Data General Eclipse System
John Woodall
David J. Dachtera
>
> Anyone know anything about these?
It's been 20+ years since I saw one, don't think I'd be much help, but
ask away - I don't know whether there are any DG-specific newsgroups.
--
David J. Dachtera
dba DJE Systems
http://www.djesys.com/
Unofficial Affordable OpenVMS Home Page and Message Board:
http://www.djesys.com/vms/soho/
This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings
is to be expected.
Feel free to exercise your rights of free speech and expression.
However, attacks against individual posters, or groups of posters, are
strongly discouraged.
Paul Brown
> John Woodall wrote:
>>
>> Anyone know anything about these?
> It's been 20+ years since I saw one, don't think I'd be much help, but
> ask away - I don't know whether there are any DG-specific newsgroups.
--
Being a geek is a state of mind
Being paid to be a geek is a state of utopia
- p...@geekstuff.co.uk
Paul Brown
> Anyone know anything about these?
I actually own one (MV2000) - yes, DG used MV designations as well.
Runs a ghastly OS called AOS/VS which I believe DG still support today
on the Aviions.
Ask away and I'll see if I can help.
Paul
Jerry Leslie
: John Woodall wrote:
: >
: > Anyone know anything about these?
: It's been 20+ years since I saw one, don't think I'd be much help, but
: ask away - I don't know whether there are any DG-specific newsgroups.
There's at least one:
comp.os.aos Topics related to Data General's AOS/VS.
--Jerry Leslie
Hans Vlems
John Woodall <len...@mindspring.com> wrote in message
news:9gegtc$fta$1...@nntp9.atl.mindspring.net...
> Anyone know anything about these?
>
the Eclipses come in two flavors IIRC. The oldest are the successors of the
Nova machines.
The Nova was to DG what the PDP-11 is for DEC. The Nova ran an OS calles
RDOS, similar to
RT-11 in functionality. The early Eclipses ran that same OS, they were
called S/...
Then came the Eclipse MV6000; it was a 32 bit machine and ran AOS.
Until 1986 or so, VMS and AOS were quite similar in functionality. The DG
hardware was as well
built (if not better) than the DEC kit.
So it depends a bit on which Eclipse you are talking about, say an S/120 or
an MV6000.
Hans Vlems
Benjamin Gawert
news:9gegtc$fta$1...@nntp9.atl.mindspring.net...
> Anyone know anything about these?
We have an Eclipse S/140 at work, running RDOS...
Benjamin
Terry Kennedy
> the Eclipses come in two flavors IIRC. The oldest are the successors of the
> Nova machines.
> The Nova was to DG what the PDP-11 is for DEC. The Nova ran an OS calles
> RDOS, similar to
> RT-11 in functionality. The early Eclipses ran that same OS, they were
> called S/...
And C/ and M/. If it isn't "MV####" it's a 16-bit Eclipse. I'd also say
that RDOS was a bit beyond RT-11 - we ran either 32 interactive BASIC
users + batch with a variety of compilers or 64 interactive BASIC users,
depending on the workload.
> Then came the Eclipse MV6000; it was a 32 bit machine and ran AOS.
I believe the MV8000 was the first one.
> Until 1986 or so, VMS and AOS were quite similar in functionality. The DG
> hardware was as well
> built (if not better) than the DEC kit.
DG's goal was always to have a less-expensive design than DEC - the Nova
(and DG as a whole) came about because the founder's design for a 16-bit
mini was rejected (for the PDP-11) while at DEC. The MV8000 was designed
to be a 32-bit system that was a lot less expensive than a VAX-11/780.
There were lots of design concessions, at least on the 16-bit systems,
to save costs. The backplanes were covered with AGA fuses, adding an I/O
controller involved wiring a device-specific I/O paddleboard onto the
backplane with 60 or so wire-wrap wires, etc. while DEC by that time
had moved to ribbon cables off the top of the cards (which DG couldn't
do because of their chassis layout and board stiffeners).
I didn't dislike the DG stuff, it's just very different.
Terry Kennedy http://www.tmk.com
te...@tmk.com Jersey City, NJ USA
Wilko Bulte
>Hans Vlems <hvl...@iae.nl> writes:
>> the Eclipses come in two flavors IIRC. The oldest are the successors of the
>> Nova machines.
>> The Nova was to DG what the PDP-11 is for DEC. The Nova ran an OS calles
>> RDOS, similar to
>> RT-11 in functionality. The early Eclipses ran that same OS, they were
>> called S/...
> And C/ and M/. If it isn't "MV####" it's a 16-bit Eclipse. I'd also say
>that RDOS was a bit beyond RT-11 - we ran either 32 interactive BASIC
>users + batch with a variety of compilers or 64 interactive BASIC users,
>depending on the workload.
>> Then came the Eclipse MV6000; it was a 32 bit machine and ran AOS.
> I believe the MV8000 was the first one.
