7 May
2001
7 May
'01
9:33 a.m.
Bruce wrote:
Rolm systems were used in some very "cool"
ruggedized applications, some
we
can even talk about now. They extended the standard
Nova instruction set
like most other 3rd party knockoffs did, but typically standard DG
software
was used for program development by us poor software
types.
"Ruggedized"? Puh-lease ;-) Norden made "ruggedized" stuff, meaning
commercial
boards stuffed into a beefed-up chassis. We made _militarized_ stuff at
ROLM,
which with of two painful exceptions were not even based on DG
implementations
(so-called "punches" of DG machines; one was a 1/2 ATR S/130 clone, the
other
was a punch of the MV8K that was so incredibly miserable that only two were
built).
Life at ROLM started with the 1601, which was essentially a Nova 800, and
progressed through the 1666, which added things like stacks but in a fashion
utterly incompatible with the way DG wedged them into the Nova 3. Most
16XX series machines went to the Navy, although as far as I can tell the
largest single buyer of them was MacDAC for use in the GLCM/SLCM (ground/sea
launch cruise missile) erector/launch system. As an interesting(?) aside,
the 16xx machines maintained the Nova I/O bus, including the utter lack of
parity or any other form of error detection. When the NAB people got wind
of
this they went (understandably) nuts and Eddie Yee had to work magic with
ROLM's version of RTOS in order for the system to qualify for use with
"specials".
As Bruce suggested, _most_ DG code would run on ROLM processors, although
ROLM also vended its own operating systems (ARTS and ARTS/32) and peddled
their own language (MSL -- The "MilSpec Language" which looked a great deal
like something between BCPL and C, minus byte pointers, which were a pain in
the ass to make work right thanks to the fact that the Nova was a word
oriented
machine).
The culmination of hardware at ROLM was the Hawk/32, which was an original
implementation of the Eagle (MV) architecture, which was supposed to be
coupled
with a painfully advanced B3-secure distributed OS named MARVIN (the
derivation
of the name is complex; basically those of us who were originally assigned
to
the project were considered to have bad attitudes so the project took its
name
from the similarly attitude-impaired android from
"The Hitchhiker's Guide to
The Galaxy". When pressed by marketing for
the underlying meaning of the
term,
we offered up "Multiprocessor Advanced Virtually Interconnected Network",
which
the marketing drones gobbled up and immediately started regurgitating to the
customer base). MARVIN hit the dustbin about the time that the MSC division
was purchased from IBM by Loral.
In a blast-from-the-past, the PlayStation II/CPU2 project brought three of
us
ROLM refugees together at Toshiba. Each of us had in tow a statue of a hawk
that was handed out upon completion of the Hawk/32 project. As each of us
jointed the project the staff would be abuzz about how we were stealing this
statue from each other -- until they realized that there were several of
these
things running around the office...
Back to the D-116; I have a complete set of prints hanging about if someone
needs
copies. They appear to have belonged to a FE; the print for the front panel
has
been attacked with color highlighters, and the microswitch that disables the
front panel when the key is in the "lock" position has a large circle drawn
around it with a like leading to handwritten text that reads:
"Dear diary: Today is fucked".
--
Chris Kennedy
chris(a)mainecoon.com
http://www.mainecoon.com
PGP fingerprint: 4E99 10B6 7253 B048 6685 6CBC 55E1 20A3 108D AB97