First DESKTOP Unix Box?
Simon Bowring
What was the very first commercially available DESKTOP Unix box?
I worked at the UK company Cifer during 1983-85 when we launched
a 68000 based Unix system III desktop model. We did the port
jointly with a company called root computers (I think, "root"
anyway).
Our early systems had a wopping 256kB of RAM. though we quickly
went to 1Mb - the process-granular virtual memory meant that 256kB
was just not enough!
ISTR HP running adverts AFTER our product releases claiming they
had the first desktop Unix systm - I think it had an integrated printer
and may have been called the 9000 series (although I think they've
subsequetly reused this designation)
TIA
Simon Bowring
Larry M Headlund
Simon Bowring <sbowrin...@mpc-data.co.uk> wrote:
>Sorry if this is a FAQ....
>
>What was the very first commercially available DESKTOP Unix box?
>
Your going to have to define DESKTOP. Do what we now call workstations
count, for example.
>I worked at the UK company Cifer during 1983-85 when we launched
>a 68000 based Unix system III desktop model. We did the port
>jointly with a company called root computers (I think, "root"
>anyway).
>
>Our early systems had a wopping 256kB of RAM. though we quickly
>went to 1Mb - the process-granular virtual memory meant that 256kB
>was just not enough!
>
>ISTR HP running adverts AFTER our product releases claiming they
>had the first desktop Unix systm - I think it had an integrated printer
>and may have been called the 9000 series (although I think they've
>subsequetly reused this designation)
Sounds like the HP Integral. This was a portable (lugable) computer about
the size and weight of a portable sewing machine. It integrated the display,
computer and printer. THE OS was in ROM. It came out before the end of 1984.
However, HP had HP-UX systems begore then, In fact, it was in the fall of 1984
that HP released the 9000/320, a 68020 based system. Previously they
had 68000 based systems: 9000/320.
IBM had 68000 based Unix systems in this time frame, also designated 9000.
Tandy (Radio Shack) had 68000 based Xenix systems.
In pure software, Xenix and Venix were Un*x OS which ran on Intel 8086
IBM PC's.
In 1983 the market also included a plethora of "Unix-like" boxes, running
under such names as Unos.
A lot will depend on your definition of "desktop" and "Unix."
--
--
Larry Headlund l...@world.std.com Mathematical Engineering, Inc.
(617) 242 7741
Unix, X and Motif Consulting Speaking for myself at most.
Christopher Stacy
Simon> Sorry if this is a FAQ....
Simon> What was the very first commercially available DESKTOP Unix box?
Simon> I worked at the UK company Cifer during 1983-85 when we launched
Simon> a 68000 based Unix system III desktop model. We did the port
Simon> jointly with a company called root computers (I think, "root"
Simon> anyway).
Three Rivers Corporation's PERQ was out in 1980 or 81, but I don't
know if the original model was a desktop or if it was larger. It had
a bitmapped display, mouse, network, and 68000 based processor running
some kind of UNIX. It was a CMU thing.
There were desktop machines from Western Digital, which were based on
systems developed at MIT. These were 68000 based workstations using
MIT's "NuBus". I don't remember what they looked like or exactly what
year they came out, but it would have been around the same time as the PERQ.
PERQ was later than the MIT Lisp Machine (which was a very high-end system,
and the first commercially available bitmap-mouse-network thing, but which
was not a desktop and not a UNIX), but predates the Xerox Star (which
I think was more of a word processor than a full-blown workstation).
I remember hearing the name Cifer back then, too, but I think it
was a little later than all that stuff.
Chris Hedley
A couple of possibilities from the 81-83 timeframe, one of which was
covered recently but I forget the exact details: one was either an
external processor for or a system using the motherboard of a BBC model
B, which I think *may* have ran UNIX of some description; the other
was a 68000-based "PC" made by, IIRC, ACT/Sirius, described in one of
the computer rags at the time as "blindingly fast." Doubtless there
were earlier ones, but my knowledge of pre-1980 microcomputers is
scant at best.
Chris.
Charles Richmond
that was based on a 68010??? I have one of those in my closet.
It only had a 20 meg disk drive. I was amazed at how *small* the
"vi" editor executable is for this box!!! It seems to be a pretty
much fully implemented "vi" editor...
--
+-------------------------------------------------------------+
| Charles and Francis Richmond <rich...@plano.net> |
+-------------------------------------------------------------+
Phil Gustafson
>>>>>> On Thu, 24 Jan 2002 16:13:36 +0000 (GMT), Simon Bowring ("Simon") writes:
> Simon> Sorry if this is a FAQ....
> Simon> What was the very first commercially available DESKTOP Unix box?
