On Sun, Mar 03, 2002 at 09:10:52PM -0800, paul beard wrote:
> Speed or any measure of performance is meaningless with a workload 
> to test with. What's the measurement in this case? It sounds like 
> subjective comfort: nothing wrong with that,  but it's hard to 
> quantify and tune for.

That quantification from that page is (quote): "My parents
say that OpenBSD is faster."

There's nothing wrong with people being satisfied with the system
they use, but that's hardly a benchmark ;-)

Such statements are hard to make anyway. It often depends on the
kind of application you run. I have done 4-way benchmark comparisons
between NetBSD, Linux, Solaris and FreeBSD on exactly the same
hardware, and each system won on a few points, and lost on
others. The results should be in the tech-perform archives somewhere.
This was over a year ago. I need to run the becnhmarks
again with up-to-date versions of the OSs (to see the impact
of UBC for NetBSD for example).

- Frank