November 2010 - Haskell-Cafe

new version of the mtl package
by Ross Paterson 08 Nov '10

08 Nov '10

After lengthy discussion on the libraries list, the mtl package has been updated to depend on the transformers package, from which mtl re-exports the monad transformers and the MonadTrans and MonadIO classes. This makes the two packages compatible. The monads-fd package is now a deprecated stub re-exporting the mtl modules. Ths transformers/mtl-2 combination is under consideration for inclusion in the Haskell Platform: http://trac.haskell.org/haskell-platform/wiki/Proposals/transformers Discussion is on the libraries list. So far there have been no objections. The new version has a few incompatibilities with the old, which will break a few packages, though the great majority build unchanged. The above web page lists the incompatibilities and suggested fixes. You might also consider whether the portable transformers package suffices for your package. As packages move to the new version, there will be increasing dependency clashes with packages that currently specify a bound on their mtl dependency excluding the new version. In some cases the dependency can be simply relaxed with no other change, or with small changes obtain compatibility with both versions. In other cases one must choose a version, or use cpp hackery to support both.

1 0

Mysterious fact
by Andrew Coppin 08 Nov '10

08 Nov '10

The other day, I accidentally came up with this: |{-# LANGUAGE RankNTypes #-} type Either x y= forall r. (x -> r) -> (y -> r) -> r left :: x -> Either x y left x f g= f x right :: y -> Either x y right y f g= g y | This is one example; it seems that just about any algebraic type can be encoded this way. I presume that somebody else has thought of this before. Does it have a name?

12 14

[issue1002] darcs support for ohloh: call for help
by Simon Michael 08 Nov '10

08 Nov '10

Many of us are interested, and Ohloh is interested, in getting ohloh.net to support darcs repositories directly. It doesn't look hard to get good-enough support working; it just requires hacking 10 or so simple, well-organised ruby files. I've made a start on github[1], and set up a collaborative how-to & notes page here: http://etherpad.osuosl.org/pCIn3uhOeW Please help! And forward this to appropriate places. Best, -Simon [1] will be merged with prior work as needed

1 3

Haskell meeting in Berlin on Tuesday Nov 9th
by Daniel van den Eijkel 08 Nov '10

08 Nov '10

Hi everybody, just wanted to send this again to the list - it's tomorrow. Cheers Daniel Am 10/31/10 6:30 PM, schrieb Sönke Hahn: > I can confirm the next meeting: > > Date: Tuesday, November 9th > Time: from 20:00 > Location: c-base, Rungestrasse 20, 10179 Berlin > > I was asked by some people from c-base, if we could do a talk about Haskell > for non-haskellers, which is a good idea, IMHO. We could discuss the details > in the mext meeting. > > Sönke >

1 0

ANN: cil-0.1.0
by Tom Hawkins 08 Nov '10

08 Nov '10

CIL (C Intermediate Language) [1], not to be confused with the Common Intermediate Language, is a mature OCaml library that parses and reduces C programs down to a simplified subset of the C language, making it easier to analyze and compile C programs. This library [2, 3] parses these results, providing a Haskell interface to CIL. -Tom [1] http://cil.sourceforge.net/ [2] http://hackage.haskell.org/package/cil [3] http://github.com/tomahawkins/cil

2 1

ANN: Copilot 0.25
by Lee Pike 08 Nov '10

08 Nov '10

Copilot[1] is a dataflow stream language built on Atom[2] for writing hard real-time C monitors of embedded systems. Copilot is a Haskell eDSL. The current release includes the following improvements: * Added true & false Specs (Spec Bool). * Removed generic const -- unneeded, since Spec is an instance of Num. * Change casting to only allow casts that (1) are guaranteed not to change the sign and (2) are to a larger type. * Added triggers---C functions to call when a monitor fails. * Removed libs from Copilot.hs. You must import these explicitly. * Fixed bugs with LTL and ptLTL libraries & examples and added documentation. * Added external arrays for sampling. * Fixed a bug in external array analysis for computing the minimum period size. * Added the ability to specify a HW clock for Atom. [1] <http://hackage.haskell.org/package/copilot> and <http://leepike.github.com/Copilot/> [2] <http://hackage.haskell.org/package/atom> Regards, Lee

