Send Beginners mailing list submissions to
beginners@haskell.org
To subscribe or unsubscribe via the World Wide Web, visit
http://www.haskell.org/mailman/listinfo/beginners
or, via email, send a message with subject or body 'help' to
beginners-requ...@haskell.org
You can reach the person managing the list at
beginners-ow...@haskell.org
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Beginners digest..."
Today's Topics:
1. Integer performance of GHC (Sebasti?n E. Peyrott)
2. Re: mayBe stuck (prad)
3. Re: Integer performance of GHC (Brandon S Allbery KF8NH)
4. Re: Integer performance of GHC (Sebasti?n E. Peyrott)
5. Re: mayBe stuck (Dean Herington)
6. Re: Integer performance of GHC (David Virebayre)
7. Re: Re: mayBe stuck (Brent Yorgey)
8. Foreign function interface? (Mathew de Detrich)
9. Re: Foreign function interface? (Mathew de Detrich)
----------------------------------------------------------------------
Message: 1
Date: Sat, 7 Aug 2010 00:32:23 -0300
From: Sebasti?n E. Peyrott <pse...@gmail.com>
Subject: [Haskell-beginners] Integer performance of GHC
To: beginners@haskell.org
Message-ID:
<aanlkti=hk_uexvop=chnwq8d1zm494ewby80zdjro...@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8
Hi, I'm looking for information on Haskell's viability for heavy
integer math. My (very) naive tests show Haskell lagging behind C even
when doing heavy (manual) inlining, compiling with special
optimizations (funbox-strict-fields) and using
continuation-passing-style when possible. Of course, being a beginner
myself, I don't really know how good my code is.
For reference, I wrote a simple MD5 implementation. It's as fast as
the one in Happstack (Happstack.Crypto.MD5), which on my computer runs
around 2.77 times slower then the base version from GNU coreutils
(md5sum).
GHC primitives appear as the next logical step in my tests, however
I've been having problems using fixed sizes for integers (Word32#
comes with no prim-ops defined?).
Do you have any insights on this? Is it reasonable to expect Haskell
integer performance to reach the levels of C with current
implementations? Which are the limiting factors?
Thanks!
------------------------------
Message: 2
Date: Fri, 6 Aug 2010 21:06:24 -0700
From: prad <p...@towardsfreedom.com>
Subject: [Haskell-beginners] Re: mayBe stuck
To: beginners@haskell.org
Message-ID: <20100806210624.630ca...@gom>
Content-Type: text/plain; charset=US-ASCII
On Fri, 6 Aug 2010 13:38:56 +0100
Brent Yorgey <byor...@seas.upenn.edu> wrote:
> you shift the burden of proving that it is safe from the compiler
> onto yourself
> ...
> Here are some off the top of my head, for
> beginners reading this who might not already know:
>
> fromJust
> head
> tail
> init
> last
> (!!)
>
i'm trying to follow this conversation. the problem is that some of
these functions don't behave nicely for some inputs right? like tail []
also, is it better to use pattern matching (x:xs) than head and tail?
could that be part of the point here too?
--
In friendship,
prad
... with you on your journey
Towards Freedom
http://www.towardsfreedom.com (website)
Information, Inspiration, Imagination - truly a site for soaring I's
------------------------------
Message: 3
Date: Sat, 07 Aug 2010 00:09:22 -0400
From: Brandon S Allbery KF8NH <allb...@ece.cmu.edu>
Subject: Re: [Haskell-beginners] Integer performance of GHC
To: beginners@haskell.org
Message-ID: <4c5cdc72.8020...@ece.cmu.edu>
Content-Type: text/plain; charset=UTF-8
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 8/6/10 23:32 , Sebastián E. Peyrott wrote:
> Do you have any insights on this? Is it reasonable to expect Haskell
> integer performance to reach the levels of C with current
> implementations? Which are the limiting factors?
Integer, or Int? In theory Integer should use Int when possible under the
covers, but that may not always do what you expect --- and Integer, being
arbitrary precision, will never be as fast as native Int math.
- --
brandon s. allbery [linux,solaris,freebsd,perl] allb...@kf8nh.com
system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu
electrical and computer engineering, carnegie mellon university KF8NH
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.10 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iEYEARECAAYFAkxc3HIACgkQIn7hlCsL25VWIgCePL5sZCEL87WWoVQXJIUlgC1d
PUwAmQEqtanpLJHa7amkz7GP9qH5rkBg
=SoCp
-----END PGP SIGNATURE-----
------------------------------
Message: 4
Date: Sat, 7 Aug 2010 01:40:25 -0300
From: Sebasti?n E. Peyrott <pse...@gmail.com>
Subject: Re: [Haskell-beginners] Integer performance of GHC
To: beginners@haskell.org
Message-ID:
<aanlktimggvb-x95a0wgqktruexsqc03q2sfcyjbmv...@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8
I meant fixed precision integers (such as Int). Sorry I wasn't clear about that.
