[Haskell-cafe] following up on space leak
Good evening, all, following up on my question regarding space leaks, I seem to have stumbled across something very promising. I said I was using this tiny function "lastOrNil" to get the last value in a list, or the empty (scheme) list if the haskell list was empty. The uses of it were all of the form lastOrNil (mapM ) so I wrote a different function mapML to do this directly: > mapML fn lst = mapMLA (List []) fn lst > where mapMLA r _ [] = return r > mapMLA ro fn (x:xs) = >do rn <- fn x > mapMLA rn fn xs This isn't an accumulator, it's a replacer (or, if you like, the "accumulation" is "drop the old one on the floor"), it starts out with the scheme empty list that I want as the default, and it never even builds the list which it'll just dump an instant later. Shazam! Memory usage dropped by roughly an order of magnitude in my little Collatz benchmark, and incidentally runtime improved by 25% or so as well. The horror! :-) Having tasted blood, I will of course be continuing to benchmark... but not tonight. Uwe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal fun [Half-integer]
Antoine Latter wrote: Personally, I've never used "runhaskell Setup sdist" and I've only ever used "cabal sdist". But I'm not sure where I learned that. I think cabal-install is a pretty standard util for people to have, and it ships with the Haskell platform now. So the big hurdle is documentation. Andrew - where does it state that "Setup sdist" is the recommended way of doing this? If it's a wiki you could go and edit it yourself. Start from the Hackage homepage: http://hackage.haskell.org/packages/hackage.html Click on "how to create a Haskell package", takes you to http://www.haskell.org/haskellwiki/How_to_write_a_Haskell_program Section 2.10.1.1 shows you how to create a tarball - using setup sdist rather than cabal sdist. Indeed, I don't think cabal-install is mentioned anywhere. If people seriously want this to become the preferred way to do things, it needs to be much more prominently documented. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] following up on space leak
On Saturday 04 July 2009, Uwe Hollerbach wrote: > Good evening, all, following up on my question regarding space leaks, > I seem to have stumbled across something very promising. I said I was > using this tiny function "lastOrNil" to get the last value in a list, > or the empty (scheme) list if the haskell list was empty. The uses of > it were all of the form > > lastOrNil (mapM ) > > so I wrote a different function mapML to do this directly: > > mapML fn lst = mapMLA (List []) fn lst > > where mapMLA r _ [] = return r > > mapMLA ro fn (x:xs) = > >do rn <- fn x > > mapMLA rn fn xs > > This isn't an accumulator, it's a replacer (or, if you like, the > "accumulation" is "drop the old one on the floor"), it starts out with > the scheme empty list that I want as the default, and it never even > builds the list which it'll just dump an instant later. Shazam! Memory > usage dropped by roughly an order of magnitude in my little Collatz > benchmark, and incidentally runtime improved by 25% or so as well. The > horror! :-) Hi, IMHO expressing mapML using StateT would be a bit cleaner ;) mapML :: (Monad m) => (a -> m List) -> [a] -> m List mapML fn lst = execStateT mapMLAs (List []) where mapMLAs = sequence_ $ map mapMLA lst mapMLA x = (lift $ fn x) >>= put -- Marcin Kosiba signature.asc Description: This is a digitally signed message part. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ORM for haskell?
