[Shootout-list] Stuff
Isaac Gouy
igouy2@yahoo.com
Fri, 22 Apr 2005 15:01:30 -0700 (PDT)
--- Jon Harrop <jon@ffconsultancy.com> wrote:
> On Friday 22 April 2005 16:42, Isaac Gouy wrote:
> > The mystery is why you "don't want to do" it on a website that's
> > always been about including a wild range of languages - GAWK to
> > Haskell.
>
> We need to ossify "a wide range of languages": My guess is that all
> of my
> programs should be implementable in D, SML, C#, OCaml, C++, Lisp,
> Java,
> Haskell, Lua, Scheme, Perl, Python and Felix. Is that wide enough?
>
> The main missing languages are C, Fortran and Pascal because I don't
> think the
> necessary data structures will be easy enough to implement in those
> languages, but I may be wrong. Other data structures can be
> substituted but
> are likely to be slower.
>
> I don't want to "change" the problems to make it easy to implement
> them in C
> and Fortran as the problems will no longer represent anything
> relevant or
> useful.
Does the Shootout exist to present problems from your book?
-snip-
> > Obviously, absence of implementation is not evidence that an
> > implementation is even difficult, let alone impossible.
>
> It is not proof. I take it to be evidence.
If you wish to be foolish I'll not argue :-)
-snip-
> For non-trivial tasks, it will be impractical to restrict
> implementations to using identical data structures and algorithms.
That could be one of the reasons why the Shootout deals in trivial
tasks.
-snip-
> I agree with what Ralph said there.
He seems to be far more interested in apples-to-apples.
-snip-
> By "geared up for Fortran" I mean it is a simple array-based
> numerical algorithm which, I believe, is enforced by the benchmark
> specification rather than objectively selected as the most
appropriate
> method (which it probably is because the task is too simple to be
> representative of most scientific computing).
Let's see - you say it's probably the most appropriate method *and* you
say that there's a specification which enforces some kind of bias to
Fortran?
Does it claim somewhere that the task is supposed to be representative
of most scientific computing?
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com