On Sat, Aug 7, 2010 at 1:09 AM, Brandon S Allbery KF8NH
<allb...@ece.cmu.edu> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On 8/6/10 23:32 , Sebastián E. Peyrott wrote:
>> Do you have any insights on this? Is it reasonable to expect Haskell
>> integer performance to reach the levels of C with current
>> implementations? Which are the limiting factors?
>
> Integer, or Int? Â In theory Integer should use Int when possible under the
> covers, but that may not always do what you expect --- and Integer, being
> arbitrary precision, will never be as fast as native Int math.
>
> - --
> brandon s. allbery   [linux,solaris,freebsd,perl]  Â
> Â allb...@kf8nh.com
> system administrator  [openafs,heimdal,too many hats]  allb...@ece.cmu.edu
> electrical and computer engineering, carnegie mellon university    KF8NH
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.10 (Darwin)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
>
> iEYEARECAAYFAkxc3HIACgkQIn7hlCsL25VWIgCePL5sZCEL87WWoVQXJIUlgC1d
> PUwAmQEqtanpLJHa7amkz7GP9qH5rkBg
> =SoCp
> -----END PGP SIGNATURE-----
> _______________________________________________
> Beginners mailing list
> Beginners@haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
------------------------------
Message: 5
Date: Sat, 7 Aug 2010 02:05:54 -0400
From: Dean Herington <heringtonla...@mindspring.com>
Subject: [Haskell-beginners] Re: mayBe stuck
To: prad <p...@towardsfreedom.com>, beginners@haskell.org
Message-ID: <a06240800c882a69f5...@[10.0.1.4]>
Content-Type: text/plain; charset="us-ascii" ; format="flowed"
At 11:04 AM -0700 8/6/10, prad wrote:
>On Fri, 6 Aug 2010 03:43:30 -0400
>Dean Herington <heringtonla...@mindspring.com> wrote:
>
>> I would write something like:
>>
>> br [] = []
>> br ss = let (h, t) = break eqD ss
>> in h : case t of
>> [] -> []
>> _ : t -> br t
>>
>thx dean. that certainly looks cleaner than mine, but i'm not sure i
>have seen this construct before.
>
>i thought let/in was used like this:
>
>aaa = let y = 1+2
> z = 4+6
> in y+z
>
>which is like
>
>aaa = y + z
> where y = 1+2
> z = 4+6
>
>in other words, you just define the parts first and use the "in" to
>define the main expression.
>
>but here it seems you are defining what appears to be the main item
>let (h,t) = break eqD ss
>to get the tuple parts and then forming your array (which really is
>the main item) using these parts as
>h : t
>with t being given 2 options.
>
>this is very interesting to me as i had not seen such a construct
>before.
Choosing between `let` and `where` is usually a matter of taste. I
use both, depending on whether I think it's clearer to present the
logic "bottom-up" or "top-down". Here's the `where`-style equivalent
of the `let` version:
br [] = []
br ss = h : case t of
[] -> []
_ : t' -> br t'
where (h, t) = break eqD ss
For a reason I probably can't articulate well, I find the `let`
version more readable in this case. Perhaps it's because I
understand the function as (1) `break`ing on '%', then (2) dealing
with the details of recursing.
Cheers,
Dean
------------------------------
Message: 6
Date: Sat, 7 Aug 2010 08:59:17 +0200
From: David Virebayre <dav.vire+hask...@gmail.com>
Subject: Re: [Haskell-beginners] Integer performance of GHC
To: Sebasti?n E. Peyrott <pse...@gmail.com>
Cc: beginners@haskell.org
Message-ID:
<aanlktimcqpxpt2ozgpw6dr6dcmddylp4gsayvsz5j...@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8
2010/8/7 Sebastián E. Peyrott <pse...@gmail.com>:
> Hi, I'm looking for information on Haskell's viability for heavy
> integer math. My (very) naive tests show Haskell lagging behind C even
Relevant links :
Benefits of the LLVM code generator (lots of other relevant posts on Dons' blog)
http://donsbot.wordpress.com/2010/02/21/smoking-fast-haskell-code-using-ghcs-new-llvm-codegen/
Archive of threads about loop unrolling :
http://www.haskell.org/pipermail/glasgow-haskell-users/2009-March/thread.html#16741
(Also check some of the links in "You might also be interested in:")
http://efreedom.com/Question/1-2978979/Haskell-math-performance
David.
