Contemporary Letter.doc.pif, Content-Type: application/mixed] The attachment file in the message has been removed by eManager.

2 1

RE: More feedback on Haskell 98 modules from the Programatica Team
by Simon Peyton-Jones 10 Aug '01

10 Aug '01

| However, I think there is a risk that name clashes may be | introduced. If module A defines and exports f, and imports | (qualified) and exports module B, which also defines f, then | a module C that imports A has two fs, both of which have the | qualified name A.f in C, even though there is no conflict in | A. This suggests that only unqualified imports should be | exported in a "module X" export list entry. The Report already covers this point (though I don't have it to hand). There must be no name clashes among the *unqualified* names of the exported things; so in your example, module A's export list is illegal. Simon

2 1

Re: More feedback on Haskell 98 modules from the Programatica Team
by Ketil Malde 10 Aug '01

10 Aug '01

"Simon Peyton-Jones" <simonpj(a)microsoft.com> writes: > Wrong, and I hope that the report is already unambiguous on this point. > If you say 'import A' then you can refer to "A.f" or "f" but definitely > not "B.f". The module qualifier is the name of the module named in the > import statement (or its "as" alias), not the name of the module where it was > originally defined. I tried this with GHC (files below), and it doesn't work the way you imply. I.e. a module export is exported with its qualification. I guess this is a bug, then? main.hs: ------------ import A main = putStr "foo" A.hs: ---------- module A (a, module B) where import qualified B a :: Int a = 3 B.hs: ---------- module B where b :: Int b = 1 transcript: ---------- Prelude> :l main Compiling B ( B.hs, interpreted ) Compiling A ( A.hs, interpreted ) Compiling Main ( main.hs, interpreted ) Ok, modules loaded: Main, A, B. (0.18 secs, 3408976 bytes) Main> b <no file>:0: Variable not in scope: `b' (0.00 secs, 0 bytes) Main> B.b 1 (0.01 secs, 0 bytes) Main> A.b <no file>:0: Variable not in scope: `A.b' (0.00 secs, 0 bytes) -kzm -- If I haven't seen further, it is by standing in the footprints of giants

1 0

Generic programming bundle released
by Ralf Laemmel 10 Aug '01

10 Aug '01

Folks, As a result of our work on functional transformation systems, large bananas, and traversal technology, we have come up with a new generic programming bundle, that is, Strafunski: http://www.cs.vu.nl/Strafunski/ The above web page lists papers (including a new draft on typed combinators for generic traversal) and download information for Strafunski. Strafunski should be useful for anyone in the Haskell community who needs to traverse syntaxes, or other representations. Strafunski is also just interesting as another incarnation of the general idea of generic programming. We appreciate your feedback. Regards, Joost Visser and Ralf Laemmel -- Dr.-Ing. Ralf Laemmel CWI & VU, Amsterdam, The Netherlands http://www.cwi.nl/~ralf/ http://www.cs.vu.nl/~ralf/

1 0