And I realize that you are not trying to replace RDBs, just building a > nicer interface to them. I am just concerned that some of the nice > properties are lost in the process. I think my main concern comes from > seeing people create databases, by automatically generating tables from > OO-classes. They invariably ends up with something not nearly as nice, > as if they had constructed the database in a more traditional fashion. > for web applications in the Internet, due to security reasons, 99% of the databases are handled exlusively by te web application. This increases the arguments in favor of spending less time in database design. Moreover, since there are no concurrent updates from different applications, (communication with other applications are done trough the middle tier of the web application rather than trough the database), the database just provides transaction coherence (for the single application) and storage. Then it is much faster to perform transactions in the application trough STM and leave the database for storage purposes. At this time the database can be substituted with advantage by files. All of this gives credit to ORM solutions and HappStack or SQLalchemy, Database design and maintenance don't worth the pain in this scenario. My package TCache http://hackage.haskell.org/package/TCache is made also around this philosophy. If the ORM has an interface such is SQLalchemy, it would be nice to have two "drivers" one for pure SQL databases, where all the primitives would be executed in the database, and other pure Haskell where the primitives are executed in memory. For example, the transactions would be executed trough STM. This driver would have configurable persistence (either in files, manualy designed databases or whatever) . Perhaps mixed drivers can be added later. This would unite the best of both worlds. It would be flexible enough to permit the change of scenario without breaking the code. This would be nice for prototyping for example. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] ANN: HalfInteger-1.1.1
Andrew Coppin wrote: I just wrote a small module for dealing with half-integers. (That is, any number I/2 where I is an integer. Note that the set of integers is a subset of this; Wikipedia seems to reserve "half-integer" for such numbers that are *not* integers.) Now, the question is... Is this useful enough to be worth putting on Hackage? It's on Hackage: http://hackage.haskell.org/package/AC-HalfInteger-1.1.1 It'll be interesting to see if I uploaded it right... o_O ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: HalfInteger-1.1.1
On Sat, Jul 04, 2009 at 12:37:21PM +0100, Andrew Coppin wrote: > It's on Hackage: > > http://hackage.haskell.org/package/AC-HalfInteger-1.1.1 > > It'll be interesting to see if I uploaded it right... o_O I guess you did, congrats! :) ...on the Haddock comments of halve and double, it should be "halve . double" and not "half . double", right? You'll see that after the first upload it is easy to do another one (hopefully). -- Felipe. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: HalfInteger-1.1.1
Felipe Lessa wrote: On Sat, Jul 04, 2009 at 12:37:21PM +0100, Andrew Coppin wrote: It's on Hackage: http://hackage.haskell.org/package/AC-HalfInteger-1.1.1 It'll be interesting to see if I uploaded it right... o_O I guess you did, congrats! :) ...on the Haddock comments of halve and double, it should be "halve . double" and not "half . double", right? You'll see that after the first upload it is easy to do another one (hopefully). Well that's a good start. Already somebody has found a bug to be fixed. And in such a trivial package... o_O Still, at least it's an easy fix. ;-) I don't think I'll bother making another upload just for this; the fix is applied to my local copy, but I'll wait until I have something more substantial to upload for the next version. "Thanks for the report." ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Monoid wants a (++) equivalent
On Jul 4, 2009, at 01:17 , Jason Dusek wrote: What is the proper name for the operation on functions of a functor, anyway? The name `fmap` seems to driven by an analogy with `map`. (.) -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu electrical and computer engineering, carnegie mellon universityKF8NH PGP.sig Description: This is a digitally signed message part ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Small Haddock question