>
> Simon> I worked at the UK company Cifer during 1983-85 when we launched
> Simon> a 68000 based Unix system III desktop model. We did the port
> Simon> jointly with a company called root computers (I think, "root"
> Simon> anyway).
>
>Three Rivers Corporation's PERQ was out in 1980 or 81, but I don't
>know if the original model was a desktop or if it was larger. It had
>a bitmapped display, mouse, network, and 68000 based processor running
>some kind of UNIX. It was a CMU thing.
>
>There were desktop machines from Western Digital, which were based on
>systems developed at MIT. These were 68000 based workstations using
>MIT's "NuBus". I don't remember what they looked like or exactly what
>year they came out, but it would have been around the same time as the PERQ.
>
it cost around $10K around 1983. I was pretty jealous of my one friend
who had one.
Phil G.
--
Phil Gustafson <ph...@panix.com>
There is a tarantula in the song, of course. I remember when you
could ask your grocer to save one for you. -- Casady
Chris Hedley
> What about the A.T. & T. UNIX box, the 6100 or 6300 or something,
> that was based on a 68010??? I have one of those in my closet.
I told you my knowledge was a bit threadbare here! Now I remember
seeing a couple of pictures of those things years ago, didn't they
appear sometime in the late '70s?
> It only had a 20 meg disk drive. I was amazed at how *small* the
> "vi" editor executable is for this box!!! It seems to be a pretty
> much fully implemented "vi" editor...
Amazing, isn't it? ISTR the V7 kernel for the PDP11 was something
like 70K in size... but tell the kids that these days an' they won't
believe yer.
Chris.
Charlie Gibbs
(Larry M Headlund) writes:
>In article <fobjevatzcpqngnpbhx.gqgoao7.pminews@news-ogw>,
>Simon Bowring <sbowrin...@mpc-data.co.uk> wrote:
>
>>Sorry if this is a FAQ....
>>
>>What was the very first commercially available DESKTOP Unix box?
>
>Your going to have to define DESKTOP.
Since it's spelled in all caps it's obviously an acronym. How
about Diminutive Electronic System Konnected To Outside Power?
(Gimme a break, misspellings are very chic among marketroids.)
>Sounds like the HP Integral. This was a portable (lugable) computer
>about the size and weight of a portable sewing machine.
I heard it described as a "lunchbox".
When I hear the term "luggable" I think of the Olivetti M18 on
which our software was born. It was about the size and weight
of a 40-pound suitcase. Unfortunately, to fit among the card
slots the handle was mounted at 90 degrees to the comfortable
orientation found on a standard suitcase (i.e. across the long
axis instead of along it). I had to lug the thing a couple of
blocks one time. My arm wound up several inches longer; after
the pain subsided I became a ping-pong pro.
--
cgi...@sky.bus.com (Charlie Gibbs)
Remove the first period after the "at" sign to reply.
I don't read top-posted messages. If you want me to see your reply,
appropriately trim the quoted text and put your reply below it.
Steve O'Hara-Smith
c...@ieya.co.REMOVE_THIS.uk (Chris Hedley) wrote:
CH> A couple of possibilities from the 81-83 timeframe, one of which was
CH> covered recently but I forget the exact details: one was either an
CH> external processor for or a system using the motherboard of a BBC model
CH> B, which I think *may* have ran UNIX of some description; the other
That would be the Torch 68k, I had a large hand in the Z80 based
original which came out in mid '82 (I can place this with precision having
escaped university in '81). I think the 68K was about a year later, but by
that time nearly every original design team member had quit for greener
pastures (I almost lead the stampede, a couple of people beat me to it).
Ah, the early 80's when everybody wanted to make you rich, but collecting
the readies could prove a mite tricksome :)
Oh yes the Torch 68k was indeed a UNIX box, the development system
we (mostly Dave Oliver until he quit) used was a thing called The Frog which
was about the size of the draw unit that goes under the desk IIRC. I cannot
remember where we got it from though (it was also a 68K based system).
Dave had fun just after he left putting two 68Ks on the same bus
with a remarkably small amount of logic, it never left CST though AFAIK
so instead of a cute 68K SMP system (nobody wanted it without software)
they wound up producing lots of IEEE488 interface boxes (lot's of people
wanted them).
--
C:>WIN | Directable Mirrors
The computer obeys and wins. |A Better Way To Focus The Sun
You lose and Bill collects. | licenses available - see:
| http://www.sohara.org/
Don Quixote
>According to Charles Richmond <rich...@ev1.net>:
>> What about the A.T. & T. UNIX box, the 6100 or 6300 or something,
>> that was based on a 68010??? I have one of those in my closet.
>
>I told you my knowledge was a bit threadbare here! Now I remember
>seeing a couple of pictures of those things years ago, didn't they
>appear sometime in the late '70s?