>> Until 1986 or so, VMS and AOS were quite similar in functionality. The DG
>> hardware was as well
>> built (if not better) than the DEC kit.
> DG's goal was always to have a less-expensive design than DEC - the Nova
>(and DG as a whole) came about because the founder's design for a 16-bit
>mini was rejected (for the PDP-11) while at DEC. The MV8000 was designed
>to be a 32-bit system that was a lot less expensive than a VAX-11/780.
Recommend reading is "The soul of the new machine" by Tracy Kidder.
W/
--
| / o / / _ Arnhem, The Netherlands email: wi...@FreeBSD.org
|/|/ / / /( (_) Bulte http://www.FreeBSD.org
Jay R. Ashworth
to Hawkeye and me and said:
> Recommend reading is "The soul of the new machine" by Tracy Kidder.
"The Soul of A New Machine". Excellent book. Much better than "The
Soul of The Internet", a book on the net which thought it was good
enough to steal the title. It wasn't.
*That* book is Hafner and (Markoff's?) "Where Wizards Stay Up Late",
andother must read.
Cheers,
-- jra
--
Jay R. Ashworth j...@baylink.com
Member of the Technical Staff Baylink
The Suncoast Freenet The Things I Think
Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
OS X: Because making Unix user-friendly was easier than debugging Windows
Hans Vlems
the MV8000 came before the MV6000.
The assembly language of the 16 bit systems was quite nice. IIRC one of the
differences
with the PDP-11 was that the Nova/Eclipse had dedicated IO instructions,
whereas the
PDP-11 used memory mapped IO. I cannot remember whether the Nova allowed
operators
to work on registers only or could also work on memory (like the PDP-11).
The 16 bit DG processors could be found in all kinds of instruments, mass
spectrometers,
NMR equipment etc. The Nova/3 was "copied" by another company (Digilab?) and
that was
used in Varian equipment.
Hans
Terry Kennedy <te...@gate.tmk.com> wrote in message
news:GF11u...@spcuna.spc.edu...
Roger Ivie
> Hans Vlems <hvl...@iae.nl> writes:
>> called S/...
>
> And C/ and M/. If it isn't "MV####" it's a 16-bit Eclipse. I'd also say
The S/ and C/ referred to the types of enhanced instructions present in
the machine. S/ machines had the "scientific" enhancements while C/
machines had the "commercial" enhancements. I don't know what the scientific
enhancements entailed, but IIRC correctly the commercial machines had
instructions for dealing with packed decimal strings. I never encountered
an M/, so I don't know what was special about them. A lab I once worked for
had an S/150 for real-time work and a C/300 for other stuff; the C/300
benchmarked slightly faster than the 11/780 on the piddly little benchmarks
I threw at it.
--
-------------------------+----------------------------------------------------
Roger Ivie | Ben Stein for president!
iv...@cc.usu.edu |
http://cc.usu.edu/~ivie/ |
-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GP dpu s:+++ a C++ UB- P--- L- E--- W- N++ o-- K-- w---
O M+ V+++ PS+++ PE++ Y+ PGP+ t++ 5++ X-- R tv++ b+++ DI+++ D-
G-- e++ h--- r+++ y+++
------END GEEK CODE BLOCK------
Alfred Falk
>
> John Woodall <len...@mindspring.com> wrote in message
> news:9gegtc$fta$1...@nntp9.atl.mindspring.net...
>> Anyone know anything about these?
Yes.
> the Eclipses come in two flavors IIRC. The oldest are the successors of
> the Nova machines.
> The Nova was to DG what the PDP-11 is for DEC. The Nova ran an OS
> calles RDOS, similar to
> RT-11 in functionality. The early Eclipses ran that same OS, they were
> called S/...
> Then came the Eclipse MV6000; it was a 32 bit machine and ran AOS.
> Until 1986 or so, VMS and AOS were quite similar in functionality. The
> DG hardware was as well
> built (if not better) than the DEC kit.
>
> So it depends a bit on which Eclipse you are talking about, say an
> S/120 or an MV6000.
Nova's were comparable, in some ways, to low-end PDP-11's, but also a lot
like PDP-8. It's designer and DG's founder was the designer of the PDP-8,
Edson DiCastro.
16-bit Eclipses (introduced ca. 1974-75) were comparable to later high-end
PDP-11's, especially with respect to memory management. Eclipse instruction
set was super-set of Nova's. (Nova's instruction set was small and rather
elegant in a limited kind of way. Some described it as the "first RISC".
Actually RISC architectures have _way_ more instructions than the Nova.)
Eclipses could run Nova programs unchanged (I mean RDOS) as long as you
wanted to stay inside 64kb. 16-bit Eclipses had various desiginatons, S/xxx
(Scientific) and C/xxx (Commercial). C/xxx had all features of
corresponding S/xxx with floating point and "commercial" instructions
standard. Models varied in memory management and hardware implementation.