This is irritating me now... Suppose I have something like the following: zero = 0 :: Int one = 1 :: Int two = 2 :: Int three = 3 :: Int How do I add Haddock comments to the end of each line? For some reason, Haddock doesn't like either of zero = 0 :: Int -- | Zero zero = 0 :: Int -- ^ Zero Either way it whinges about parse errors. How do I make it shut up and stop being so dense at me? :-} It's damned obvious what I want it to do... ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cont, ContT and IO()
> process [] = return () > process (file:files) = do x <- doit file > if x>0 then process files > else return () Or use a fold: > process' = foldl op True files > op True file = doit file > op False _ = False ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cont, ContT and IO()
>> process' = foldl op True files >> op True file = doit file >> op False _ = False Please pardon me. 'doit' should surely be able to do some IO: > import Data.Foldable > import System.IO > process' = foldlM op True files > op True file = doit file > op False _ = return False were DoIt has the type FilePath -> IO Bool ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re[2]: [Haskell-cafe] Cont, ContT and IO()
Hello Matthias, Saturday, July 4, 2009, 6:39:30 PM, you wrote: > Or use a fold: >> process' = foldl op True files >> op True file = doit file >> op False _ = False foldM, probably, otherwise you will need to execute all actions before running fold -- Best regards, Bulatmailto:bulat.zigans...@gmail.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Re: Cont, ContT and IO() - Code on hpaste
Hi, I've put the code that I wish to transform from using exceptions to using continuations on hpaste: ?http://hpaste.org/fastcgi/hpaste.fcgi/view?id=6515#a6515 thanks Günther Am 04.07.2009, 01:11 Uhr, schrieb Günther Schmidt : Hi, I've got an IO action, some file system IO, traversing one level only and iterating over files found. I wish to build in an "early" exit, ie. if an IO action in the loop encounters a particular value I want it to abort the loop. Now so far, pls don't shoot, I have done this by throwing IO Exceptions and catching them. I'm trying to rewrite this using Continuatios / callCC but can't figure out where to place what. I certainly don't have the intuition yet and funny enough not even in RWH I could find some Cont/ContT examples. Would someone please draw me an example? Günther -- Erstellt mit Operas revolutionärem E-Mail-Modul: http://www.opera.com/mail/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: Cont, ContT and IO() - Code on hpaste
Hi Günther, here is a solution with the Maybe Monad: http://hpaste.org/fastcgi/hpaste.fcgi/view?id=6515#a6515 Matthias. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: Cont, ContT and IO() - Code on hpaste
P.S. See http://en.wikibooks.org/wiki/Haskell/Monad_transformers for some documentation. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: Cont, ContT and IO() - Code on hpaste
P.P.S. Strange it does not seem to work with the paste. So here comes
the solution by mail:
module Consolidator.BusinessLogic.ConflictsResolved
(consolidateDuplicates) where
import System.FilePath
import System.Directory
import Control.Monad (filterM)
import Control.Exception (throwIO)
import System.Environment
import Data.Maybe
import Control.Monad
import Control.Monad.Trans
newtype MaybeT m a = MaybeT { runMaybeT :: m (Maybe a) }
instance (Monad m) => Monad (MaybeT m) where
(>>=) tmb_v f =
MaybeT (runMaybeT tmb_v
>>= \b_v -> case b_v of
Nothing -> return Nothing
Just v -> runMaybeT $ f v
)
return = MaybeT . return . return
instance MonadTrans MaybeT where
lift mon = MaybeT (mon >>= return . Just)
abort :: String -> MaybeT IO a
abort reason = do lift . putStrLn $ reason
MaybeT (return Nothing)
{-
The traversal is one directory deep only.
I try to find out if every immediate subdirectory contains exactly one
"*.gdr" file,
and collect the path names in a list, sgls.
Afterwards I append the contents of each such file to another file.
I want to abort the whole process as soon as I encounter a directory
that does not
include exactly one *.gdr file.
Currently I'm throwing exceptions but I'd prefer to rewrite this code
to use continuations.
