[Shootout-list] Directions of various benchmarks

[Brent Fulgham]

--- José Bollo <jose.bollo@tele2.fr> wrote:
> I disagree with the rule of "100 LOC limit".
> 
> What if such language formatted using its proper
> style rule has 103 LOC?
> 
> Does it really care? One can see the LOC and exam
> the code for its elegance. just one clic. That is 
> enougth for me.

I can't make a compelling case for 100 LOC.  It was
just a WAG we put together to try to bound the size
of a typical microbenchmark.  Since Java is the
lingua fraca of most universities these days (only
implying that most people can at least glance at a
Java program and figure out what it does), it made
sense to describe the benchmark algorithm using Java
as the 'example' and 100 LOC as the limit in Java.

I would expect Haskell or O'Caml to be much terser,
and perhaps Eiffel and Ada to be a bit more verbose.

In no case have we rejected a shootout submission
because of LOC.

On the other hand, we have encouraged Jon and others
to reduce the scope of tests to stay within the 100
LOC for proposed new test.

I think this has worked fairly well.

If you wish, we can enforce the use of some kind of
Pseudo-Code for describing all future shootout tests.
As a starting point, I suggest MMIX
(http://www-cs-faculty.stanford.edu/~knuth/mmix.html)
;-)

But seriously -- should we require all benchmark
proposals state the algorithm using some kind of
formal language?

-Brent