Early eighties, no? And made by Convergent Technologies, OEM'd by
AT&T ISTR.
K.J. Seefried
<cst...@swingandcircle.com> wrote:
>
>Three Rivers Corporation's PERQ was out in 1980 or 81, but I don't
>know if the original model was a desktop or if it was larger.
>
I've never seen a PERQ that wasn't a deskside box.
To the original question, the earliest I can think of offhand is
Microsoft (shudder) Xenix in 1980. There were some 8086 machines
(Altos?) that ran variations on V7 that might have been earlier.
The 68000 was introduced in 1979. Don't remember if there was a Unix
for it before the 68010 was introduced.
If one could argue that machines like the PDP-11/23 or 11/34 are
desktops, which could push us back to the V6 days (or earler...I know
V4 (circa 1974) ran on the 11/45). Most people wouldn't consider such
machines desktops, however.
When did Venix come out on the DEC Pro? Much later than 1980, I
think.
Ken Seefried
Liam Greenwood
>To the original question, the earliest I can think of offhand is
>Microsoft (shudder) Xenix in 1980. There were some 8086 machines
>(Altos?) that ran variations on V7 that might have been earlier.
Yep. My first unix experience was on an Altos 586, running
Xenix 1.3 IIRC. An 8086 box, 6 serial ports and a 10 MEG hd. That
would have been mid 1982. We ran an issues systems at the library
on it, loosely coupled to the Council B1800 which ran the OPAC and
backend of the issue system. It happily ran 4 teminals with sub-sec
response time for the scanning and recording of the issues/returns
data. Only lookup it had to do locally was the Borrower name on
issue, and the reserved book list on returns.
Cheers, Liam
Al Kossow
> The 68000 was introduced in 1979. Don't remember if there was a Unix
> for it before the 68010 was introduced.
>
Unisoft did many, many ports of V7 before the 68010 was released.
Dual systems had a 68K S100 board that ran it. It also ran on Lisa.
There were lots of people who used it on the SUN Multibus board.
I'm drawing a blank as to who was Unisoft's earliest customers.
--
Charles River Data Systems was pretty early in the market, too.
CBFalconer
>
... snip ...
>
> When I hear the term "luggable" I think of the Olivetti M18 on
> which our software was born. It was about the size and weight
> of a 40-pound suitcase. Unfortunately, to fit among the card
> slots the handle was mounted at 90 degrees to the comfortable
> orientation found on a standard suitcase (i.e. across the long
> axis instead of along it). I had to lug the thing a couple of
> blocks one time. My arm wound up several inches longer; after
> the pain subsided I became a ping-pong pro.
<Thread-Drift Warning>
Then there was the luggable IBM machine that implemented APL and
cost somewhere around 20 kilobucks. Feasable luggability still
lay in the future with Osborne and Kaypro, and 2 kilobucks. Still
later Compaq reared its head in the same field.
--
Chuck F (cbfal...@yahoo.com) (cbfal...@XXXXworldnet.att.net)
Available for consulting/temporary embedded and systems.
(Remove "XXXX" from reply address. yahoo works unmodified)
mailto:u...@ftc.gov (for spambots to harvest)
Foobar T. Clown
>
> [...] Simon Bowring ("Simon") writes:
> Simon> Sorry if this is a FAQ.... What was the very first
> commercially available DESKTOP Unix box?
>
> a bitmapped display, mouse, network, and 68000 based processor
> running some kind of UNIX. It was a CMU thing.
All true, except that it didn't use a 68000, and it didn't run UNIX (at
least, not to begin with), and it wasn't exactly a C-MU thing.
Three Rivers Computer Corporation was located about a mile and a half
from the C-MU campus. It was founded by C-MU alumni, but AFAIK, none
of them was employed by the university at the time. The PERQ work-
station was designed to meet the requirements of an RFP published by
the C-MU computer science department, but the work was all done off-
campus. I kind of recall that C-MU only reluctantly bought PERQs. They
had no choice, really -- no one else responded to their RFP.
There were two processors in the original PERQ I workstation; the main
processor and the I/O processor. The I/O processor was a Z80, and it
handled disk requests, two RS-232 ports, the keyboard, the mouse, and
a GPIB bus. The I/O processor and all of its peripheral interfaces
occupied one slot in the backplane (circuit boards were approximately
18 inches by maybe 12 inches.) The original I/O board was quickly
replaced by one which also contained a three megabit ethernet interface.
IIRC, the so-called EIO board was a ten layer circuit board (ugly).
One whole backplane slot was filled with memory chips -- 256 K WORDS in
the first model, quickly replaced by a 512K word board and then a one
megaword board. Words were 16 bits, and the memory was *NOT* byte
addressable.
