|
|
Up |
|
|
  |
Author: flatpvtflatpvt Date: Jul 22, 2008 16:35
Hi folks...
Being a C/C++ hobbyist developer for almost 6 years now (I've been
also working with some Assembly and Python too), I've become somewhat
bored at imperative languages lately, so I decided it was about time
to take a closer look at functional programming languages...
After some research, I was able to pick the ones I thought would be
more interesting for me (Erlang, Haskell, OCaml and F#). Although I
found Haskell the easiest for me to understand, there are some other
points I think I should take into consideration.
One of them is SMP. I found this topic the most confusing one... Some
say Erlang does better at parallel execution, some say Haskell
supports it and others say Haskell has no support at all.
I've been reading some papers about functional language's SMP support
and I must say this may be one of the things I'm most interested in on
FP.
Performance is always important, but as I'm not aiming to produce any
mainstream software with it I think they all do run fairly well for my
needs.
|
| Show full article (1.80Kb) |
|
| | 11 Comments |
|
  |
Author: Ertugrul SöylemezErtugrul Söylemez Date: Jul 22, 2008 17:38
> Being a C/C++ hobbyist developer for almost 6 years now (I've been
> also working with some Assembly and Python too), I've become somewhat
> bored at imperative languages lately, so I decided it was about time
> to take a closer look at functional programming languages...
> After some research, I was able to pick the ones I thought would be
> more interesting for me (Erlang, Haskell, OCaml and F#). Although I
> found Haskell the easiest for me to understand, there are some other
> points I think I should take into consideration.
Hello dear fellow and welcome to the world of functional programming. I
don't know F# and OCaml too well, but I can tell you a few things about
the others you mentioned, particularly Haskell, which is my main
language.
> One of them is SMP. I found this topic the most confusing one... Some
> say Erlang does better at parallel execution, some say Haskell
> supports it and others say Haskell has no support at all.
> I've been reading some papers about functional language's SMP support
> and I must say this may be one of the things I'm most interested in on
> FP.
|
| Show full article (5.70Kb) |
|
| | no comments |
|
  |
Author: Benjamin L. RussellBenjamin L. Russell Date: Jul 24, 2008 23:48
On Wed, 23 Jul 2008 02:38:13 +0200, Ertugrul Soylemez
wrote:
>flatpvt@ gmail.com wrote:
>
>[...]
>
>My favorite Haskell example is the Lucas-Lehmer primality test for
>Mersenne numbers [1], which involves checking an element of an infinite
>sequence:
Your message seems to be missing the reference to [1]. If you don't
mind, could you please provide the missing reference?
-- Benjamin L. Russell
|
| |
| no comments |
|
  |
Author: Jon HarropJon Harrop Date: Jul 25, 2008 17:09
> As you may have noticed, I'm only starting with FP, so if any of you
> have any suggestions or appointments that might help me to choose the
> right language, I'd be really grateful.
As you want to learn rather than write production-quality code, I recommend
replacing F# with Scheme and Mathematica.
F# is little more than a tiny subset of OCaml with no decent tools, few
libraries and vendor lock-in that is specifically designed to let you build
products for Windows in a language better than Java/C#.
Of course, if you ever do want to write production code in a decent language
then F# is head and shoulders above of all other functional languages
simply because it has more rich users and a commercial DLL market.
|
| |
| no comments |
|
  |
Author: Ertugrul SöylemezErtugrul Söylemez Date: Jul 25, 2008 19:24
Benjamin L. Russell Yahoo.com> wrote:
> On Wed, 23 Jul 2008 02:38:13 +0200, Ertugrul Soylemez
> wrote:
>
>> My favorite Haskell example is the Lucas-Lehmer primality test for
>> Mersenne numbers [1], which involves checking an element of an
>> infinite sequence:
>
> Your message seems to be missing the reference to [1]. If you don't
> mind, could you please provide the missing reference?
Indeed, I've forgotten to mention it, sorry. But as you may have
guessed, it's just a reference to Wikipedia. Anyway, here it is:
[1] http://en.wikipedia.org/wiki/Lucas–Lehmer_test_for_Mersenne_numbers
Greets,
Ertugrul.
