Benchmarking

[Dan Melomedman]

Eli Scott wrote:
> The benchmark that was presented had all the hallmarks of science - a
> repeatable method, data, and conclusions based on that data.  Whether or
> not it was *good* science is up for debate, one which I'll stay out of.

Someone said if you write software and do not experiment or profile,
then the project may just turn into a lengthy and wasteful experiment
instead.

-- 
Rather fail with honor than succeed by fraud. -Sophocles