The main CPU occupied another whole backplane slot all by itself. It
was built mostly from 74LSxxx TTL. There was one LSI part: A micro-
sequencer which (I think) was made by AMD, and there was a whole bunch
of 70nS static RAM for the control store.
The control store was writable. It was orgainized as 4K words with
some really weird number of bits per word (numbers like 34 and 28 and
19 keep popping into my head). A later model expanded the control
store to 16 K words. There was some kind of interface which allowed
the Z80 on the I/O processor to write to the control store at boot time,
but there was also an interface which allowed the CPU to re-write its
own microinstructions.
I personally know of two different instruction sets that were
implemented in microcode. The first one to be used was called the Q-
code instruction set. It was more or less a straightforward implemen-
tation of the UCSD P-Code virtual machine. (Most PERQ software was
written in Pascal, and compiled with the UCSD compiler.) The other one
I know of was the C-MU SPICE Lisp instruction set. There may have been
a third, but more on that below.
There were three operating systems used on PERQs that I know of. (No,
I am not counting Aaron Wohl's port of CPM because that ran entirely on
the Z80, and it didn't use the main CPU at all.) The first OS -- called
POS -- was written by Three Rivers. "POS" stood for "Piece Of Shit."
It was written for a demo, and it was never intended to be sold as a
product. The company's founders -- I swear to God, I am NOT making this
up -- expected to sell computers with *NO* operating system. They were
prepared to sell and support the micro-assembler; the Pascal microcode,
libraries, and compiler; and they expected to sell and support the
bootstrap loader. Everything else was supposed to be up to the
customer. Nobody below the top level of management was surprised when
POS was added to the product catalog, and the definition of "POS" was
changed to "Perq Operating System."
The second OS was Accent. Accent probably was the coolest OS ever.
Let me say that again, it's important.
Accent was the COOLEST OS *EVER* devised by Man!!!!
Well, it wasn't "Man" so much as it was Rick Rashid -- a research asso-
ciate at C-MU (IIRC) -- and a handful of other C-MU folk. Accent was
such a great idea that when C-MU abandoned the PERQs, they ported it to
other hardware and called it "Mach." Mach was licensed by NeXT computer
corp, and became the core of NeXTStep, and that was eventually purchased
by Apple computer where it has mutated into Mac OS X. While much of
that was happening, Rick Rashid accepted a top-level engineering job at
a large and famous software house, where he put most of the best ideas
from Mach (and a whole lot of compromises) into an entirely new
operating system called Windows NT.
The third OS was called QNX, and it came out of ICL Ltd. in England.
I have never seen a PERQ running QNX, but I understand it to be a port
of something called PNX which I understand was somewhat like UNIX.
AFAIK, the PNX source code was written in the C programming language.
I do not know whether the ICL folk wrote their own microcode, or whether
they wrote a C compiler that targeted the Q-code instruction set.
ICL entered the picture when C-MU dropped out, and Three Rivers (by then
called "PERQ Systems, Inc.") sought out another large customer to pay
the bills. Some time after, PERQ Systems ceased operations as the
result of a business deal between themselves and a third company. ICL
went on to develop something called the Perq 3B (Perq II was just a new
skin wrapped around the same old circuit boards), while the third
company completed the design of what would have been the Perq 3A except
they named it something else. I know nothing of the Perq 3B, but the
Perq 3A was, TA DA!, A 68020 machine running real, honest-to-God, AT&T
UNIX. (It was intended to be a whole lot more than that, but that's
another story.)
The Perq 3A was never sold *AS* a product, but rather, it was sold *IN*
a product -- a computerized typesetting and photo-lithography system
that sold for several millions of dollars a pop.
> PERQ was later than the MIT Lisp Machine (which was a very high-end
> system, and the first commercially available bitmap-mouse-network
> thing, but which was not a desktop and not a UNIX) [...]
I've never seen a MIT lisp machine, but I've seen a Symbolics machine --
pretty much the same thing. It was six feet tall, weighed several
hundred pounds (not counting the separate, floor-standing disk drive),
and it had special power and air conditioning requirements. It *WAS* a
personal computer -- only one person could use it at one time -- but it
sure as hell was no DESKTOP personal computer.
-- Foo!
Foobar T. Clown
>
> One true UNIX desktop was a 68000-based box from Fortune Systems.
> IIRC it cost around $10K around 1983. I was pretty jealous of my one
> friend who had one.
When did AT&T market their "UNIX PCs" (e.g., the 3B1, 3B2, etc.)?
-- Foo!
Christopher Stacy
CBFalconer> Charlie Gibbs wrote:
>>
CBFalconer> ... snip ...
