June 2008 - Haskell-Cafe

Re: [Haskell-cafe] Lambda and closures in PHP -- could someone please comment?
by PR Stanley 18 Jun '08

18 Jun '08

> > Now, if you, Jules, Alex or some other wannabe Hitler have a problem with > > my freedom of expression then your best solution is to saddle up > and get the > > hell out yourselves. This is the wrong place for setting up your tinpot > > dictatorship, Doctor! > >Just more evidence... > >http://en.wikipedia.org/wiki/Godwin's_law > >Paul: I think you're missing a big painfully obvious point here, Mr >smarty pants. No, actually, I think I'll leave it at that. I just >can't be bothered. However, I'll say this - anyone sticking his >bloody big hooter into my business in the future will get a similar reaction! Paul >Luke

1 0

Compiling large code with old machine
by Samuel Silva 18 Jun '08

18 Jun '08

Hello I'm using GHC to compile around 700K of Haskell Code generated by HaXml. How I compile this code. My machine is Windows-XP(512MB RAM, 1.5GHz) running GHC-6.8.2. How much time it spend to compile this file? I spent more than 1 hour and it doesn't finished. What flags make compiling fast? I try with -H500m but dont't expect. It is possible GHC running endless cycle at compilation stage? Thanks. -- Don't hug that, Hugs ME! Samuel Silva

6 6

ANNOUNCE: cabal-install 0.5
by Duncan Coutts 18 Jun '08

18 Jun '08

cabal-install 0.5 ================= cabal-install version 0.5 is out: http://haskell.org/cabal/download.html or get it from hackage: http://hackage.haskell.org/cgi-bin/hackage-scripts/package/cabal-install If you are already using a cabal-install pre-release then you can just: $ cabal update $ cabal install cabal-install New features in cabal-install ============================= Command line improvements ------------------------- The most immediately noticeable thing is that the command line interface now has all the commands that runhaskell Setup.hs has. Of course it still has the features to download and install packages from hackage. It also gained an upload command. So it now provides a command line interface to the whole Cabal/Hackage system. There’s no need to use runhaskell Setup.hs ever again. There is also bash command line completion support included which I find is a great time saver. Installing and upgrading ------------------------ The next big thing is that it includes a new package dependency resolution system that finds correct and sensible solutions more of the time and has better default behaviour. The new behaviour should be similar to other package managers that people are used to. For example, suppose you’ve got xmonad-0.5 installed and version 0.7 is the latest on hackage, then $ cabal install xmonad will install xmonad-0.7. The older version, xmonad-0.5, will remain installed. The behaviour of install is to upgrade as little as possible to satisfy your request, but sometimes you want to upgrade all the dependencies too. Supposing now that we have xmonad-0.7 installed, but we’re still using X11-1.4.1 and the latest version on hackage is X11-1.4.2, then $ cabal upgrade xmonad will install X11-1.4.2 and *re-install* xmonad-0.7, this time built against the newer X11-1.4.2. So in general, the install command will install the latest version of things but will try and use any existing installed versions of dependencies while the upgrade command will also try to use the latest versions of all dependencies. As a special case, cabal upgrade on its own will try to upgrade all the packages that you have installed. For both command there is a --dry-run flag so you can see what would be installed without actually doing it. Hugs ---- Yes, it even works with hugs. That is, cabal-install built by ghc can manage the installation of packages for hugs. In principle cabal-install should be able to be run by hugs but currently the zlib binding is using a function that hugs does not support. Note that for hugs, Cabal does not know what packages are already installed because there is no equivalent of the package database that ghc has. So that means cabal-install cannot do very sensible installation planning. It should work ok so long as all the dependencies are already installed. Windows ------- Yes, it even works on Windows. The one caveat is that cabal-install cannot currently upgrade itself because Windows makes it hard for a process to overwrite its own executable file. It needs more complex trickery with the Win32 API. In the meantime the workaround is to rename the cabal.exe file first eg to cabal-foo.exe, then run cabal-foo install cabal-install. Build reporting --------------- One feature that made it into this release is build reporting. cabal-install keeps logs of all packages that you install (at least packages from hackage, not local ones). It records a bit of information about each package, in particular whether the outcome was successful or not. You can see these build reports in ~/.cabal/packages/$server/build-reports.log. For example, there is one from my machine for xmonad: package: xmonad-0.7 os: linux arch: x86_64 compiler: ghc-6.8.2 client: cabal-install-0.5.1 flags: -testing small_base dependencies: X11-1.4.2 base-3.0.1.0 containers-0.1.0.1 directory-1.0.0.0 mtl-1.1.0.0 process-1.0.0.0 unix-2.3.0.0 install-outcome: InstallOk docs-outcome: NotTried tests-outcome: NotTried The plan in the longer term is to let people upload these build reports to hackage so we can get a wider range of testing data about the packages on hackage. Duncan

