danial_thom at yahoo.com
Wed Jan 11 10:00:56 PST 2006
--- Freddie Cash <fcash at ocis.net> wrote:
> On January 11, 2006 08:41 am, Danial Thom
> > I really do know what I'm talking
> > about. The freebsd "team" no longer has the
> > talent to get this project done.
> Just once, on one of the many lists you post
> this stuff to, I'd like to
> see some supporting evidence *from you*. A
> description of your test lab,
> the hardware used, the software used, the
> testing methodology, some hard
> numbers, something reproducable. Something
> more than just "I know what
> I'm talking about, you're all wrong."
> Nowhere on any of the Linux, DFly, or FreeBSD
> mailing lists that I
> frequent have you provided any of this.
> Even googling for your name doesn't provide
> anything concrete from you
> regarding this issue. Until that happens, its
> impossible for anybody to
> take you seriously. And these threads will
> just devolve into "I'm right"
> "No you're not" "Yes I am, you're dumb" "No,
> you're dumb" verbal nonsense
> (as has happened three times so far in the past
> 2 months).
> Freddie Cash
> fcash at ocis.net
We seem to have the same discussions over and
over and no one really seems to have a clue, and
the same wrong advice keeps being given. I've
outlined simple tests many times, including in
THIS thread. You just read selectively like
everyone else, because you prefer not to hear
that you've been doing everything wrong your
The proof of the pudding is in the test. Set up a
box with 1 route and 2 NICs and pump traffic
through it until it starts to drop packets. Then
pop another disk on and try with another O/S. Its
not rocket science. FreeBSD 4.x wins hands down.
Then you need to do other things, as a control,
to see how different features work. Doing
fastforwarding, for example, may cause all of the
code and memory being used to sit in the cpu
cache, which may give very different results from
a larger scale test that doesn't. Intels blow AMD
away out of the cache, while AMD does better out
of it, so when you see a benchmark you have to be
careful of what your conclusions are, and what
your environment is.
Frankly I don't see how any of you guys, who rely
on freebsd for your businesses, don't do any
testing yourselves. People slapping up AMD64 dual
systems without having any idea if its any better
than their old box. It mindless. "oh, its 64bit
so it much be faster than 32bit". Its just
Today's freebsd "team" is completely different
than it was during the 4.x heydays; all the great
minds behind it are gone; yet you still follow it
blindly as if its just some project that any ole
programmer can do. Its really quite fascinating.
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
More information about the freebsd-isp