- Haskell - Haskell.org

HList with new records, in darcs.haskell.org
by oleg＠pobox.com 02 Jul '06

02 Jul '06

The HList library of strongly typed heterogenous lists, records, type-indexed products (TIP) and co-products is now accessible via DARCS http://darcs.haskell.org/HList/ The library contains a new representation for open records. Previously (Record.hs) a record is a Scheme-like heterogenous associative list, that is, the list of pairs of field label and the corresponding field value. The HList library uses the type of the label, therefore, the run-time representation of the label is actually `undefined'. Thus, a record built as rec1 = label1 .=. True .*. lab2 .=. 'c' .*. emptyRecord has the run-time representation HCons (undefined,True) (HCons (undefined,'c) HNil) In the new representation, RecordP.hs in the repository, the record is a heterogeneous list of field values. Label _types_ are also bundled in a heterogenous list of types, which annotate the record type as a phantom type. Thus if we write rec2 = label1 .=. True .*. lab2 .=. 'c' .*. emptyRecordP rec2 now has the run-time representation of HCons True (HCons 'c HNil) That is, labels are nowhere to be found in the run-time representation. Field labels become compile-time-only entity. Of course, by making labels Typeable and using the methods of that class it is possible to access the label information at run-time, if so desired. The new representation of the records supports the old record interface; for example, both (rec1 .!. lab2) and (rec2 .!. lab2) evaluate to 'c', the value of the field lab2. RecordP are similar to TIP; however, the values of RecordP are indexed by the label types rather than by their own types. RecordP have an advantage for representing tables -- homogenous collections of rows, which in turn are records of fields (columns): newtype Table ls vs = Table [RecordP ls vs] All records have exactly the same field types in the same sequence. The run-time representation of the table is just the collection of all the field values. The database schema, labels and the column types, is a compiled-time entity and used for checking and compiling of database operations.

1 0

Comments from Brent Fulgham on Haskell and the shootout
by Simon Marlow 02 Jul '06

02 Jul '06

Brent Fulgham has given me permission to share this message that he sent to us about his views on Haskell and the Great Computer Language Shootout. There's some nice advocacy material in here. I'd like to publicly thank Brent for all his work on the shootout - benchmarking 50 or so different language implementations and getting any kind of useful results at all is a significant feat. The shootout results have also influenced our direction in the Haskell community and GHC in particular. Here's the message, for your enjoyment... -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Dear Drs. Hudak and Peyton-Jones: (Feel free to forward these comments to anyone you think would enjoy them). A couple of years ago I revived the dormant "Great Computer Language Shootout" website because I wanted to create an up-to-date comparison of language implementations to use as ammunition against my nay- saying coworkers over the state of the art in Computer Science. Back in those days, I was mainly interested in the performance of Objective Caml and Lisp, but was also interested in a new (to me) language called Haskell . Haskell had (and still has) a beautiful, mathematical notation that was very easy to read, but it was such a poor performer that I despaired of ever getting to use it for anything but simple test programs. While it was conceptually satisfying, pragmatism seemed destined to keep it on the shelf for the foreseeable future. I wanted to write to inform you how shocked I was to see the great advances in performance in the Glorious Haskell Compiler over the last year or so. Of course, we have also benefited from some great contributions by the folks on the Haskell-Cafe mailing list. The thing I find most staggering about the compiler is that its quite brilliant at identifying wasted work. Many of the original shootout test cases were written to test the performance of simple scripting languages, and had grown a set of hacks over the years to keep the run times large enough to measure as processor speeds improved and language implementations became more sophisticated. In many cases, tests involved calculation or large values in loops, which would cause Perl and Python (and even C, Java, and C++) to chug merrily away burning cycles so that I could measure the timings in seconds, rather than microseconds. Then along came GHC, which began achieving enormous wins over the other languages because the lazy nature of the evaluation threw out the unneeded calculations, handing over the final useful result without breaking a sweat. Consequently, we had to revise many of the tests to force the programs to perform real work that could not be avoided, in some cases by forcing intermediate values to be printed in the solutions so we could be sure the work was actually being done. With those changes, Haskell began to fall back in speed to more moderate levels. However, in the last few releases of GHC (and the last several rounds of tweaks by the Haskellers on the Cafe list), GHC has risen to the top of the heap, approaching native C in performance overall. What I find quite appealing about this is that the performance is consistent across platforms. While Java's "HotSpot" compiler makes a good showing on Intel, my AMD test machine yields consistently worse results than the Pentium IV system. However, GHC performs equally well on both systems, yielding a similar fractional performance against C in both cases[1,2]. I think this is great because it means that applications written in Haskell can be reasonably expected to yield good performance on all of the common x86 platforms without customizations. Java and several other languages seem to yield varying performance, probably because of advanced optimizations available on the Intel platform that have been hand-coded into those systems. At any rate, I just wanted to write a quick fan letter about Haskell, and let you know how impressed I am with its performance. Looks like it's time for me to start taking a serious look at Haskell for some real-world problems at work now! Best regards, Brent Fulgham http://shootout.alioth.debian.org/ [1] "http://shootout.alioth.debian.org/gp4/benchmark.php? test=all&lang=all" [2] "http://shootout.alioth.debian.org/debian/benchmark.php? test=all&lang=all" -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2.2 (Darwin) iD8DBQFEnLtkzGDdrzfvUpURAhRBAKCHWiIAJVONyKhDQgG/rBvSmGGDWgCeIufW G2nx/o1ruiXT5Yn9EcNxsFk= =U7WI -----END PGP SIGNATURE-----