>>
>> When I hear the term "luggable" I think of the Olivetti M18 on
>> which our software was born. It was about the size and weight
>> of a 40-pound suitcase. Unfortunately, to fit among the card
>> slots the handle was mounted at 90 degrees to the comfortable
>> orientation found on a standard suitcase (i.e. across the long
>> axis instead of along it). I had to lug the thing a couple of
>> blocks one time. My arm wound up several inches longer; after
>> the pain subsided I became a ping-pong pro.
CBFalconer> <Thread-Drift Warning>
CBFalconer> Then there was the luggable IBM machine that implemented APL and
CBFalconer> cost somewhere around 20 kilobucks. Feasable luggability still
CBFalconer> lay in the future with Osborne and Kaypro, and 2 kilobucks. Still
CBFalconer> later Compaq reared its head in the same field.
The IBM 5100!
I used one a few times back in the mid-1970s at Scientific Timesharing.
The screen was like 32x12, but logically it was 64x12 (or maybe it was
like 80x16, I forget) but you toggled a switch back and forth to see
the left and right sides of the screen! Of course, short lines were
fine for APL. It also had BASIC in the ROM, but I didn't use that.
I was told that the machine was actually emulating an IBM mainframe.
Mass storage on the device was a tape cartridge.
This was all long before the 68000 though!
Christopher Stacy
around 1972. It used a proprietary processor constructed from TTL.
It had a keyboard, graphics display (not sure what kind of generator,
but it could draw and play 3D Tic Tac Toe), and a casette tape drive.
I think it had a few Kbytes of memory. People programmed actual
applications on them, but it was intended as an intelligent terminal,
and I mostly used it to talk to an IBM/370.
The Datapoint 2200 was supposed to use the 8008, the first complete
microcomputer CPU chip, which was commissioned for this project.
But the two companies contracted for it, Intel and TI, were unable
to deliver working versions of the chip in time.
Christopher Stacy
>>>>> On Fri, 25 Jan 2002 03:58:33 GMT, Foobar T Clown ("Foobar") writes:
Foobar> I've never seen a MIT lisp machine, but I've seen a Symbolics machine --
Foobar> pretty much the same thing. It was six feet tall, weighed several
Foobar> hundred pounds (not counting the separate, floor-standing disk drive),
Foobar> and it had special power and air conditioning requirements. It *WAS* a
Foobar> personal computer -- only one person could use it at one time -- but it
Foobar> sure as hell was no DESKTOP personal computer.
Right: only the user's console sat on the desktop.
The machine was located at the end of the console cables,
in an air-conditioned machine room. I am pretty sure it was
the first computer that was commercially available with a
bitmapped display, mouse, network, and of course local disk.
It features a large address space, several (36 bit) megawords
of memory, and an 80-300 MB hard drive. The CPU was a microcoded
stack machine that implemented Lisp, which the entire operating
system and all applications were written in.
(Later versions of the Lisp Machine were, of course,
even more powerful.)
Frank McConnell
> What was the very first commercially available DESKTOP Unix box?
This is a good question....
> I worked at the UK company Cifer during 1983-85 when we launched
> a 68000 based Unix system III desktop model. We did the port
> jointly with a company called root computers (I think, "root"
> anyway).
...but I think you need to look a little bit earlier into the 1980s
for the answer.
Sun Microsystems was founded in 1982 and I'm thinking it was not
the only outfit building Unix systems around the Stanford
SUN design. Forward Technologies and Codata come to mind, but
I'm not sure they were building desktops.
How about the Callan Unistar 100? That was clearly a desktop, a sort
of fat deep terminal with room for a Multibus chassis and some disk
drives. I've got some manuals for that, dating from 1982.
> ISTR HP running adverts AFTER our product releases claiming they
> had the first desktop Unix systm - I think it had an integrated printer
> and may have been called the 9000 series (although I think they've
> subsequetly reused this designation)
I wonder which you are thinking of: the HP Integral, aka HP 9807, a
sort of lunchbox portable 68000-based Unix system; or the HP 9020,
later 9000 Series 520, a desktop system (perhaps deskside if you
wanted any serious disk space) built around HP's 32-bit FOCUS
processor which could run BASIC or HP-UX. The 9020 was announced
in September 1982 and shipped in 1982. The Integral was a bit
later, I'm thinking 1985 or 1986.
Pictures of both are at <http://www.reanimators.org/hp/index.html>,
the Integral up top and the 520 underneath.
-Frank McConnell
daven...@sourplum.org.uk
> A couple of possibilities from the 81-83 timeframe, one of which was
> covered recently but I forget the exact details: one was either an
> external processor for or a system using the motherboard of a BBC model
> B, which I think *may* have ran UNIX of some description; the other
> was a 68000-based "PC" made by, IIRC, ACT/Sirius, described in one of
> the computer rags at the time as "blindingly fast." Doubtless there
> were earlier ones, but my knowledge of pre-1980 microcomputers is
> scant at best.