------------------------------
Message: 7
Date: Sat, 7 Aug 2010 09:26:10 +0100
From: Brent Yorgey <byor...@seas.upenn.edu>
Subject: Re: [Haskell-beginners] Re: mayBe stuck
To: beginners@haskell.org
Message-ID: <20100807082609.ga27...@seas.upenn.edu>
Content-Type: text/plain; charset=us-ascii
On Fri, Aug 06, 2010 at 09:06:24PM -0700, prad wrote:
> On Fri, 6 Aug 2010 13:38:56 +0100
> Brent Yorgey <byor...@seas.upenn.edu> wrote:
>
>
> > you shift the burden of proving that it is safe from the compiler
> > onto yourself
> > ...
> > Here are some off the top of my head, for
> > beginners reading this who might not already know:
> >
> > fromJust
> > head
> > tail
> > init
> > last
> > (!!)
> >
> i'm trying to follow this conversation. the problem is that some of
> these functions don't behave nicely for some inputs right? like tail []
All of these functions are "partial", which means that they are
undefined (crash with an error) for some inputs. If possible you
should try to program only with "total" functions (functions which are
defined for every input) since then you know your program will not crash.
>
> also, is it better to use pattern matching (x:xs) than head and tail?
> could that be part of the point here too?
Yes, that's exactly right.
-Brent
------------------------------
Message: 8
Date: Sat, 7 Aug 2010 19:21:44 +1000
From: Mathew de Detrich <dete...@gmail.com>
Subject: [Haskell-beginners] Foreign function interface?
To: beginners@haskell.org
Message-ID:
<aanlktikz_rg4obmse=9takm41ctnhcurtcn5yn2d4...@mail.gmail.com>
Content-Type: text/plain; charset="iso-8859-1"
---------- Forwarded message ----------
From: Mathew de Detrich <dete...@gmail.com>
Date: Sat, Aug 7, 2010 at 7:19 PM
Subject: Re: [Haskell-beginners] Foreign function interface?
To: aditya siram <aditya.si...@gmail.com>
Actually it is possible to use C++ with FFI without having to use a C
Wrapper (albeit its quite complicated)
http://www.haskell.org/haskellwiki/Cxx_foreign_function_interface
<http://www.haskell.org/haskellwiki/GHC/Using_the_FFI>Also refer to
http://qthaskell.berlios.de/ (thats a practical example of QtHaskell, which
is a haskell binding to a C++ library, being Qt)
<http://qthaskell.berlios.de/>
On Sat, Jul 31, 2010 at 6:49 AM, aditya siram <aditya.si...@gmail.com>wrote:
> > Can it interface with C++ as well as C?
> No you have to wrap your C++ code in C.
> -deech <Beginners@haskell.org>
>
> _______________________________________________
> Beginners mailing list
> Beginners@haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://www.haskell.org/pipermail/beginners/attachments/20100807/495f596f/attachment-0001.html
------------------------------
Message: 9
Date: Sat, 7 Aug 2010 19:22:32 +1000
From: Mathew de Detrich <dete...@gmail.com>
Subject: Re: [Haskell-beginners] Foreign function interface?
To: beginners@haskell.org
Message-ID:
<aanlktikh0c7hsk5rtdkeg05h2fbp-fwfosqujbxcq...@mail.gmail.com>
Content-Type: text/plain; charset="iso-8859-1"
eh woops, sorry about what happened above with the mailing list, still
figuring out mutt =D
On Sat, Aug 7, 2010 at 7:21 PM, Mathew de Detrich <dete...@gmail.com> wrote:
>
>
> ---------- Forwarded message ----------
> From: Mathew de Detrich <dete...@gmail.com>
> Date: Sat, Aug 7, 2010 at 7:19 PM
> Subject: Re: [Haskell-beginners] Foreign function interface?
> To: aditya siram <aditya.si...@gmail.com>
>
>
> Actually it is possible to use C++ with FFI without having to use a C
> Wrapper (albeit its quite complicated)
> http://www.haskell.org/haskellwiki/Cxx_foreign_function_interface
>
> <http://www.haskell.org/haskellwiki/GHC/Using_the_FFI>Also refer to
> http://qthaskell.berlios.de/ (thats a practical example of QtHaskell,
> which is a haskell binding to a C++ library, being Qt)
>
> <http://qthaskell.berlios.de/>
>
> On Sat, Jul 31, 2010 at 6:49 AM, aditya siram <aditya.si...@gmail.com>wrote:
>
>> > Can it interface with C++ as well as C?
>> No you have to wrap your C++ code in C.
>> -deech <Beginners@haskell.org>
>>
>> _______________________________________________
>> Beginners mailing list
>> Beginners@haskell.org
>> http://www.haskell.org/mailman/listinfo/beginners
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://www.haskell.org/pipermail/beginners/attachments/20100807/688e4999/attachment.html
------------------------------
_______________________________________________
Beginners mailing list
Beginners@haskell.org
http://www.haskell.org/mailman/listinfo/beginners
End of Beginners Digest, Vol 26, Issue 16
*****************************************
