On Aug 17, 2010, at 12:33 AM, Jason Dagit wrote:
So next I would use heap profiling to find out where and what type
of data the calculation is using.
I did.
I would do heap profiling and look at the types.
All retained data is of type ARR_WORDS. Retainer profiling shows that
the majorit
On Tue, Aug 17, 2010 at 3:53 AM, Richard O'Keefe wrote:
> On Aug 17, 2010, at 12:37 AM, Roel van Dijk wrote:
>>
>> phi = (1 + sqrt 5) / 2
>> fib n = ((phi ** n) - (1 - phi) ** n) / sqrt 5
>>
>> The use of (**) should make the complexity at least O(n). Please
>> correct me if I'm wrong (or sloppy).
Hello all,
Does GHC expose any primitives for things like atomic compare-and-swap?
I can't seem to find anything in the docs. I'm wondering if it's
possible, for example, to implement things like the wait-free concurrent
queue from [1] or a lock-free wait-free hash table like the Azul people
are d
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 8/17/10 00:11 , Ivan Lazar Miljenovic wrote:
> On 17 August 2010 12:23, Jens Petersen wrote:
>> I may be missing something but wouldn't it be less confusing and simpler
>> just to have now one single unique tarball (ie darcs-beta-2.4.98.3.tar.gz)?
On 17 August 2010 12:23, Jens Petersen wrote:
> Thanks for releasing the betas. :)
>
> I checked that
> http://hackage.haskell.org/packages/archive/darcs-beta/2.4.98.3/darcs-beta-2.4.98.3.tar.gz
> and
> http://darcs.net/releases/darcs-2.4.98.3.tar.gz
> again seem to be identical for all purposes a
Thanks for releasing the betas. :)
I checked that
http://hackage.haskell.org/packages/archive/darcs-beta/2.4.98.3/darcs-beta-2.4.98.3.tar.gz
and
http://darcs.net/releases/darcs-2.4.98.3.tar.gz
again seem to be identical for all purposes and yet actually different tarballs.
The same was true for d
On Aug 17, 2010, at 12:37 AM, Roel van Dijk wrote:
>
> phi = (1 + sqrt 5) / 2
> fib n = ((phi ** n) - (1 - phi) ** n) / sqrt 5
>
> The use of (**) should make the complexity at least O(n). Please
> correct me if I'm wrong (or sloppy).
Using the classic
x**0 = 1
x**1 = x
I've tested haskell on CJK system. For CJK characters, the IO can work great
on file, but awfully on console. I think it is the matter from Winodws.
We can not set utf8 on Windows like linux, instead of that they use code
page. It looks that greek characters doesnot support on your system. You
shal
On 16.08.10 14:44, Daniel Fischer wrote:
Hi Bulat,
On Monday 16 August 2010 07:35:44, Bulat Ziganshin wrote:
Hello Daniel,
Sunday, August 15, 2010, 10:39:24 PM, you wrote:
That's great. If that performance difference is a show stopper, one
shouldn't go higher-level than C anyway :)
*all* spe
There's a Fibonacci Heap: http://en.wikipedia.org/wiki/Fibonacci_heap
Not sure what else though :)
On Mon, Aug 16, 2010 at 11:14 PM, Antoine Latter wrote:
> On Mon, Aug 16, 2010 at 1:37 PM, Andrew Coppin
> wrote:
> >
> > This neatly leads us back to my second assertion: In all my years of
> >
On Mon, Aug 16, 2010 at 9:00 AM, Sebastian Fischer <
s...@informatik.uni-kiel.de> wrote:
> [CC-ing café again]
>
>
> On Aug 16, 2010, at 5:52 PM, Daniel Fischer wrote:
>
> I am a bit concerned about the memory usage.
>>>
>>
>> Of your implementation of the matrix power algorithm?
>>
>
> Yes.
>
>
On Mon, Aug 16, 2010 at 1:37 PM, Andrew Coppin
wrote:
>
> This neatly leads us back to my second assertion: In all my years of
> computer programming, I've never seen one single program that actually
> *needs* the Fibonacci numbers in the first place (let alone in
> arbitrary-precision).
>
I thin
Hi,
according to this page:
http://hackage.haskell.org/package/pango-0.11.0
pango should work with cairo 0.11.0 (or any other 0.11.* version)
So it seems that the problem is that cabal tried to use pango-0.11.1,
and I am guessing that it does not backtrack and try an older version
if a build fails
Don Stewart wrote:
So that's pretty simple.
This is one of those "2 > 3 for sufficiently large 2" things, eh? ;-)
'sifflet' requires cairo ==0.11.0 and pango.
Indeed.
But since
pango doesn't work with cairo 0.11.0, the package can't be built.
Uh... it seemed to build just fine for me (w
Kevin Jardine wrote:
The latest version of the Haskell Platform is Haskell Platform
2010.2.0.0.
However, even with the latest version,
cabal install cabal-install
installs cabal in the wrong place (not in extralibs/bin) under Windows
at least so it is impossible to upgrade cabal.
Well, th
Don Stewart wrote:
Can I ask why you chose 2010.1.0.0 ?
That's easy; the installer is already on my harddrive.
I hope the download site for the HP isn't leading people astray.
Nope. I just hadn't noticed that what I downloaded a few months ago was
no longer the latest, that's all.
