And now, some sad news - I've apparently had a death in the family. One
of those four wonderful CMD SCSI adapters I brought home two weeks ago
aparently was DOA, so I'm down to three. Pity me (ducking).

First, the VS4000/60 issue:

I've finally put together all the components I needed (read: MMJ cable &
adaptor) to get a serial console out of my VS4000/60, and I managed to
get it netbooting - almost perfectly. It's only got one problem.

After the kernel (1.4L-DMASCSI-CRT kernel, 19991115 shapshot) loads, the
KB is non-responsive. It works fine at the >>>, and the F5/Halt key
still makes the system stop, and the Hold Screen button still makes the
lights flash (Micro-Term VT220 emulating terminal), but the Login:
prompt doesn't accept any characters. Not even enter. Of course, this
leaves the system pretty much dead in the water, as I can't telnet in as
root, and that's the only usable account built-in. It appears that I
can't run pwd_mkdb from the nfs host system to reflect changes in the
master.passwd either, but I didn't really expect that.

Could somone point me to a more recent kernel (preferably that will
still work with the last snapshot binaries that are on the NetBSD ftp
site)? Or is this problem completely unheard of?

I even tried using the console KB (ie, the one plugged directly into the
VS) - it doesn't do squat with the serial console switch thrown.

Second, the UV3800 issue:

Just finished putting together a 3800 for a friend of mine. It's got a
CMD 423 SCSI controller in it. When it goes to read the bootblocks,
while it's doing the xxxx+xxxxx[xxxx]etc stuff, (also 1.4L-DMASCSI-CRT
kernel, 19991115 shapshot), it shows all sorts of wierd messages that
look like SCSI commands (sorry, didn't get a chance to write any down).
Does this sound familar, and is it just some debug stuff that got left
active? Other than this, the machine works fine.

--
 -J. Buck Caldwell