On Tue, Dec 30, 2008 at 12:14 AM, Bob Armstrong <bob at jfcl.com> wrote:
> Glen Slick wrote:
>I basically followed along with the installation info I found here and
>gettting a basic bootable system installed was fairly straight
>forward.
>
>http://www.itsecuritygeek.com/itsgeek/comments/43bsd-quasijarus-on-simh-vax
Nice.
This procedure is for a MicroVAX, though. You're
not going to be able to
boot directly from magtape on a 11/7xx VAX.
If you already know this (or still have the battle scars yourself),
what we did back in the day, for installs, we booted 9-track tapes on
PDP-11s (RK05 or RL02 if you didn't have tape, but the media charge
was expensive), TK50s or RX50s on early MicroVAXen, and "console
media" for large VAXen. The 11/78x models had RX01s via the
internal PDP-11 console processor, the 11/725 and 11/730 had TU58 by
the internal 8085 console "processor", the 11/750 came right up via
ROM and could boot TU58s, and ISTR the 86xx had RL02, but I don't know
by what attachment. You had to order your OS with the right console
media to be able to boot up a "standalone restore" program which was
then used to (sometimes) prep the disk and to restore the first
saveset off of, commonly, a 9-track tape. Early on, VMS might have
been small enough to fit on console media entirely, but by the time I
was doing it (VMS 4.x), we had a tape and a pack of TU58s for our
11/750. Our UNIX distros (System III, 4BSD...) always came on
magtape.
I still have a pile of TU58s from our 11/750, but when I went to read
them off a few years ago, I did not achieve 100% for any tape set. If
anyone would still have a stack of VMS install floppies for a 11/78x,
I think those might be more robust and quite possibly still legible.
What I can't recall is how hard it is (i.e. - how manual and how much
esoteric knowledge you have to have at your fingertips) to make a
standalone restore kit once you have a running system. Eventually,
and I forget when it started, you could create a VMS "SYSE" directory
structure on your system disk for standalone backup with a provided
script, but that doesn't help you get a completely bare machine up and
running.
This aspect of VAXen has hampered me now and then over the years -
there's a lot of little fiddly software bits you have to have for
installs and complete recoveries - stuff you never need when the
machine is working fine. Unfortunately, I don't think I have
everything we did when I did this every day or I'd have more working
machines. It appears, looking back on things, that a long-term
advantage of the MicroVAX architecture was that they did _not_ have
console processors, so they had to be able to boot any MSCP device, so
disks and diskettes and tape interfaces all were designed to fit the
bill. The 11/7xx line does _not_ have that advantage, so takes more
resources to bootstrap from the factory-fresh state.
So if anyone who has resources from before 1995 (VMS before 6.0 or
older versions of Ultrix or 4BSD) wants to write me off-list, I think
there's a need to pull out what we all might have to see what
combinations of systems and operating systems we can cover. With the
present state of simh, it seems that the 11/780 is the way to go
(there was a recent thread on the simh mailing list about this sort of
thing and I think it's safe to say that emulation of other 11/7xx
machines is distant or worse). Unfortunately for me, I don't happen
to have *any* 11/780-specific resources, but I do have plenty of OS
magtapes, etc. from about 1985-1995, FWIW, and quite a few 11/750 and
11/730-specific resources since that's what we had then.
*Is* anyone on the list sitting on a pile of ancient OS install kits?
If so, have they been read into disk and tape image files already?
-ethan