[Shootout-list] Stuff

[Isaac Gouy]

--- Jon Harrop <jon@ffconsultancy.com> wrote:
> On Tuesday 26 April 2005 09:34, Bengt Kleberg wrote:
> > Isaac Gouy wrote:
> > > While we're speculating, does anyone want implementations using
> > > different algorithms with different running times?
> >
> > yes, i do. _if the benchmark design allows it_...
> 
> For me, these "same thing" benchmarks are more interesting than the
> "same way" 
> benchmarks because I'd attack problems using different techniques
> depending 
> upon the language I was writing in. For example, I'd be much more
> inclined to 
> use hash sets than balanced binary trees in imperative languages.
> 
> Indeed, the "same way" benchmarks are very skewed because they
> restrict the 
> programmer to using the "lowest common denominator" of features only
> present 
> in all of the languages. Thus, they don't let more modern languages
> exploit 
> more modern features.


Ralph's posting to comp.lang.functional summarized the inherent
stupidity of large-scale cross-language benchmarking:

- compare language performance using the same algorithms and data
structures; and the comparison will be unfair because for some language
the 'standard' approach would use a different algorithm or data
structure.

- compare language performance using different algorithms and data
structures; and then it's a comparison of algorithms and data
structures and programming skill, not a comparison of language
performance.


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