I recall torch having a Z80 based BBC 2nd processor in that time frame,
but that would have been CP/M based
The Acorn NS16032 2nd processor was later (we - the Cambridge Uni Comp Soc -
saw one late 84/early 85 IIRC) - this would have been capable of running
UNIX (1MB RAM, 20MB Winchester) but was running Tripos.
Dave
Brian Inglis
>Sorry if this is a FAQ....
>
>What was the very first commercially available DESKTOP Unix box?
>
>I worked at the UK company Cifer during 1983-85 when we launched
>a 68000 based Unix system III desktop model. We did the port
>jointly with a company called root computers (I think, "root"
>anyway).
From Salus "A Quarter Century of Unix" p133:
Among the users who had helped himself was Haruhisa Ishida at the
University of Tokyo. "I was the first user and licensee of UNIX
in Japan," he told me. "It was in 1976 that I signed the
agreement with AT&T Bell Labs. The Unix was version 6, which I
implemented on an LSI-11 with only slight modification."
The DEC LSI-11 was the desktop micro version of the PDP-11, so I
think we have a winner, unless someone has an earlier instance.
--
Thanks. Take care, Brian Inglis Calgary, Alberta, Canada
Brian....@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca)
fake address use address above to reply
tos...@aol.com ab...@aol.com ab...@yahoo.com ab...@hotmail.com ab...@msn.com ab...@sprint.com ab...@earthlink.com ab...@cadvision.com ab...@ibsystems.com u...@ftc.gov
spam traps
Simon Bowring
>Well, it wasn't "Man" so much as it was Rick Rashid -- a research asso-
>ciate at C-MU (IIRC) -- and a handful of other C-MU folk. Accent was
>such a great idea that when C-MU abandoned the PERQs, they ported it to
>other hardware and called it "Mach." Mach was licensed by NeXT computer
>corp, and became the core of NeXTStep, and that was eventually purchased
>by Apple computer where it has mutated into Mac OS X. While much of
>that was happening, Rick Rashid accepted a top-level engineering job at
>a large and famous software house, where he put most of the best ideas
>from Mach (and a whole lot of compromises) into an entirely new
>operating system called Windows NT.
Fascinating! How does this square with the popular folklore that
DEC's VMS architect Dave Cutler was largely responsible for NT's
kernel, which borrowed heavily from VMS (which borrowed from RSX-11
which borrowed from IBM's TSO or TSC or something)?
Simon
John Everett
sbowrin...@mpc-data.co.uk says...
>
>Sorry if this is a FAQ....
>
>What was the very first commercially available DESKTOP Unix box?
>
>I worked at the UK company Cifer during 1983-85 when we launched
>a 68000 based Unix system III desktop model. We did the port
>jointly with a company called root computers (I think, "root"
>anyway).
Unix boxes predate the timeframe you're talking about. Looking through my
notebook from 1981 to 1982, I find references to Altos ACS 8600 (8086 box)
and a proposed 68000 box, Intel (8086 box), Fortune Systems 32:16 (I believe
a 68000 box), Dual Systems, Zilog System 8000, Onyx (68000 box), and Paradyne
(Z8000 box). I also recall visiting Bell Labs (now Lucent) in Naperville, IL
to look at a Unix desktop called perhaps the 3B2.
I'm not sure who was first, but in the timeframe I'm refering to there were
many, many Silicon Valley Fiction Factories promising to build Unix boxes.
Some of them even did.
--
jeverett<AT>wwa<DOT>com (John Everett) http://www.wwa.com/~jeverett
Pete Fenelon
> The 68000 was introduced in 1979. Don't remember if there was a Unix
> for it before the 68010 was introduced.
What was the Tandy thing with a 68k running Xenix and a Z80 doing the
I/O? That was out very early in the 80s....
pete
Simon Bowring
>From Salus "A Quarter Century of Unix" p133:
>
>Among the users who had helped himself was Haruhisa Ishida at the
>University of Tokyo. "I was the first user and licensee of UNIX
>in Japan," he told me. "It was in 1976 that I signed the
>agreement with AT&T Bell Labs. The Unix was version 6, which I
>implemented on an LSI-11 with only slight modification."
>
>The DEC LSI-11 was the desktop micro version of the PDP-11, so I
>think we have a winner, unless someone has an earlier instance.
That's certainly the best so far. Definately Unix and definately
Desktop (but not sure about "commercial")!
Thanks very much.
