On Sat, Sep 10, 2005 at 02:36:41PM -0400, Christian von Kleist wrote:
> gary rolland wrote:
> >NetBSD changed my teams life for the better.
> 
>      This is a great story!  I think we all have similar reasons to 
> thank the NetBSD team.
> 
> --
> c v k @ z y b x . c o m

I second that!

After using Linux for 10 years and giving up on the 2.4.x kernel
versions ever becoming stable, we switched to FreeBSD.  This was after
considerable research and noticing a pattern of a mass exodos from Linux
to FreeBSD and essentially no movement in the other direction.  Our
problems went away for about a year or two until we started experiencing
a similar pattern all over again with the FreeBSD 5.x releases for
a long stretch of time with all of both our servers and work stations
crashing all the time.  It was deja vu of our Linux experiences all over
again.  Only this time, we recognized the pattern and were not willing
to be patient enough to fight with it for years hoping it would improve
only to find out it would not.  By the way, that is not a prediction.
I hope the FreeBSD community recognizes their problems and does not
continue down the same path that Linux did.  

At that point we started researching NetBSD and OpenBSD more
extensively.  After finding that the culture and attitudes of the NetBSD
development community aligned much more closely to ours and reading
about the NetBSD philosophy of stability and clean code being the
highest priority, we finally started switched to NetBSD, one machine at
a time, until most all our problems went away.  We are elated with the
results and we have been able to get real work done again over the last
year and have actually been able to do backups without our NFS server
crashing :-).  

During the last year that we ran Linux, the kernel 2.4.0 was released as
a production kernel at a point when it was so buggy that it was unusable
(same thing they did with 2.2.0 several years earlier).  I, as
diplomatically as I could, posted to the kernel mailing list, in reply
to another person who complained about the kernel freezes, suggesting
they follow Linus's original plan of only putting experimental untested
code in odd numbered kernels and keep the even numbered production
kernels as stable as possible.  The only kind of responses I got back
were replies like, "Welcome to software", and comments about how I am
forgetting about the fun factor, and that, if they tested the code in
the development kernels first, it would take to long.  Some of them were
fairly sarcastic.  Even Alan Cox, the primary kernel developer at the
time (next to Linus Torvalds, Linux's creator) and an employee of
RedHat, said that he was still running a several year old 2.2.x kernel
on his machine because of the instability.

We really hope that NetBSD community maintains their philosophy of and
professionalism and never starts going down the path that Linux did.

Thank you for a wonderful product.
It has been a life saver.


-- 
Vincent Stemen
Avoid the VeriSign/Network Solutions domain registration trap!
Read how Network Solutions (NSI) was involved in stealing our domain name.
http://inetaddresses.net/about_NSI.html