1 0

Re: [Haskell-cafe] Compiling large code with old machine
by Samuel Silva 18 Jun '08

18 Jun '08

Hello On Tue, Jun 17, 2008 at 10:24 PM, Niels Aan de Brugh <nielsadb(a)gmail.com> wrote: > On Tue, 2008-06-17 at 13:19 +0100, Samuel Silva wrote: >> >> I'm using GHC to compile around 700K of Haskell Code generated by HaXml. >> How I compile this code. >> My machine is Windows-XP(512MB RAM, 1.5GHz) running GHC-6.8.2. >> >> How much time it spend to compile this file? >> I spent more than 1 hour and it doesn't finished. > > Just out of curiosity, is all that 700K just one big file or did you > split it up into multiple files? It's just two big files with 700K each generated by DtdToHaskell from two big DTD files around 70K each. > Are you using any "expensive" features? I'm not a GHC exert, but I'd > expect things like type manipulations (e.g. phantom types) to be > expensive (please correct me if this assumption is wrong). This large module only contains datatype declarations and instance classes for HTypeable, XmlContent, XmlAttributes and XmlAttrType. > Bottom line: if compiling is so slow consider changing the code > generator so the result is easier to process by GHC instead of fiddling > with command line switches. Since you've chosen to generate the code it > will probably change often and you'll run into this problem more than > once. > > Regards, > Niels > > Each Haskell file have almost 15000 lines. It is huge. It's compiling very fast(some minutes) using -H300m without cabal and -H200m with cabal. GHC it is amazing. This was a valuable help. -- Don't hug that, Hugs ME! Samuel Silva

1 0

ANN: Mueval 0.3
by Gwern Branwen 17 Jun '08

17 Jun '08

Hiya everyone. So I've uploaded Mueval 0.3 (release early, release often) <http://hackage.haskell.org/cgi-bin/hackage-scripts/package/mueval-0.3>. What's new? Well, I changed the printing output to be cleaner, and I made printing the inferred type optional (through a --print-type flag). In addition, I've made sure it can run (for me, at least) under -N4, and the watchdog threads should be sturdier. (Plus extra tests.) However, I've been hearing that a number of people (3 at last count) have been having a lot of trouble with the ResourceLimits stuff. So, I have a few requests of people who install mueval: * First, does it work out of the box, with no tweaking of the resource limits? * Second, if it doesn't work out of the box, which resource limits did you need to disable or increase? * Thirdly, if you changed the resource limits, could you run ./build.sh and tell me whether all the tests pass? Finally, I've been pondering removing the rlimits entirely since they are both very unportable, empirically derived, and apparently don't even work consistently on Linux. Any thoughts? -- gwern PABX MEMEX 868 1* TSCI @ shelter Yukon el ASLET

1 0

Polymorphic QuickCheck arguments
by Hugo Pacheco 17 Jun '08

17 Jun '08

Hi all, There is something about polymorphic tests in QuickCheck that I do not understand. If you write the simplest dummy test function tst :: a -> Bool tst _ = True and evaluate it we get > verboseCheck tst 0: () 1: () ... How come did the polymorphic value a get instanciated to ()? Is this done via the Testable type class? Could someone please explain this to me? I'm not saying it does not make sense, but I would like to understand how it works. Being possible to change this behavior, this is, assign another "default type" whenever polymorphic types occur, would be awesome. Cheers, hugo -- www.di.uminho.pt/~hpacheco

