October 2010 - Haskell-Cafe

commutativity
by Patrick Browne 01 Nov '10

01 Nov '10

Hi, Below are two questions on commutative operations in Haskell. infixl 5 `com` com :: Int -> Int -> Int x `com` y = (x + y) commutative com a b = (a `com` b) == (b`com`a) -- 1 + 3 == 3 + 1 -- This gives true by virtue of the value of LHS and RHS being equal after the plus operation -- Question 1 -- commutative com 1 3 -- This also gives true. Is it because of commutative equation or because of the plus operation? -- Question 2 -- In Haskell can commutativity be specified as a property of infix operations? This message has been scanned for content and viruses by the DIT Information Services E-Mail Scanning Service, and is believed to be clean. http://www.dit.ie

4 4

[ANNAUNCE] ghcjs-0.1.0 Haskell to Javascript compiler
by Victor Nazarov 01 Nov '10

01 Nov '10

I've been working on this for some month and I think now I'm ready to share the results. http://github.com/sviperll/ghcjs Haskell to Javascript translator ================================ Project aims to provide solution to * compile modern Haskell libraries to Javascript files and use them in Ajax applications or * develop entire Ajax application in Haskell language Building -------- Code builds as standard haskell package $ runghc Setup configure $ runghc Setup build $ runghc Setup install Usage ----- To compile Haskell module to Javascript use `ghcjs` command. $ ghcjs Test.hs This command is merely equivalent to the following $ ghc --make Test.hs but it compiles to Javascript instead of native code. See examples folder for an example of loading and running haskell code from browser. Status ------ The code is in alpha stage. Feel free to experiment with it as you wish. Implementation -------------- Compiler is implemented as [GHC](http://www.haskell.org/ghc/) backend using GHC API. And been tested with GHC 6.12.1. -- Victor Nazarov

6 12

Edit Hackage
by Andrew Coppin 01 Nov '10

01 Nov '10

Today I uploaded a package to Hackage, and rediscovered something that you already know: I'm an idiot. More specifically, I copied the Cabal description from another package and then updated all the fields. Except that I forgot to update one. And now I have a package which I've erroneously placed in completely the wrong category. Is there any danger that at some point in the future, it might be possible to edit a package /after/ it has been released on Hackage? There are several reasons why you might wish to do this (beyond realising you did something wrong five minutes after hitting the upload button): The maintainer might change. The homepage might move. A better package might come along, making the existing one obsolete. Or you might want to stick a message on the package saying "hey, this version has a serious bug, please use the next version up instead!" (I guess you wouldn't want to actually change the code content of a package. That leaves the potential for several /different/ variants of "the same" package floating around, depending on when people installed it. That would be Bad.) Is there any danger that we might be able to do these kinds of things sometime soon?

16 25

Red links in the new haskell theme
by Victor Oliveira 01 Nov '10

01 Nov '10

Hi Cafe, I really liked the new colors of haskell theme, but... Is really red a good color for links? At least for me, red links looks like broken or already visited ones. And the worst is hackage docs. It is really eye tiring to read. It's just a thought. Maybe it just with me. What about you? []s Victor

14 27

who's in charge?
by Günther Schmidt 31 Oct '10

31 Oct '10

Hi all, this may be an odd question to some, but I think it's actually quite an un-extraordinary one. Who's in charge? Of Haskell I mean. If there was some alien from Planet Java to land on Planet Haskell and demand to be taken to our leader, whom would we take him to? Günther

32 55

Generic traversals for GADTs
by Ozgur Akgun 31 Oct '10

31 Oct '10

Café, SYB-style libraries (and especially Uniplate) make it very easy to run generic traversals (queries/transformations) on ADTs. > data Expr = ... > x :: Expr > f :: Expr -> Expr > transform :: (Expr -> Expr) -> Expr -> Expr > transform f x :: Expr -- applies f to x (and its children) in a bottom-up manner Uniplate is very powerful and doesn't require any hand written type-class instances to get going reasonably efficiently. I can simply re-implement the functionality of Uniplate myself, modulo requiring hand-written instances. (I didn't want to go into that road, since I don't need it for this argument.) I did this exercise, because I wanted to understand why we don't have similar generic traversal libraries for GADTs. And, it didn't take long :) I can post my trials but they are really baby steps and they wouldn't be helpful. My question is: Considering how useful generic traversal libraries and GADTs are, has anybody tried implementing a generic traversal library (like Uniplate or similar) for GADTs? Is it a hard (or impossible, or not efficient, or problematic in some other way) thing to implement, or is it just not a requirement for anyone else? Best, Ozgur