Simon
Jim
<fu...@gazonk.del> wrote:
> Three Rivers Computer Corporation was located about a mile and a half
> from the C-MU campus. It was founded by C-MU alumni, but AFAIK, none
> of them was employed by the university at the time. The PERQ work-
> station was designed to meet the requirements of an RFP published by
> the C-MU computer science department, but the work was all done off-
> campus. I kind of recall that C-MU only reluctantly bought PERQs. They
> had no choice, really -- no one else responded to their RFP.
<snip>
> There were three operating systems used on PERQs that I know of. (No,
> I am not counting Aaron Wohl's port of CPM because that ran entirely on
> the Z80, and it didn't use the main CPU at all.) The first OS -- called
> POS -- was written by Three Rivers. "POS" stood for "Piece Of Shit."
> It was written for a demo, and it was never intended to be sold as a
> product. The company's founders -- I swear to God, I am NOT making this
> up -- expected to sell computers with *NO* operating system. They were
> prepared to sell and support the micro-assembler; the Pascal microcode,
> libraries, and compiler; and they expected to sell and support the
> bootstrap loader. Everything else was supposed to be up to the
> customer. Nobody below the top level of management was surprised when
> POS was added to the product catalog, and the definition of "POS" was
> changed to "Perq Operating System."
>
> The second OS was Accent. Accent probably was the coolest OS ever.
>
> Let me say that again, it's important.
>
> Accent was the COOLEST OS *EVER* devised by Man!!!!
<snip>
Thanks for the PERQ history.
I used a PERQ system in the1982 to 1985 timeframe. It was a large
brownish box, actually built into the desk. I think it used 8 inch
floppies, and had what looked like a 14 inch hard drive. I also
remember InTran (or possibly Three Rivers - it's been a long time!)
sending us replacement boards when there was a hardware failure, and I
remember replacing the I/O board with the Z80 on it, and the hard drive
controller board.
It was OEMed by InTran, the company we got the machine from. It was a
graphics workstation, connected via ethernet to a Xerox 9700 laser
printer. We used it to create fonts and electronic forms for the
printer, bypassing the functional but extremely time consuming FDL
language on the printer.
I've been trying to identify the model of this machine and the OS that
it used for some time. It had a 4 button "puck" as the mouse, that
you had to use on the graphics tablet. Each button was a different
color - red, green, blue and yellow as I recall. The software we used
was mainly InTran's stuff, used to actually create forms and fonts for
the Xerox printer. However, I do remember using the "shell" command
from a menu that put you at a command prompt. I also remember a disk
utility called "Scavenger", that was probably a defragger of some kind.
While it was running, the "progress indicator" was a bird flying up and
down the screen. It also had a good chess game, and a rather
outstanding pool game!
This thing was quite incredible in it's time. The InTran software was
all menu driven, point and click, with an ethernet connection to the
Xerox 9700 printer. This printer, BTW, was at the time the fastest
cut sheet laser printer in the world - 2 pages per second (120 ppm) -
continuously. I ran these things for several years.
Does any of this ring any bells? Is this enough info for you (or
anyone else) to identify this PERQ and the OS it used?
Ahhhhh, memories.......
Jim
Eric Smith
> The IBM 5100!
> I used one a few times back in the mid-1970s at Scientific Timesharing.
> The screen was like 32x12, but logically it was 64x12 (or maybe it was
> like 80x16, I forget) but you toggled a switch back and forth to see
> the left and right sides of the screen!
No, it's 64x16, which by default are all visible at once. You can flip a
switch to go into 32x16 showing the left half, or the right half.
Unfortunately it doesn't make the characters double width, it just displays
spaces inbetween, so it's not particularly useful.
There was also a switch that made it display the first 512 bytes of RAM
in hexadecimal instead of the normal text buffer. Used for field service.
I'm not sure why they put the switch right on the front panel, rather than
inside the machine like the single step control.
> It also had BASIC in the ROM, but I didn't use that.
They were available with APL, BASIC, or both. The machines with both
actually have a front panel switch to select.
> I was told that the machine was actually emulating an IBM mainframe.
They simulate enough of a 360 to run a modified version of APL\360.
For BASIC they apparently simulate a System/3 or something similar.
> Mass storage on the device was a tape cartridge.
> This was all long before the 68000 though!
Yup. The IBM 5100 was introduced in 1975. It was pretty advanced
for a desktop machine at that time. It was followed by the 5110,
which supported optional 8-inch floppy drives, a few more I/O devices,
and had some improvements to the BASIC interpreter including better string
support. I'm not sure how much they improved the APL. The 5120 was
a repackaging of the 5110 with the disk drives and a bigger monitor
in a single box. The 5120 looks about the same as a System/23
Datamaster.
Eric Smith
> One true UNIX desktop was a 68000-based box from Fortune Systems.