3 4

send in material for the HWN!
by Brent Yorgey 17 Jun '08

17 Jun '08

Hi all, Do you know of a blog post about Haskell that wasn't syndicated on Planet Haskell? Do you have a Haskell-related story to tell or announcement to make? Do you have a link to some great pictures from the latest meetup/hackathon/talk/conference? A funny story about Don Stewart? If so, consider submitting it to the Haskell Weekly News. The HWN isn't really about Haskell so much as it is about the Haskell *community* -- and that means you! In addition to the above sorts of submissions, I am also thinking of adding a "Community News" section with announcements of (not-necessarily-Haskell-related) major events in the lives of community members -- but obviously this depends on submissions, so if you get a new job/move/start a new company/graduate/increment your age/have a comet named after you/bring forth a new life (aka future Haskeller) into the world, please share, and let the community celebrate with you! Submissions can be sent to byorgey at seas dot upenn dot edu. Be sure to include "HWN" in the subject line. For now I'm planning to put an issue out every Wednesday afternoon, so the submission deadline is Wednesday at 11am GMT (6am Eastern US). -Brent

1 0

How to do this in FP way?
by Magicloud Magiclouds 17 Jun '08

17 Jun '08

Hello, I am getting familiar with FP now, and I have a "program design" kind of question. Say I have something like this in C: static int old; int diff (int now) { /* this would be called once a second */ int ret = now - old; old = now; return ret; } Because there is no "variable" in Haskell. So how to do this in a FP way? Thanks.

15 23

Suggestion for implementation of vector library
by Olex P 17 Jun '08

17 Jun '08

Hi guys, Sorry for a silly questions but I didn't find a proper answer in Google. I've started to learn Haskell and would like to implement a library for work with vectors. I found different implementations of this stuff but all of them made just for fun in a short as possible terms using lists or tuples. I suppose there should be a better way to go. First idea came to my mind was using of classes. Something like this: [code] -- Template Vector class class Vector v where (<+>) :: v -> v -> v (<->) :: v -> v -> v (<*>) :: v -> v -> v (*>) :: v -> Float -> v -- other methods here -- Vector3 instance -- Declare new Vector3 type data Vector3 = Vector3 (Float, Float, Float) instance Vector Vector3 where (<+>) (Vector3 (x1, y1, z1)) (Vector3 (x2, y2, z2)) = Vector3 (x1 + x2, y1 + y2, z1 + z2) (<->) (Vector3 (x1, y1, z1)) (Vector3 (x2, y2, z2)) = Vector3 (x1 - x2, y1 - y2, z1 - z2) (<*>) (Vector3 (x1, y1, z1)) (Vector3 (x2, y2, z2)) = Vector3 (x1 * x2, y1 * y2, z1 * z2) (*>) (Vector3 (x, y, z)) f = Vector3 (x * f, y * f, z * f) length (Vector3 (x, y, z)) = sqrt (x * x + y * y + z * z) -- the rest of methods [/code] What I don't like here is using of data type constructors when even simple expression like v1 + v2 becomes too long (<+>) (Vector3 (1,2,3)) (Vector3 (4,5,6)) Do I really need a data type constructor (in this particular case)? Or better to declare a vector3 type as: type Vector3 = (Float, Float, Float)? Next question is how to make one instance which derives from several classes? For example from Eq, Num and Vector (if I want to overload +, -, <, >, == etc.) Is it good idea to use classes and ad hoc polymorphism or better to use parametric polymorphism? How you would implement such library? (not just for educational purposes but for development of _very_good_extensible_ software) Thank you, Alex.

5 5

1/0
by Evan Laforge 17 Jun '08

17 Jun '08

So, I know this has been discussed before, but: > 1/0 Infinity > 0/0 NaN ... so I see from the archives that Infinity is mandated by ieee754 even though my intuition says both should be NaN. Every other language throws an exception, even C will crash the program, so I'm guessing it's telling the processor / OS to turn these into signals, while GHC is turning that off. Or something. But then what about this note in Control.Exception: (NOTE: GHC currently does not throw ArithExceptions except for DivideByZero) Doesn't this imply GHC should be throwing DivideByZero? Why is it in the stdlib if it's turned off? And why would it be turned off? I suppose I could accept the above since Infinity *is* an exceptional value meant to represent things like 1/0, and it can be nice to see NaNs in your output instead of crashing everything (this is, I assume, why ieee754 has these modes in the first place), but how about this: > round (0/0) :: Integer (huge negative number) Ok, so integral types don't have that exceptional value. Shouldn't trying to convert NaN or Infinity to an Integral throw something? Is it a performance thing? I'd think if you're converting to Integer you don't really need hardware level performance anyway, so a couple of checks wouldn't kill anyone.

10 18