1 0

Re: internship opportunities in France
by Lee Pike 08 Nov '10

08 Nov '10

Hi, On a (possibly-related) note: Galois, Inc. and the National Institute of Aerospace (NIA) will likely be hosting a Visiting Scholar this summer for a joint NASA-sponsored project, located in Virginia, on the East coast of the U.S. (Since this particular opportunity is a visiting scholar position, we pay all travel/living expenses, but no salary, so one does not need a visa/work permit to work in the U.S.) We had two visiting scholars last summer, from Germany and France respectively, publishing a paper and even flying some aircraft <http://www.galois.com/blog/2010/09/22/copilot-a-dsl-for-monitoring-embedded…>. The project is a Haskell DSL project focused on monitoring embedded systems. Interested folks can contact Alwyn Goodloe (Cc'ed) at the NIA and myself, off the list. Regards, Lee > Hello folks ! > I am a Computer Science student looking for an internship of 6 months here > in France. Does anybody know of any company working with Haskell ? > Thanks in advance :D, > > Lorenzo Fundaró García

2 2

How can I use MACID in my existing application?
by Dupont Corentin 08 Nov '10

08 Nov '10

Hello, I'm wondering how can I use Happstack's MACID in my application without breaking everything. I have a monad like that: type Comm = StateT Communication IO type GameState a = StateT Game Comm a and many functions like: foo :: GameState () foo = do lift $ putComm <some message to player's channel> modify <someAction> The state of the game is stored in Game. Comm is used as an abstraction to communicate over several channels with players. Whereas MACID asks to use: type *Update* state = Ev<http://happstack.com/docs/0.5.0/happstack-state/Happstack-State.html#t%3AEv>( StateT<http://happstack.com/docs/0.5.0/mtl/Control-Monad-State-Lazy.html#t%3AStateT>state STM <http://happstack.com/docs/0.5.0/base/GHC-Conc.html#t%3ASTM>) How can I use this without modifying everything?? I understand that MACID must record the <someAction> from above but the message should not. Thanks for help! Corentin

4 5

"Best practices" for Text
by Ivan Lazar Miljenovic 08 Nov '10

08 Nov '10

With the recent discussions on the libraries mailing list about the API of Bryan O'Sullivan's text library, and the fact that I'm looking at using it soon for the first time, are there any recommended best practices or tips & tricks people have come up with? For starters, I'm considering writing a polyparse [1] instance for Text. However, even with the current Bytestring instances for polyparse there seems to be an emphasis on character-based parsing. Is that the correct way of doing things? For example, what would be the best way to try to parse a text value when you don't care about case? At the moment, with my String-based parser, I'm doing repeated comparisons on Char values [2,3]. [1]: http://hackage.haskell.org/package/polyparse [2]: http://hackage.haskell.org/packages/archive/graphviz/2999.10.0.1/doc/html/s… [3]: http://hackage.haskell.org/packages/archive/graphviz/2999.10.0.1/doc/html/s… -- Ivan Lazar Miljenovic Ivan.Miljenovic(a)gmail.com IvanMiljenovic.wordpress.com

2 2

ANNOUNCE: seal-module
by Joachim Breitner 08 Nov '10

08 Nov '10

Hi Cafe, for those who are not following planet haskell: I uploaded a module providing a template haskell solution to the configuration problem to hackage: http://hackage.haskell.org/package/seal-module It turns the style you prefer to write in (pure code that uses configuration values as if they were global constants) into the style that is semantically correct (pure code with configuration values as an additional parameter). The module is well documented, so as soon as hackage has built the documentation you can read how it works on http://hackage.haskell.org/packages/archive/seal-module/0.1.0.1/doc/html/La… BTW: Is there a way to avoid having to indent a declarations passed to a top-level splice? Currently, one has to write: splice [d| some = declaration |] but splice [d| some = declaration |] would be nicer. Greetings, Joachim -- Joachim "nomeata" Breitner mail: mail(a)joachim-breitner.de | ICQ# 74513189 | GPG-Key: 4743206C JID: nomeata(a)joachim-breitner.de | http://www.joachim-breitner.de/ Debian Developer: nomeata(a)debian.org

1 0