--
nightmare = unsafePerformIO (getWrongWife >>= sex)
|
| |
| no comments |
|
  |
Author: Benjamin L. RussellBenjamin L. Russell Date: Jul 28, 2008 01:07
On Sat, 26 Jul 2008 01:09:55 +0100, Jon Harrop ffconsultancy.com>
wrote:
>> As you may have noticed, I'm only starting with FP, so if any of you
>> have any suggestions or appointments that might help me to choose the
>> right language, I'd be really grateful.
>
>As you want to learn rather than write production-quality code, I recommend
>replacing F# with Scheme and Mathematica.
Speaking of Mathematica, I am also interested in learning this
tool/language, but find it rather expensive. Because I am not a
student, educator, or government employee, I am not eligible for any
of Wolfram Research's discounts. I would prefer to learn Mathematica
with a tutorial or book by myself, rather than in a classroom setting,
in my spare time after work.
Any suggestions?
-- Benjamin L. Russell
|
| |
| no comments |
|
  |
Author: Jon HarropJon Harrop Date: Jul 28, 2008 08:59
Benjamin L. Russell wrote:
> On Sat, 26 Jul 2008 01:09:55 +0100, Jon Harrop ffconsultancy.com>
> wrote:
>>> As you may have noticed, I'm only starting with FP, so if any of you
>>> have any suggestions or appointments that might help me to choose the
>>> right language, I'd be really grateful.
>>
>>As you want to learn rather than write production-quality code, I
>>recommend replacing F# with Scheme and Mathematica.
>
> Speaking of Mathematica, I am also interested in learning this
> tool/language, but find it rather expensive. Because I am not a
> student, educator, or government employee, I am not eligible for any
> of Wolfram Research's discounts.
Yes, Mathematica is extremely expensive for non-students. Indeed, it is
probably cheaper to enroll yourself as a student just to buy Mathematica!
|
| Show full article (1.69Kb) |
| no comments |
|
  |
Author: Ben FranksenBen Franksen Date: Jul 30, 2008 02:34
> After some research, I was able to pick the ones I thought would be
> more interesting for me (Erlang, Haskell, OCaml and F#).
The most important differences: Haskell is non-strict and purely functional,
the others are strict and impure (i.e. functions can have side-effects). F#
and Ocaml are quite similar (this is hearsay, I am not an expert). Erlang
was designed for control systems, (soft) realtime stuff etc. and is the
only one that directly supports distributed programs.
> One of them is SMP. I found this topic the most confusing one... Some
> say Erlang does better at parallel execution, some say Haskell
> supports it and others say Haskell has no support at all.
> I've been reading some papers about functional language's SMP support
> and I must say this may be one of the things I'm most interested in on
> FP.
Erlang was designed mainly for concurrency, but supports SMP, too.
The quasi-standard Haskell implementation GHC offers a lot of SMP support,
including high-level combinators like 'par' that are pure and completely
abstract from things like threads etc (but you have explicit threads, too,
for concurrent programming).
|
| Show full article (2.85Kb) |
| no comments |
|
  |
Author: Jon HarropJon Harrop Date: Jul 30, 2008 07:35
Ben Franksen wrote:
> The quasi-standard Haskell implementation GHC offers a lot of SMP support,
> including high-level combinators like 'par' that are pure and completely
> abstract from things like threads etc (but you have explicit threads, too,
> for concurrent programming).
Apparently GHC's standard library is not parallelized at all, which
surprised me.
> I don't know much about SMP support for Ocaml and F#.
SMP support in OCaml is non-existent. SMP support in F# is awesome, far
better than any other functional language.
> Due to its integration with .Net, F# can be expected to have good library
> support.
|
| Show full article (2.93Kb) |
| no comments |
|
  |
|
|
  |
Author: William D. NeumannWilliam D. Neumann Date: Jul 31, 2008 06:13
On Jul 28, 3:07 am, Benjamin L. Russell Yahoo.com>
wrote:
> Speaking of Mathematica, I am also interested in learning this
> tool/language, but find it rather expensive. Because I am not a
> student, educator, or government employee, I am not eligible for any
> of Wolfram Research's discounts. I would prefer to learn Mathematica
> with a tutorial or book by myself, rather than in a classroom setting,
> in my spare time after work.
>
> Any suggestions?
Yes. Go to a University and offer a student $20 to buy you a copy
with their student discount.
|
| |
| no comments |
|
|
|
|