-}
consolidateDuplicates :: FilePath -> MaybeT IO ()
consolidateDuplicates fp
= do dirs <- lift (getDirectoryContents fp)
recs <- lift (filterM doesDirectoryExist $ map (fp ) $
filter (not . flip elem [".", ".."]) dirs)
sgls <- mapM checkForSingle recs
let cpy = fp "Korrigiert.gdr"
lift (copyFile (fp "Konsolidiert.gdr") cpy)
lift (mapM_ (\sgl -> do
str <- readFile sgl
appendFile cpy str) sgls)
checkForSingle :: FilePath -> MaybeT IO FilePath
checkForSingle fp = do
cnt <- lift (getDirectoryContents fp)
let fltr = filter ((== ".gdr") . takeExtension)
case fltr cnt of
[] -> abort ("The directory " ++ fp ++ " is empty")
[f] -> return (fp f)
_ -> abort ("There is more than one file in the directory " ++ fp)
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Haskell Weekly News: Issue 124 - July 4, 2009
--- Haskell Weekly News http://sequence.complete.org/hwn/20090704 Issue 124 - July 04, 2009 --- Welcome to issue 124 of HWN, a newsletter covering developments in the [1]Haskell community. Announcements HLint 1.6. Neil Mitchell [2]announced the release of [3]HLint 1.6, a tool for automatically suggesting improvements to Haskell code. Haskell Implementers Workshop: accepted talks. Simon Marlow [4]announced that the list of talks at the [5]Haskell Implementers Workshop 2009 has now been posted. bloxorz clone. Patai Gergely [6]announced a [7]Haskell clone of the game "bloxorz", written by Viktor Devecseri. Fun with type functions. Simon Peyton-Jones [8]announced that he, Ken Shan, and Oleg have finished Version 2 of their [9]paper "Fun with Type Functions", which gives a programmer's tour of what type functions are and how they are useful. If you have a moment to look at, and wanted to help them improve it, leave comments on the linked wiki page. package Boolean: Generalized booleans. Conal Elliott [10]announced [11]Boolean, a new package for generalized booleans, which provides type classes with generalizations of Boolean values and operations, if-then-else, Eq and Ord. TernaryTrees-0.1.1.1 - An efficient ternary tree implementation of Sets and Maps. Alex Mason [12]announced the release of [13]TernaryTrees, a package that extends Data.Set ad Data.Map with some ternary tree structures, one of the more efficient ways of storing strings in a set. 6.12.1 planning. Simon Marlow [14]announced plans for a release of GHC 6.12.1, sometime around September. If you have the time and inclination to help with any of the listed features, please get involved! regular-0.1. José Pedro Magalhães [15]announced the release of the [16]regular library. Many generic programs require information about the recursive positions of a data type, such as generic fold, generic rewriting, and the Zipper data structure. Regular provides a fixed point view on data which allows these definitions for regular data types. It also serves as the basis for a [17]generic rewriting library. Google Summer of Code Progress updates from participants in the 2008 [18]Google Summer of Code. Haddock improvements. Isaac Dupree has made it easier to generate Haddock documentation [19]for non-exported functions, posted an [20]overview of the issues involved in getting proper cross-package documentation working, and his current [21]plan. EclipseFP. Thomas Ten Cate has done a lot of work on EclipseFP, including some [22]cosmetic updates and getting [23]error reporting to work better. space profiling. Gergely Patai is [24]working on a network protocol for his profiling grapher tool, so that other tools can monitor the profiling information. haskell-src-exts. Niklas Broberg has [25]released [26]haskell-src-exts version 1.0.0! fast darcs. Petr Rockai has completed [27]quite a bit of work on darcs, including a [28]beta release of darcs 2.3. Discussion Monoid wants a (++) equivalent. Bryan O'Sullivan [29]suggested adding a more concise operator to the Monoid class for 'mappend', leading to a long, bike-shed-ish (but hopefully still useful) discussion. Reflections on the ICFP 2009 programming contest. Justin Bailey began a [30]discussion on results and experiences from the ICFP 2009 programming contest. Blog noise [31]Haskell news from the [32]blogosphere. Blog posts from people new to the Haskell community are marked with >>>, be sure to welcome them! * Gergely Patai: [33]Playing and learning. * Ketil Malde: [34]A set of tools for working with 454 sequences. * Sebastian Fischer: [35]FP Overview. * Magnus Therning: [36]Making a choice from a list in Haskell, Vty (part 1). * David Amos: [37]Conjugacy classes, part 1. * Well-Typed.Com: [38]GHC and Windows DLLs. * Manuel M T Chakravarty: [39]Converting typed term representations: from HOAS to de Bruijn.. * >>> Ivan Uemlianin: [40]Haskell: sort and sortBy. * Gregory Collins: [41]Building a website with Haskell, part 3. * Michael Snoyman: [42]Hack sample- chat server. * Luke Palmer: [43]On the By functions. * Magnus Therning: [44]Dataenc finally making it into Debian. * Thomas ten Cate: [45]New build instructions. * Erik de Castro Lopo: [46]Three More for the Debian New Queue. * >>> Yuval Kogman: [47]What Haskell did to my brain. * Greg Bacon: [48]FFI: calling into kernel32.dll. * Greg Bacon: [49]Setting up a simple test with Cabal. * Ketil Malde: [50]Dephd updates. * Bryan O'Sullivan: [51]What's in a