5 4

Re: ANNOUNCE: HNOP 0.1
by Paul Johnson 01 Jul '06

01 Jul '06

> One interesting line of development would be to spin off the core > functionality into a separate library, to provide no-op services to > other Haskell applications. Actually this would be useful, and probably belongs in Control.Monad (although I'd call it "noOp" rather than "noop"). I was recently working on code that took monadic actions as parameters, so that under the right circumstances the appropriate action would be executed. On several occasions I had to pass "return ()" as a null action. Having an explicit "noOp :: Monad m => m ()" would have made the code shorter and clearer. Paul.

1 0

Re: ANNOUNCE: HNOP 0.1
by Paul Johnson 01 Jul '06

01 Jul '06

> Now we just need the Godot version, in which nothing happens twice. Oops. That should have been "... in which nothing happens, twice." Alternatively I just need a type checker for my English. Paul.

1 0

ANNOUNCE: HNOP 0.1
by Ashley Yakeley 01 Jul '06

01 Jul '06

HNOP: Haskell No Operation A first version of HNOP 0.1 is now available under a simple permissive license. This version should be considered "beta" quality, though I don't know of any bugs. http://semantic.org/hnop.tar.gz HNOP does nothing. Here's a sample session to illustrate: $ ./hnop $ The code is written entirely in plain Haskell 98 and makes no use of FFI or impure functions. The source is available in a darcs repository: darcs get http://semantic.org/hnop/ Possible applications include generated code size comparison for compilers, and as a starting point for more complex Haskell projects. -- Ashley Yakeley Seattle, WA

12 17

Re: ANNOUNCE: HNOP 0.1
by Paul Johnson 30 Jun '06

30 Jun '06

Now we just need the Godot version, in which nothing happens twice. Or are Godots considered harmful? Paul.

1 0

NEW: hpodder
by John Goerzen 30 Jun '06

30 Jun '06

Hello everyone, I'm pleased to announce the first release of hpodder. hpodder is a podcast downloader (podcatcher) written in pure Haskell. I wrote it because I was unsatisfied with the other podcatchers for Linux. I am using hpodder for my own purposes already. hpodder homepage: http://quux.org/devel/hpodder documentation: http://darcs.complete.org/hpodder/doc/hpodder.pdf darcs repo: http://darcs.complete.org/hpoder feature list: · Convenient, easy to learn, and fast command-line interface (it's sim- ple to do simple things, and advanced things are possible) · Automatic discovery of feed metadata such as title · Full history database for accurate prevention of duplicate downloads and tracking of new episodes · Conversion tools to convert your existing feed list and history from other applications to hpodder. Supported applications and formats include: castpodder and ipodder. · Most operations can work fully automatically across your entire pod- cast database, or they can work manually as well. · Automatic updating of ID3 (v1 and v2) tags based on metadata in the podcast itself. This important feature is available through iTunes but is often missed by other podcatchers. · hpodder operations can be easily scripted or scheduled using regular operating system tools. · Fully customizable naming scheme for downloaded episodes, including a name collision detection and workaround algorithm. · Automatic support for appending .mp3 extensions to MP3 files that lack it. · Numerous database and history inquiry tools · Small, minimalist footprint · Power users and developers can interact directly with the embedded Sqlite3 database used by hpodder. The database has a simple schema that is developer-friendly. · Support for resuming interrupted downloads of podcasts · hpodder is SAFE and is designed with data integrity in mind from the beginning. It should be exceedingly difficult to lose a podcast episode, even in the event of a power failure.