> On Aug 16, 9:27 pm, Andrew Coppin wrote:
> > And so today, just for giggles, I tried to get Sifflet to work. Along
> > the way, I encountered a number of... "glitches", if you will.
> >
> > First of all, I tried to get it to work on Windows. I fired up a new
> > Windows VM and installed Haskell
The latest version of the Haskell Platform is Haskell Platform
2010.2.0.0.
However, even with the latest version,
cabal install cabal-install
installs cabal in the wrong place (not in extralibs/bin) under Windows
at least so it is impossible to upgrade cabal.
Having said that perhaps it is for
$ cabal update
Downloading the latest package list from hackage.haskell.org
$ cabal install sifflet
Resolving dependencies...
cabal: cannot configure pango-0.11.1. It requires cairo >=0.11.1 && <0.12
and glib >=0.11.1 && <0.12
For the dependency on cairo >=0.11.1 && <0.12
And so today, just for giggles, I tried to get Sifflet to work. Along
the way, I encountered a number of... "glitches", if you will.
First of all, I tried to get it to work on Windows. I fired up a new
Windows VM and installed Haskell Platform 2010.1.0.0. It seems that
(finally) this includes
Using the latest version (WinGHCi 1.0.6) and GHCi 6.12.3, I get:
*
Prelude>* putStrLn "ß"
ß
*
Prelude>* print "ß"
"\223"
*
Prelude>* putStr "αγδ"
*** Exception: : hPutChar: invalid argument (character is not in the
code page)
and this is consistent with output I get when using GHCi on a con
Roel van Dijk wrote:
On Sat, Aug 14, 2010 at 5:41 PM, Andrew Coppin
wrote:
(Then again, the Fibonacci numbers can be computed
in O(1) time, and nobody ever needs Fibonacci numbers in the first place, so
this is obviously example code.)
A bit off-topic, but I don't think there exists
Since we're off-topic...
Any sequence of numbers given by a linear recurrence equation with
constant coefficients can be computed quickly using asymptotically
efficient matrix operations. In fact, the code to do this can be
derived automatically from the recurrence itself.
Here is what you
[CC-ing café again]
On Aug 16, 2010, at 5:52 PM, Daniel Fischer wrote:
I am a bit concerned about the memory usage.
Of your implementation of the matrix power algorithm?
Yes.
Making the fields of Matrix strict should help:
data Matrix a = Matrix !a !a !a
Making all fields strict makes
[continuing off topic]
On Aug 16, 2010, at 3:13 PM, Daniel Fischer wrote:
You can calculate the n-th Fibonacci number in O(log n) steps using
Integer
arithmetic to get correct results.
Yes, I was delighted when I saw this for the frist time. It works be
computing
/1 1\^n
\1 0/
On Monday 16 August 2010 14:37:34, Roel van Dijk wrote:
> On Sat, Aug 14, 2010 at 5:41 PM, Andrew Coppin
>
> wrote:
> > (Then again, the Fibonacci numbers can be computed
> > in O(1) time, and nobody ever needs Fibonacci numbers in the first
> > place, so this is obviously example code.)
>
> A bi
Hi Bulat,
On Monday 16 August 2010 07:35:44, Bulat Ziganshin wrote:
> Hello Daniel,
>
> Sunday, August 15, 2010, 10:39:24 PM, you wrote:
> > That's great. If that performance difference is a show stopper, one
> > shouldn't go higher-level than C anyway :)
>
> *all* speed measurements that find Has
On Sat, Aug 14, 2010 at 5:41 PM, Andrew Coppin
wrote:
> (Then again, the Fibonacci numbers can be computed
> in O(1) time, and nobody ever needs Fibonacci numbers in the first place, so
> this is obviously example code.)
A bit off-topic, but I don't think there exists a function that
computes fi
Hello,
I have uploaded new versions of the primes package to Hackage:
http://hackage.haskell.org/package/primes
Version 0.2.0.0 significantly improves the memory requirements (and as
a result also the run time) compared to 0.1.1. The run time is now
almost linear and when streaming an i
Have you read Wouter Swierstra's "Data Types A La Carte" [1]?
Whether it uses basic and easy parts of Haskell depends on your
mindset. You need to wrap your head around the fixpoint. It requires
at least the MultiParamTypeClasses language extension.
Regards,
Roel
1 - http://www.cs.nott.ac.uk/~ws
I think that's a wonderful idea.
{pe'i le sibdo ku xmagu} (pardon me if my lojban is horrible; I'm
practicing).
On Mon, Aug 16, 2010 at 8:19 AM, Andrew U. Frank <
fran...@geoinfo.tuwien.ac.at> wrote:
>
> may i suggest that the description of the package, where it lists the
> depreciated functio
Hi,
The expression problem [1] can be described as the ability to add new
variants (either constructors or methods) to a data type without
changing the existing code. The Haskell and OO language issues are well
described at [1]
It seems that the expression problem does not exist in Maude[2].
My q
On 14/08/2010 22:29, Yitzchak Gale wrote:
Lennart Augustsson wrote:
So it's a bug in the garbage collector. It's closing a handle that
clearly is still reachable, otherwise this would not have happened.
Simon Marlow wrote:
The handle is in fact not reachable from the roots, because the thre
33 matches
Mail list logo
