As a side-effect of having to reinstall large pieces of my system, I
discovered what my reboot problem was.  I had auto-mounting BFFS
partitions on a slow-to-startup disk.  Now, it seems that BFFS
requires >= 37.175 kickstart, and if that wasn't available the
software error requester would turn up.  I never saw this when
coldbooting because the drive wasn't ready so early, due to it's
slow spinup time.  I never saw it when warmbooting from ADOS as
I always had zkicked 37.175.  It was just NetBSD reboots that
triggered the requester!

As you might have guessed by the above message filippa (my NetBSD
system) is alive again, however an important disk has crashed
for me, my local kernel tree *and* the CVS archive of the kernel
tree!  I can't imagine why I put both of them on the same disk!
Especially as I don't have a backup of the disk.  Worse yet, my
wife had her home directory there... shudder...  Does anybody have
a defect CONNER CP-3100?  I think my disk is intact, and that the
controller board has died, so if I get an extra controller board,
maybe I can recover the data there.

No, now I must try to get a scheduling scheme for how I should take
backups in the future :-) Bye, bye... (strangely, I have 10+ years
of Unix sysadm experience, and have seen lots of disk crashed, but
still, I *hate* managing my home system properly, *that* system
*should* just run...)

Niklas

------------------------------------------------------------------------------