Re: [Haskell-cafe] Small Haddock question
On Sat, Jul 4, 2009 at 08:22, Andrew Coppin wrote: > This is irritating me now... Suppose I have something like the following: > > zero = 0 :: Int > one = 1 :: Int > two = 2 :: Int > three = 3 :: Int > > How do I add Haddock comments to the end of each line? For some reason, > Haddock doesn't like either of > > zero = 0 :: Int -- | Zero > zero = 0 :: Int -- ^ Zero > > Either way it whinges about parse errors. How do I make it shut up and stop > being so dense at me? :-} It's damned obvious what I want it to do... I think top-level definitions can only be commented in one way: -- | Emptiness and void, as an Int. zero = 0 :: Int Denis ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Flow Graphs for Language-independent Code Representation
Hi *^o^*, I'm working on a source code transformation project for numerical automatic differentiation for Fortran and C. I would love to know the best Haskell way/package available today to represent procedural (non-OO) code in a language-independent manner. Any tips or resource, paper references are most welcome! Regards, Cetin Sert ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: HalfInteger-1.1.1
Alexander Dunlap wrote: Couple of suggestions: - You should put an (Integer i) => constraint on the halve function so that it becomes impossible to create invalid HalfIntegers. Right. Currently you can *make* such a HalfInteger. You just won't be able to *do* anything with it afterwards. It would probably be more sensible to just add a constraint there. - The documentation for toHalfInteger is truncated. Also, why can't you make the rounding more predictable? ...wow. OK, I'm looking at the source, and it seems I must have got distracted at that moment or something, because I just 100% didn't actually finish writing the documentation! o_O That's pretty special. ;-) As I *should* have written, the rounding for something like toHalfInteger 0.25 is kind of unpredictable; but if you do toHalfInteger 0.5, the result is *guaranteed* to be exact. All the functions are carefully tuned to work correctly on things which really are half integers, and to give plausible results otherwise. Just don't let your life depend on 0.25 being mapped to exactly a half or to exactly zero... Nice work! Heh, thanks. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] ANN: AC-Vector, AC-Colour and AC-EasyRaster-GTK
OK, so having released AC-HalfInteger, I got slightly carried away and released three other small packages. These are packages that many programs I write all end up using. I'm forever copying these files, so I made them into actual bonafide packages. http://hackage.haskell.org/package/AC-Vector-1.1.1 This provides two types, Vector2 and Vector3, which are unboxed vectors of Doubles, with arithmetic, dot product and cross product, and a few other useful items. http://hackage.haskell.org/package/AC-Colour-1.1.1 This provides two types, Colour and Colour8. Both implement simple RGB colour types with arithmetic. Colour has unboxed Double fields, and Colour8 has unboxed Word8 fields. My usual workflow is to do all the image generation with Colour, and to convert to Colour8 just before the data hits the I/O channels. You can, however, do arithmetic directly on Colour8. (I haven't extensively tested that it works properly though...) http://hackage.haskell.org/package/AC-EasyRaster-GTK-1.1.1 This is a layer over Gtk2hs. As you all probably know, Gtk2hs provides a Cairo binding that makes vector graphics wonderfully easy. However, *bitmapped* graphics is darned tricky. I basically had to sit in the #haskell channel with Duncan for a few hours trying to figure out how the hell to do it. This knowledge is now codified in the above package. Load it up and you don't need to know a thing about GTK; you can just create an ImageBuffer, write some pixels to it (efficiently!), save it to disk or display it on screen. (But you *can* access the underlying GTK+ resources if you wish...) In other news, it appears that the batch job to generate the documentation just ran, so you can view it all online. :-D Comments, suggestions, random flames, etc... [I'm particularly curios to know what Duncan will make of the GTK thing...] ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: AC-Vector, AC-Colour and AC-EasyRaster-GTK
