Okay, thanks to the many helpful suggestions from everyone, including
some help from someone at Dig^H^H^HCompaq, I got the SRM console
installed on the 164LX referred to in my previous two messages.  It
works fine, and the NetBSD/alpha install floppies work, the install
scripts work (modulo a few slight rough edges), and we believe we now
have NetBSD/alpha installed on the disk.

The problem is, it won't boot from it.

The machine has two IDE controllers.  There is only one disk, which is
jumpered as master and cabled to one of the controllers.

"sh dev" shows dqa0 and dqa1 (the floppy and the network - dva0 and, if
memory serves, ewa0).  Attempting to boot from dqa0 produces a
complaint that it can't send a Read request; trying dqa1 loads and
starts the primary boot, which then prints something and immediately
clears the screen.  I *think* what it's printing is a similar complaint
about being unable to send a Read request, but it's very hard to be
sure.

Since the use we ultimately want to put the box to does not involve IDE
disk, and the documentation says things about troubles booting from
IDE, we borrowed a PCI SCSI controller (an Adaptec 2940UW, I think)
from an Intel machine and stuck it in the Alpha.  Upon booting from the
floppies, I could disklabel and newfs the SCSI drive, mount both it and
the IDE, copy over the installed system, and then installboot the SCSI
drive.

The problem then is, the console code seems incapable of seeing the
SCSI disk.  "sh conf" lists the controller as hanging off the PCI bus,
correctly (it knows it's an Adaptec whatever-it-really-is), but it
shows no drives attached to it, and "sh dev" still doesn't show it.
Disconnecting the IDE drive entirely doesn't help (not surprisingly).

I spent a while poking at the SRM help stuff, without result.  Is there
some magic incantation I need to mutter to get this to behave, or are
we just stuck booting from floppy at present?

					der Mouse

			       mouse@rodents.montreal.qc.ca
		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B