Later DG/xxx (Desktop Generation) were based on single chip implementation
and micro-nova bus, IIRC C/x3x instruction set. Operating systems include
RDOS (background/foreground) comparable to RT-11
AOS, like Unix, a child of Multics. comparable to RSX-11.
32-bit Eclipse (introduced ca. 1979 as MV/8000) was designed to run AOS/VS.
It would also run 16-bit AOS, as though it were a C/350. DG's unix (DG/UX?)
would also run on these, including a port that run on top of AOS/VS.
Interesting tid-bit: some of the first, unix-based Graphics Workstations
were made by Control Data. They had DG/4000's inside. DG also sold them
under their own name. That market was eventually taken over by Apollo and
Sun using M68000 processors.
----------------------------------------------------------------
A L B E R T A Alfred Falk fa...@arc.ab.ca
R E S E A R C H Information Systems Dept (780)450-5185
C O U N C I L 250 Karl Clark Road
Edmonton, Alberta, Canada
http://www.arc.ab.ca/ T6N 1E4
http://www.arc.ab.ca/staff/falk/
Chris Scheers
>
> In article <GF11u...@spcuna.spc.edu>, Terry Kennedy <te...@gate.tmk.com> writes:
> > Hans Vlems <hvl...@iae.nl> writes:
> >> RT-11 in functionality. The early Eclipses ran that same OS, they were
> >> called S/...
> >
> > And C/ and M/. If it isn't "MV####" it's a 16-bit Eclipse. I'd also say
>
> The S/ and C/ referred to the types of enhanced instructions present in
> the machine. S/ machines had the "scientific" enhancements while C/
> machines had the "commercial" enhancements. I don't know what the scientific
> enhancements entailed, but IIRC correctly the commercial machines had
> instructions for dealing with packed decimal strings. I never encountered
> an M/, so I don't know what was special about them. A lab I once worked for
> had an S/150 for real-time work and a C/300 for other stuff; the C/300
> benchmarked slightly faster than the 11/780 on the piddly little benchmarks
> I threw at it.
In general, the S/ machines had extended floating point (trig, etc.).
The C/ machines had decimal instructions (notably, EDIT).
The M/ machines had both.
Various Eclipse benchmarks generally only looked good if you could fit
your application in 32KW. Once it got bigger than that, performance
degraded rapidly.
-----------------------------------------------------------------------
Chris Scheers, Applied Synergy, Inc.
Voice: 817-237-3360 Internet: ch...@applied-synergy.com
Fax: 817-237-3074
Terry Kennedy
> The S/ and C/ referred to the types of enhanced instructions present in
> the machine. S/ machines had the "scientific" enhancements while C/
> machines had the "commercial" enhancements. I don't know what the scientific
> enhancements entailed, but IIRC correctly the commercial machines had
> instructions for dealing with packed decimal strings.
Yup. Note that the lines became blurred later on, with "unauthorized
model conversions" made by swapping various processor boards between
boxes. Eclipse S/350, anyone?
Sometimes you wound up with a model conversion as a result of some
necessary upgrade. For example, the S/200 had 3 processor board sets,
one for core memory, one for semiconductor without ECC, and one for
semi with ECC. Installing the semi-with-ECC got you a S/230 upgrade,
with everything except the LED's instead of incandescent lamps on the
front panel (and the badge, of course).
The Nova 4 and the Eclipse S/140 didn't even need a board-swap to be
the same system - it was originally some PROMs, but eventually the spares
folks stocked only the Eclipse flavor board.
> I never encountered
> an M/, so I don't know what was special about them. A lab I once worked for
> had an S/150 for real-time work and a C/300 for other stuff;
It was a rare beast. I think the only model ever made was the M/600, and
AFAIK very few were made. It was to be a "super" Eclipse but died quickly
once the writing appeared on the wall regarding a DG 32-bit system.
> the C/300
> benchmarked slightly faster than the 11/780 on the piddly little benchmarks
> I threw at it.
They were clever little boxes, and with well-written code (or at least
"code optimized for speed" - the source made little or no sense to anyone
looking at it - there was even a book floating around with all the hacks
to do all sorts of useful (and useless) things). But once you wound up
needing to do something complex in a large memory space, the VAX would
eat it for lunch.
What killed them for us (at SPC, when I was there) was that DG pretty
much said "Surprise! We're going 32-bit. No more maintenance on 16-bit
stuff, buy a MV". This was in 81 or thereabouts. At the time we were
running the world's largest XBASIC shop (64 terminals on one S/200 - or
S/230 if you prefer) with off-peak batch replacing 32 terminals. We were
doing our own software maintenance and DG was actually sending me STR's
for me to fix the code and send them the fixes.
But DG's attitude toward 16-bit users pissed us off, and we switched
to DEC, buying a PDP-11/24 to test the waters and then an 11/44. At the
peak of DEC-style computing at SPC, we had 3 11/44's, 2 11/70's, and 3
VAX 8650's running. Now there's only a DS20, and that may be on its way
out, to be replaced by a gaggle of buggy NT servers. Feh.