> "Helge MEINHARD, Chorus, CERN-ECP" <meinhard@afsmail.cern.ch>
> writes:
> > Is there any special meaning to the 'b'
> >partition of the system disk assigned by the NetBSD kernel? 
> 
> Have you tried doing a ``pstat -s'' to confirm that the kernel
> isn't trying to swap on the B partition?
How to check that with a miniroot without the pstat binary? Seriously, I 
think it was a bad idea to try with the 'b' partition. I redid using 'd', 
and all is fine for the NetBSD labelled disk.

For the Ultrix labelled disk (rz2), I also copied the stuff (the full 
base.tar etc.) onto the 'd' partition. In fact, I'm getting a little 
further than before (NetBSD does seem to respect 'b' as a swap partition 
even on an Ultrix disk). The messages I'm getting look as follows:

mount -r -v /dev/rz2d /tmp/ultrix
  rz2: WARNING: using ULTRIX partition information
  rzopen: ENXIO on rz2d unit 2 port 3
  # partions 8, size of 0 = -1337960448
  /dev/rz2d on /tmp/ultrix: Device not configured
!@#@@$@%^%#!^%^?? How to convince NetBSD to mount the Ultrix partition?

> >Again nope... It's exactly the same with mount -r. The miniroot I'm using 
> >is the one dated May 31st, 1995, from gregorio.stanford.edu. I then tried 
> >with an Exabyte, but although the special files in /dev are there, the 
> >/netbsd kernel on the said miniroot does not seem to have tape support 
> 
> I don't understand. There are tape in the configuration file for the
> kernel on the miniroot (see below).  If someone wants to send me a
> SCSI tape drive :), I can check them out ...
Sorry, my ignorance. I tried to address tz0 on SCSI address 0 rather than 
5. Having corrected that, the 8205XL gets recognized as a connected tape 
unit. However, trying to read the tape that I wrote under Ultrix with 
exactly the same device, NetBSD complains about non-matching block lengths:

  tz0: Incorrect Block Length, expected 1024, got 16384
  cp: /dev/rmt0: Input/output error

Thanks anyway to everybody who has given hints so far. More highly welcome. 

Helge