On Sat, Jul 04, 2009 at 06:56:44PM +0100, Andrew Coppin wrote: > http://hackage.haskell.org/package/AC-Colour-1.1.1 Why don't you use colour[1]? [1] http://hackage.haskell.org/package/colour -- Felipe. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: AC-Vector, AC-Colour and AC-EasyRaster-GTK
Felipe Lessa wrote: On Sat, Jul 04, 2009 at 06:56:44PM +0100, Andrew Coppin wrote: http://hackage.haskell.org/package/AC-Colour-1.1.1 Why don't you use colour[1]? [1] http://hackage.haskell.org/package/colour A few reasons: 1. I never knew it existed. ;-) 2. It's mind-blowingly complex. 3. It doesn't appear to provide arithmetic over colours. 4. It's parameterised over the component type; my library is hard-coded to specific types for speed. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: AC-Vector, AC-Colour and AC-EasyRaster-GTK
On Sat, Jul 4, 2009 at 8:38 PM, Andrew Coppin wrote: > A few reasons: > > 1. I never knew it existed. ;-) A good reason. However, it's good to do a quick search over Hackage before uploading (or before writing) so you know what's out there. Also, if you hadn't used an "AC-" prefix, you'd have had a name collision. Is there a particular reason why you want your name in the package name rather than just the author field? > 2. It's mind-blowingly complex. Colour *is* complex. Which is why I'm so glad Russell O'Connor did all the hard work for me :) > 3. It doesn't appear to provide arithmetic over colours. It provides darken, blend and addition (though addition is called mappend rather than (+)). signum, abs and fromInteger don't make a huge amount of sense for colours. > 4. It's parameterised over the component type; my library is hard-coded to > specific types for speed. My feeling would be to trust the specializer until it lets me down. Has it let you down in the past? BTW, the EasyRaster package looks useful. --Max ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: AC-Vector, AC-Colour and AC-EasyRaster-GTK
Max Rabkin wrote: On Sat, Jul 4, 2009 at 8:38 PM, Andrew Coppin wrote: A few reasons: 1. I never knew it existed. ;-) A good reason. However, it's good to do a quick search over Hackage before uploading (or before writing) so you know what's out there. Fair enough. ;-) Also, if you hadn't used an "AC-" prefix, you'd have had a name collision. Is there a particular reason why you want your name in the package name rather than just the author field? Well, for example, there's seemingly half a dozen unrelated packages all called "binary", which is just confusing. I wanted to make sure my packages had unique names. (I mean, so do the existing binary packages, just not very useful ones...) 2. It's mind-blowingly complex. Colour *is* complex. Which is why I'm so glad Russell O'Connor did all the hard work for me :) Well, no, because now I'm going to have to spend a few hours trying to find out what CIE is before I can even use that library. I think really it's just aimed at a different problem. It looks like it's trying to specify actual real-world colours. [It's news to me that this isn't fundamentally impossible...] I'm only trying to specify colours on a computer screen. And as we all know, computer screens aren't calibrated in any way, and the same RGB value looks different on each display. But then, I'm only trying to write a fractal generator, so CIE specifications are somewhat overkill here. ;-) 3. It doesn't appear to provide arithmetic over colours. It provides darken, blend and addition (though addition is called mappend rather than (+)). signum, abs and fromInteger don't make a huge amount of sense for colours. Yeah, I implemented signum and so forth for colours and vectors, but they're not particularly meaningful... [Insert remark here about Haskell's numeric class hierachy.] So mappend gives you colour addition [with the perplexing comments about "gamut", presumably some kind of small mammal?], but there's no subtraction? No multiplication? No linear blending? 