VMS is not the evil empire

[Dave Cornutt]

In article <2528 at pdn.UUCP> reggie at pdn.UUCP (George W. Leach) writes:
 > Another problem we had was that we were
 > sold a machine with 16 Mg of main memory.  Ultrix only could address 8 Mg!

Sounds like the old MS780E memory interleaving trick.  Back in 1983, the
company that I worked for at the time bought a 780 with one of the first
E controllers (and only 2M of memory!).  When we put Interactive IS/3
(a SysIII-based system) up on it, we noticed the same problem...only half
of the memory was visible.  After a couple of hours of probing around in
our (fortunately new) VAX Hardware Handbook, we discovered how to enable
the memory interleving on the controller and make all of the memory
visible.  I don't remember the details now, but all it took was a couple
of register pokes.  We put the necessary commands in the console boot
file and it worked fine.

The amusing thing was that, when we booted up VMS (I think the current
version at the time was 2.3), it didn't do it right either!  And whenever
the FS people ran the micro diags, they were always stunned with micro
#2 told them that it couldn't find the memory controller.  I then had
to show them the secret, mysterious micro #3 disk.  (I once spent 20
minutes on the phone with someone from Remote Diag trying to convince
them that there was such a thing.  I eventually gave up, hung up the
phone, and called back in the hopes that I wouldn't get the same person.
Fortunately, I didn't.)

This was all five years ago.  Fortunately, things appear to have changed
for the better since then.

-- 
Dave Cornutt, AT&T Bell Labs (rm 4A406), Holmdel, NJ (Note new address!)
UUCP:{ihnp4,allegra,cbosgd}!hotly!dkc (path stolen from Shelley)
"The opinions expressed herein are not necessarily my employer's, not
necessarily mine, and probably not necessary"