Re: [Chicken-users] benchmarks from the Gambit world

[John Cowan]

felix winkelmann scripsit:

> Scheme implementors are often performance freaks (Clinger, Feeley)
> and put too much emphasis in raw performance. This is in part because
> Scheme traditinionally is a playground for compiler writers and
> because pure Scheme is so useless that one is tempted to use it only
> for benchmarking... ;-)

I think there's a psychological (indeed Freudian) explanation as
well.  The Lisp/Scheme community, like the Fortran community and
some others (as distinct from the Cobol, Perl, and Python communities)
suffered a psychological wound in early stages: outsiders said "Your
language is slow, ergo you suck", and it has left a permanent obsession
with performance at all costs long after it mattered in practical terms.
Look how proud Lispers still are that the MacLisp compiler once outperformed
DEC Fortran on numerical code!

> Scheme implementers would do good by concentrating on things that really
> count (like providing extensions, foreign language bindings and better
> development environments). Scheme's metaprogramming facilities (combined
> with a good ffi) is more than enough to allow specialized and highly
> optimized code generation that is sufficiently fast for all purposes.

Agreed.  Thanks for doing those things.

-- 
John Cowan  address@hidden  http://ccil.org/~cowan
And now here I was, in a country where a right to say how the country should
be governed was restricted to six persons in each thousand of its population.
For the nine hundred and ninety-four to express dissatisfaction with the
regnant system and propose to change it, would have made the whole six
shudder as one man, it would have been so disloyal, so dishonorable, such
putrid black treason.  --Mark Twain's Connecticut Yankee