--On Sonntag, 19. Oktober 2003 16:54 +0200 Andreas Roedl
Benchmarking a OS which is explicitely not recommended for
production and contains lots of debugging hooks seems not
very smart, the release notes would have told him to use
4.8 instead which is faster for those reasons.
I dont see anything in the graphs which would make FreeBSD
(the forthcoming 5.2) serving applications any slower than
Linux 2.6 in the real world, and for shure not less stable.
FreeBSD also seems to be more responsive under heavy load
from my expirience in the past 5 years, while Linux 'feels'
faster when idle. Fefe's benchmark just proved this again ;)