Supermicro Hardware and FreeBSD
tedm at toybox.placo.com
Wed Jan 5 22:44:14 PST 2005
> -----Original Message-----
> From: owner-freebsd-questions at freebsd.org
> [mailto:owner-freebsd-questions at freebsd.org]On Behalf Of Boris
> Sent: Wednesday, January 05, 2005 8:50 AM
> To: questions at freebsd.org
> Subject: RE: Supermicro Hardware and FreeBSD
> One system cost me 3 months salary in Russia. Is this how you
> treat your users? Why can't your developer use the machine they
> used to make 5.3 work?
YOU are not PAYING the FreeBSD developers to develop for your
particular SuperMicro motherboard. If you were, then you would
have a leg to stand on. Since you are not purchasing FreeBSD,
there does NOT exist any kind of implied warranty or fitness
for merchantability between the developers and you, and therefore
the developers don't owe you anything.
> Everyone tell me to use LINUX. Now I know why. You support bad
> slow version and not good one. Very stupid people.
Unless you were paying RedHat or another Linux distributor the
situation is exactly the same there as well.
I would guess that if -I- had one of these motherboards that
I would have no problems running FreeBSD on it. So far I've
not heard anything specific from anyone on how it works more
slowly on SuperMicro boards. For all we know the people seeing
it run slow are running beta copies that were compiled with
all the debugging code turned on, or they are running the
GENERIC kernel instead of custom-compiling their own.
I should also point out as well that FreeBSD 5.X is probably
going to be slower in any case than FreeBSD 4.X simply because
the kernel does more so it's bigger. FreeBSD 4.X was slower
than 3.X and 3.X was slower than 2.X, and so on. All of this
was because the demand from the userbase is for more and more
features to be added into FreeBSD over the years, not fewer
and fewer features. And adding more features to software makes
it bigger, and bigger code runs slower in general because there's
more instructions for the procesor to go through.
All of this of course is relative since many parts of FreeBSD 5
are more efficient than 4 - and if your software app happens
to use one of those bits a lot, then it's going to run faster.
But if your software apps use bits of FreeBSD that in 4.X
were already optimal, then you probably won't see a speed
More information about the freebsd-questions