1 0

Re: [Haskell-cafe] Parsing workflow
by Stephen Tetley 31 Oct '10

31 Oct '10

On 31 October 2010 16:53, Nils Schweinsberg <ml(a)n-sch.de> wrote: > Am 31.10.2010 17:27, schrieb Stephen Tetley: >> >> Left factoring! :-) > > Stupid question: Whats that? :) > Actually a good question... Its a standard grammar transformation - if you have two alternative productions that share a common prefix A -> a b | a c Left factoring produces: A -> A_part (b | c) A_part -> a Left factoring can quickly lead to the grammar getting quite convoluted, but using 'try' can lead to very fragile parsers that stop working in mysterious ways when you extend them. Because parser combinators are just functions, you can do questionable hacks to make left factoring a bit less tedious: parseA :: Parser A parseA = do a <- a_part (parseB a <|> parseC a) parseB :: a -> Parser A parseB a = do b <- b_part return $ B a b parseC :: b -> Parser A parseC a = do c <- c_part return $ C a c

1 0

Current thinking on CompositionAsDot issue in haskell prime?
by Alexander Kjeldaas 31 Oct '10

31 Oct '10

Hi haskellers. Reading through the Haskell Prime suggestions, one that caught my eye is the CompositionAsDot issue. I'm especially thinking of the Pro issue: * Paves the way for using . as the selection operator in improved record or module systems I think I recognize this issue from common lisp. Basically the code becomes verbose because accessor functions usually need to redundantly encode the name of the module or struct in its name. The alternative is to require users to import qualified, and that doesn't deal with the occasional intra-module collision. This is a typical large-software-system issue and it annoys me a little that Haskell seems to have inherited the problems from common lisp. I worry that in large systems, a very conservative coding standard is needed with quite verbose naming conventions as that's the only safe way to avoid these conflicts in a large system while allowing refactoring code etc. What is the current thinking on this issue? Thanks, Alexander

9 12

Haskell meeting in Berlin
by Sönke Hahn 31 Oct '10

31 Oct '10

Hi all! There will be an informal Haskell meeting in Berlin. Date: Thursday, October 28th Time: from 20:00 Location: c-base, Rungestrasse 20, 10179 Berlin The bi-monthly lisp meeting takes place at the same time and place [1]. This is the first of hopefully more meetings to come, so we will probably discuss the details of following meetings. If you're in Berlin and want to talk to other Haskell users (or want to yell at lisp users for being from the dark side of the force), please come by. Cheers, Sönke [1] http://www.c-base.org/calender/phpicalendar/month.php

3 4

cabal mystery (#562?)
by Conor McBride 30 Oct '10

30 Oct '10

Hi I've just installed cabal-install (just as a user: I am nowhere near root) on our unix server at work. That went fine. Clearly, the sensible thing to do next is get hold of an up-to-date package list. So, I tried conor@cafe:~$ cabal update Downloading the latest package list from hackage.haskell.org cabal: Codec.Compression.Zlib: premature end of compressed stream Here's the verbose version. conor@cafe:~$ cabal -v3 update Downloading the latest package list from hackage.haskell.org Sending: GET http://hackage.haskell.org/packages/archive/00-index.tar.gz HTTP/1.1 User-Agent: cabal-install/0.8.2 Host: hackage.haskell.org proxy uri host: www-cache.strath.ac.uk, port: :8080 Creating new connection to www-cache.strath.ac.uk:8080 Received: HTTP/1.0 200 OK Date: Thu, 28 Oct 2010 12:59:17 GMT Server: Apache/2.2.9 (Debian) mod_python/3.3.1 Python/2.5.2 Last-Modified: Thu, 28 Oct 2010 12:38:17 GMT ETag: "1888003-20bb4e-493ac9dbe6040" Accept-Ranges: bytes Content-Length: 2145102 Content-Type: application/x-tar Content-Encoding: x-gzip X-Cache: MISS from rouble-1.cc.strath.ac.uk X-Cache-Lookup: MISS from rouble-1.cc.strath.ac.uk:8080 Via: 1.1 rouble-1.cc.strath.ac.uk:8080 (squid/2.7.STABLE7) Connection: close Downloaded to /home/s3/conor/.cabal/packages/hackage.haskell.org/00-index.tar.gz cabal: Codec.Compression.Zlib: premature end of compressed stream So I poked about a bit, and I see it's a known issue. As a user, I'm wondering what to do. Just now, wget http://hackage.haskell.org/packages/archive/00-index.tar.gz is chugging away quite nicely. Anyhow, is there (a) something I can do to avoid this problem? I couldn't quite grasp what to do from the TRAC (b) some way to persuade cabal to use the file I'm currently downloading? Any tips to keep the gremlins at bay gratefully appreciated. Thanks Conor

3 2