Re: Benchmarks
On Nov 11, 11:27 pm, CBFalconer <cbfalco...@yahoo.com> wrote:
James Kanze wrote:
Paul Hsieh <websn...@gmail.com> wrote:
... snip ...
Outside of crypto-hashes, Bob Jenkin's function and my
function, the quality of everything else out there is truly
pathetic.
As I said, I'm interested, because I've been collecting
benchmarks of various hashing functions. Post a link to the
algorithm, and I'll add it in.
One of the original purposes of hashlib was to make testing hash
functions easy. It is much more than just that now, but still
fills the purpose. hashlib includes a couple of very simple, and
fast, hashes for text strings as a convenience for users. All
available on my site.
If they're already widely known hashing algorithms, I've
probably already gotten them. But it might be interesting to
compare our results---I'll just make a guess (:-)) that your
library is written in C, so it will definitely have a different
implementation from mine.
(Just curious about one thing: how do you benchmark. I use
virtual functions to impose a firewall on optimization; do you
use pointers to functions, or some other technique?)
--
James Kanze (GABI Software) email:james.kanze@gmail.com
Conseils en informatique orient=E9e objet/
Beratung in objektorientierter Datenverarbeitung
9 place S=E9mard, 78210 St.-Cyr-l'=C9cole, France, +33 (0)1 30 23 00 34