Computer Language Shootout - concurrency

[Bengt Kleberg]

On 2005-12-06 10:19, Peter-Henry Mander wrote:

> Also, as mentioned here:
> http://www.erlang.org/ml-archive/erlang-questions/200512/msg00026.html
> Some of the benchmarks are tailored so that some strengths of erlang are
> *not* demonstrated. I quote:
> "We started off using 1500 threads but too many
> language implementations failed or timed out."
> 
> <rant>
> 
> Why is there not another concurrency test with 5000 threads and more
> that demonstrate the absolute dominance of erlang? How can the
> benchmarks be useful unless they highlight those languages with unique
> strengths such as concurrency in erlang? Are there other benchmarks that
> have been 'politically-corrected' or 'dumbed-down' to be overly
> inclusive?
> 
> </rant>

it is (was) even a faq entry that states that it a goal to make as many 
as possible of the languages pass as many as possible of the tests. we 
might think that this is 'dumbed-down' but the only alternative offered 
is to take the shootout codebase and do it better yourself.


> I'm trying to find time to rectify what erlang code I can, but it's a
> very low priority. But until _all_ the benchmarks are corrected and
> peer-reviewed the ranking is utterly pointless. If anyone cares (and
> obviously some of us do) we should extract our proverbial digit and
> 'fix' the results by submitting better code, and persuade the benchmark
> designers to add better and wider-ranging concurrency tests, since
> concurrency is becomming a hot topic these days with multicore and cell
> processors.

been there, done that. failed. i wish you luck.


bengt