1 1

Speed of ByteString.Lazy
by Chad Scherrer 29 Jun '06

29 Jun '06

I have a bunch of data files where each line represents a data point. It's nice to be able to quickly tell how many data points I have. I had been using wc, like this: % cat *.txt | /usr/bin/time wc 2350570 4701140 49149973 5.81user 0.03system 0:06.08elapsed 95%CPU (0avgtext+0avgdata 0maxresident)k 0inputs+0outputs (152major+18minor)pagefaults 0swaps I only really care about the line count and the time it takes. For larger data sets, I was getting tired of waiting for wc, and I wondered whether ByteString.Lazy could help me do better. So I wrote a 2-liner: import qualified Data.ByteString.Lazy.Char8 as L main = L.getContents >>= print . L.count '\n' ... and compiled this as "lc". It doesn't get much simpler than that. How does it perform? % cat *.txt | /usr/bin/time lc 2350570 0.09user 0.13system 0:00.24elapsed 89%CPU (0avgtext+0avgdata 0maxresident)k 0inputs+0outputs (199major+211minor)pagefaults 0swaps Wow. 64 times as fast for this run, with almost no effort on my part. Granted, wc is doing more work, but the number of words and characters aren't interesting to me in this case, anyway. I can't imagine (implementation time)*(execution time) being much shorter. Thanks, Don! -- Chad Scherrer "Time flies like an arrow; fruit flies like a banana" -- Groucho Marx

4 5

Forcing Type Class Equality
by john lask 29 Jun '06

29 Jun '06

** this is literate haskell hoping someone can help me. What I am trying to do is >class Ba1 a >ba1 :: Ba1 a => a -> IO () >ba1 x = print "helllo" what I wish to do is declare another function >class Foo a >proxy :: Foo a => a -> IO () so that I can do something like >proxy x = ba1 x this does not work in hugs and gives ERROR "test.lhs":19 - Cannot justify constraints in explicitly typed binding *** Expression : proxy *** Type : Foo a => a -> IO () *** Given context : Foo a *** Constraints : Ba1 a which is entirely expected. The trick is to tell hugs that Foo and Ba1 are the same so how do I do that ? I have tried >instance Ba1 a => Foo a >instance Foo a => Ba1 a which should define equality but to no avail ?? Can this be done ? I thought the problem might be drawing the inference Foo == Ba1 and to try and force the inference i did: >class Foo a >proxy1 :: (Foo a,Ba1 a) => a -> IO () >proxy1 x = ba1 x >proxy2 :: Foo a => a -> IO () >proxy2 x = proxy x This typed checked but Main> :load test Main> proxy2 ERROR - *** The type checker has reached the cutoff limit while trying to *** determine whether: *** Foo a *** can be deduced from: *** () *** This may indicate that the problem is undecidable. However, *** you may still try to increase the cutoff limit using the -c *** option and then try again. (The current setting is -c40) any ideas out there ?? _________________________________________________________________ Meet Sexy Singles today @ Lavalife - Click here http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Flavalife9%2Eninemsn%2Ecom%2…

2 2

ANN: MissingH 0.14.4
by John Goerzen 28 Jun '06

28 Jun '06

I am pleased to announce a new version of MissingH. Changes since the last announcement include: * Fix a bug in ConfigParser.merge and write a test for it. * Fix cabal incompatibilies with GHC 6.4.2 * IO/StatCompat: Fixed missing Data.Bits import on mingw32 * Updated COPYING file with new FSF address * Cmd.hs: Commented out code that was never used * Import new package MD5 0.2.7 from Ian Lynagh * ConfigParser gives more helpful error messages when possible -- it now includes section and option in most error messages * Unit tests updated for new ConfigParser output * MissingH.Str: New function escapeRe that takes a string and makes a regular expression that matches the string literally. * MissingH.IO.HVFS: New function vDoesExist * MissingH.List: New function hasAny * New modules: MissingH.Path.Glob and MissingH.Path.WildMatch and tests for them * 716 total unit tests now MissingH is available for download from http://quux.org/devel/missingh It is also included with several Linux and BSD distributions.

1 0