[Shootout-list] Stuff

[Isaac Gouy]

--- Brent Fulgham <bfulg@pacbell.net> wrote:
> 
> --- Bengt Kleberg <bengt.kleberg@ericsson.com> wrote:
> > different people have different "practically
> > important sets of problems".  the interesting 
> > question is who gets to decided which sets should
> > be in the shootout, and which should be removed?
> 
> The answer to this used to be 'me'.  However, I don't
> think the shootout serves much purpose if it only
> reflects the desires of a single individual.
> Therefore, I've been trying to make this a more
> 'democratic' process.
> 
> So, the answer should be 'we all do'.  However, there
> are some constraints:
> 
> 1.  Ideally, a new test will fit into the existing
> testing framework with minimal effort.
> 2.  A new test should not test the same stuff as
> existing tests.
> 3.  A new test should have a minimal set of
> implementations before being added to the main
> page of the shootout.

4. A new test should (usually) replace an old test
There are even more possible benchmarks than possible language
implementations.

30 benchmarks with implementations in all the languages, 
rather than 100 benchmarks with 5 languages each.


> > ...deleted
> > > (The obvious problem benchmark is pidigits because
> > > it requires arbitrary precision arithmetic.)
> > 
> > would it not be possible to do arbitrary precision
> > arithmetic in languages that does not have it 
> > built-in, by coding it oneself?
> 
> Yes.  The 'loc' metric would then show that the
> language is less friendly to this type of work than
> others.

Write a library and make it available open-source - why is that a
language issue?

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com