4. It's parameterised over the component type; my library is hard-coded to specific types for speed. My feeling would be to trust the specializer until it lets me down. Has it let you down in the past? Heh, my colour library includes a custom floor implementation that talks to the GHC primops directly because calling floor is too slow... [In case that sounds like idle talk, I had a program go from 10 seconds to less than 1 second just by using this function. There's a few tickets about it on the GHC Trac.] BTW, the EasyRaster package looks useful. Well, I'd like to think so... It doesn't do anything you couldn't do yourself if you spend a day or two trying to grok the GTK complexity. But it's much easier to just import some code somebody else has already written. ;-) Certainly when I'm in the middle of trying to build a complicated bit of software, figuring out how to just write a few pixels onto the screen is a low priority. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] following up on space leak
On 7/4/09, Marcin Kosiba wrote: > On Saturday 04 July 2009, Uwe Hollerbach wrote: >> Good evening, all, following up on my question regarding space leaks, >> I seem to have stumbled across something very promising. I said I was >> using this tiny function "lastOrNil" to get the last value in a list, >> or the empty (scheme) list if the haskell list was empty. The uses of >> it were all of the form >> >> lastOrNil (mapM ) >> >> so I wrote a different function mapML to do this directly: >> > mapML fn lst = mapMLA (List []) fn lst >> > where mapMLA r _ [] = return r >> > mapMLA ro fn (x:xs) = >> >do rn <- fn x >> > mapMLA rn fn xs >> >> This isn't an accumulator, it's a replacer (or, if you like, the >> "accumulation" is "drop the old one on the floor"), it starts out with >> the scheme empty list that I want as the default, and it never even >> builds the list which it'll just dump an instant later. Shazam! Memory >> usage dropped by roughly an order of magnitude in my little Collatz >> benchmark, and incidentally runtime improved by 25% or so as well. The >> horror! :-) > > Hi, > IMHO expressing mapML using StateT would be a bit cleaner ;) > > mapML :: (Monad m) => (a -> m List) -> [a] -> m List > mapML fn lst = execStateT mapMLAs (List []) > where > mapMLAs = sequence_ $ map mapMLA lst > mapMLA x = (lift $ fn x) >>= put > > -- > Marcin Kosiba Yeah, I'm sure there are more-elegant ways to write this, I'm still very much a beginner in haskell. I'm just very thrilled by the reduction in memory usage! Uwe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: AC-Vector, AC-Colour and AC-EasyRaster-GTK
On Sat, Jul 4, 2009 at 9:18 PM, Andrew Coppin wrote: >>> 2. It's mind-blowingly complex. >>> >> >> Colour *is* complex. Which is why I'm so glad Russell O'Connor did all >> the hard work for me :) >> > > Well, no, because now I'm going to have to spend a few hours trying to find > out what CIE is before I can even use that library. The sRGB function makes a Colour from RGB (actually sRGB, which is a "standardised" RGB -- basically RGB where the exact frequency and power of each channel is specified -- but you can pretend your monitor's RGB is sRGB. > So mappend gives you colour addition [with the perplexing comments about > "gamut", presumably some kind of small mammal?] The gamut of a device is the range of representable colours (a monitor's gamut looks something like a parabola with a flat base in XYZ space, whereas a printer's is much more complex and variable). This makes sense. If you double a monitor's brightest white, you don't get a colour twice as bright: you get the same colour. > but there's no subtraction? > No multiplication? No linear blending? affineCombo can do subtraction, again with the gamut warning. darken does scalar multiplication; it probably doesn't do componentwise multiplication, which doesn't make much sense if you're trying to work in a coordinate-independent setting, though I admit RGB-multiplication can be handy. > Heh, my colour library includes a custom floor implementation that talks to > the GHC primops directly because calling floor is too slow... > > [In case that sounds like idle talk, I had a program go from 10 seconds to > less than 1 second just by using this function. There's a few tickets about > it on the GHC Trac.] Fair enough. Can your implementation not be turned into a patch? BTW, I'm also working on Haskell fractals. You might be interested in looking at my fractal package (though it's currently undocumented, and has no GUI). --Max ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: AC-Vector, AC-Colour and AC-EasyRaster-GTK
