[Shootout-list] Re: Integer overflow, massivly parallell

[Isaac Gouy]

--- Bengt Kleberg <bengt.kleberg@ericsson.com> wrote:
> Date: Thu, 30 Sep 2004 08:46:25 +0200
> From: Bengt Kleberg 
 
> > But it's OK to dumb down until Erlang can get a good score?
 
> you are almost correct. i am willing to dumb down a test to make 
> _the-best-language_ look good. any test, any language.
> you see, i want to find the gold nuggets. if the test is so harsh as
> to make all languagaes appear to be silt, it is no fun. the others,
> the chaff, they do not deserve a helping hand. imho.
> 
> ok?

If Erlang is best it will head the ranking in an inclusive test -
there's no need to craft an exclusive test.


> > People will scan the test, see that none of the languages they
> > recognise are included and stop reading. If they recognise some
> > names they might even wonder for a second why those mainstream
> > languages perform worse than those other unknown languages.
> 
> are you so sure that no mainstream langauge will be able to spawn 
> threads fast enough to reach n withing 300 seconds? i am not.

http://www.lissett.com/ben/copier1.htm


> and, iirc, they will still be featured on the score page.
> with an ugly red cross of shame, bwha ha ha :-)

A good test will rank all the languages. There will be people who are
more interested in finding out if X is better than Y than in proving
Erlang is best. 

(Time-out is a pragmatic failure in the testing approach - ideally we
would provide a ranking for even the slowest languages.)





		
_______________________________
Do you Yahoo!?
Declare Yourself - Register online to vote today!
http://vote.yahoo.com