Paul Johnson wrote: Andrew Coppin wrote: Well, no, because now I'm going to have to spend a few hours trying to find out what CIE is before I can even use that library. I think really it's just aimed at a different problem. It looks like it's trying to specify actual real-world colours. [It's news to me that this isn't fundamentally impossible...] I'm only trying to specify colours on a computer screen. And as we all know, computer screens aren't calibrated in any way, and the same RGB value looks different on each display. But then, I'm only trying to write a fractal generator, so CIE specifications are somewhat overkill here. ;-) Your display may not be calibrated, but those used for graphic design certainly are. Indeed. And if you're in any kind of position where you *care* about such things, you should be using color, not AC-Colour. If, however, you just want to throw together pretty pictures, AC-Colour is simpler and easier. Different libraries for slightly different tasks. ;-) On the package naming front: I appreciate your wish to avoid just having another "colour" library. But "AC_Colour" doesn't help much. "Simple_colour" might be better. Mmm, yeah. Naming everything with "AC" precludes name clashes and doesn't require too much thinking. Coming up with a better name requires thinking about what actually makes your package unique. And, of course, if another package comes along, that analysis may change. (E.g., I seem to recall there's a "newbinary" package which has actually been long since superceeded - so not so "new" any more!) If I name my package simple-colour, and then somebody else makes an even simpler one... the name becomes kind of meaningless. (Admitedly there's not too much danger of this happening...) I just like the idea of having definitely unique, distinctive package names. Otherwise I'd have to come up with stuff like geovector and trivicolour and so on... Arguably EasyRaster-GTK should have been a sufficiently unique name by itself though. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Optimizing Compiler for the ICFP 09 contest's VM
The byte code for the virtual machine of this years ICFP specified a language with single assignment per simulation step. Interestingly most memory locations get overwritten each simulation step before they are read. That means, those locations don't have to be remember between steps. Also locations that never get overwritten (e.g. location associated with Noops), are constant. Thus the variables state of the simulation is orders of magnitude smaller than the naive 2^16 * 32 bit + 1 bit. I wrote a small program that analyses the dataflow of a byte code program (and initial memory setup) for the VM. After analyzing my program emits Haskell code to run the given byte code. If anyboby is interested, I can document my program and put it online somewhere. I also made pretty graphs of the dataflow with graphviz. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Optimizing Compiler for the ICFP 09 contest's VM
2009/7/4 Matthias Görgens : > The byte code for the virtual machine of this years ICFP specified a > language with single assignment per simulation step. Interestingly > most memory locations get overwritten each simulation step before they > are read. That means, those locations don't have to be remember > between steps. Also locations that never get overwritten (e.g. > location associated with Noops), are constant. Thus the variables > state of the simulation is orders of magnitude smaller than the naive > 2^16 * 32 bit + 1 bit. > > I wrote a small program that analyses the dataflow of a byte code > program (and initial memory setup) for the VM. After analyzing my > program emits Haskell code to run the given byte code. > > If anyboby is interested, I can document my program and put it online > somewhere. I also made pretty graphs of the dataflow with graphviz. Hi Matthias, it would be nice to blog a little post about what you found and the pretty graphs :) Thu ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ANN: AC-Vector, AC-Colour and AC-EasyRaster-GTK
On Jul 4, 2009, at 15:01 , Max Rabkin wrote: On Sat, Jul 4, 2009 at 8:38 PM, Andrew Coppin wrote: 3. It doesn't appear to provide arithmetic over colours. It provides darken, blend and addition (though addition is called mappend rather than (+)). signum, abs and fromInteger don't make a huge amount of sense for colours. I don't see a good meaning for signum offhand, but fromInteger could take an X11-encoded RGB value and abs could produce grayscale brightness. -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu electrical and computer engineering, carnegie mellon universityKF8NH PGP.sig Description: This is a digitally signed message part ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