> IIRC it cost around $10K around 1983. I was pretty jealous of my one
> friend who had one.
"Foobar T. Clown" <fu...@gazonk.del> writes:
> When did AT&T market their "UNIX PCs" (e.g., the 3B1, 3B2, etc.)?
The 7300 (later very slightly modified to be a 3B1) was the UnixPC,
and was introduced in 1985. It used the 68010.
The 3B2 family runs Unix, but is NOT a "UnixPC". They use WE 32000
processors. There were also 3B5 and 3B20 boxes.
Pete Fenelon
> Unfortunately it doesn't make the characters double width, it just displays
> spaces inbetween, so it's not particularly useful.
>
in that case! Any suggestions? :)
pete
Floyd Davidson
Certainly the most successful of those mentioned above was
indeed the AT&T 3B2 series of minicomputer using an AT&T
designed WE32000 cpu. There were 3B2 desktop computers used
within AT&T prior to disvestiture in 1984, but that year marks
their entry into the open market.
They went through several models over the next decade, and even
managed a contract to supply the 3B2-600 model to the US Air
Force. As late as 1989 or 1990 the price tag was still up
around $40K each and since it was virtually impossible to sell
one at that price AT&T began tricks like bundling a 3B2 with
certain telco equipment as a "controller" in order to move them
out of the warehouse. The product was discontinued in 1993.
AT&T also produced another minicomputer, the 3B5 (discontinued
in 1990), that predated the 3B2. And several desktop
microcomputers (PC6300, PC7300/3B1, and 386/WGS models).
The 3B5 (and upgraded model, a 3B15) was probably the first
32bit microprocessor machine in production (with a Bellmac-32A
cpu, the first single chip 32bit cpu with a 32 bit bus, which
was released in 1980), but that was prior to the break up of the
Bell System and they were not allowed to market the system to
non-Bell companies until 1984.
--
Floyd L. Davidson <http://www.ptialaska.net/~floyd>
Ukpeagvik (Barrow, Alaska) fl...@barrow.com
Heinz W. Wiggeshoff
THINKing ahead! DBCS - Double Byte Character Set!
Back in the 70's, I used a nifto neato glass terminal that could be
switched to show ASCII control characters as they were received.
Great for debugging intricate displays in text mode.
Andrew Carol
<cst...@swingandcircle.com> wrote:
> The IBM 5100!
> I used one a few times back in the mid-1970s at Scientific Timesharing.
> The screen was like 32x12, but logically it was 64x12 (or maybe it was
> like 80x16, I forget) but you toggled a switch back and forth to see
> the left and right sides of the screen! Of course, short lines were
> fine for APL. It also had BASIC in the ROM, but I didn't use that.
> I was told that the machine was actually emulating an IBM mainframe.
> Mass storage on the device was a tape cartridge.
>
> This was all long before the 68000 though!
I remember visiting the Landstuhl base library in Germany (I was in the
7th grade) when I discovered the IBM 5100 users manual. I must have
checked it out a dozen times.
I was totally in love with the idea of a computer with a high level
language that you could actually carry somewhere.
--- Andy
Peter Maydell
> some really weird number of bits per word (numbers like 34 and 28 and
> 19 keep popping into my head).
CPU registers (what the microcode sees; it can obviously fake any
width register for user code) are 20 bits wide (24 bits on the rare
2-T4). Microinstructions are 48 bits wide. The width of the bus
to the memory board is 16 bits. It's fastest to transfer to or
from memory a quadword at a time (64 bits). The CPU architecture
more-or-less requires that the instruction set exposed to users
is a bytecode of some form.
I think that memory is word addressed (ie location X+1 is 16 bits
forward from location X) rather than byte addressed, but I could
be wrong. I'd check my schematics but I've just moved house and
they're buried in a cardboard box somewhere :-<
[On the bright side I will now have room to actually house the PERQ...]
>A later model expanded the control
> store to 16 K words. There was some kind of interface which allowed
> the Z80 on the I/O processor to write to the control store at boot time,
> but there was also an interface which allowed the CPU to re-write its
> own microinstructions.
I don't think you could write to the control store using the Z80;
at least not with the PERQ 1. At boot up there was a small amount of
ROM mapped into the control store, which contained test and
bootstrapping code; at some point this code would flip a switch
to permanently unmap the ROM in favour of RAM. Most of the microcode
in the control store would be loaded from disk (or from another
PERQ via PERQlink if you were doing serious microcode development.)
I'd be very surprised if the Z80 could access the control store,
since the control store writes are one microcycle per third of a
48bit microinstruction, and transfers across the bus between memory,
IO and CPU boards don't run that fast.
> There were three operating systems used on PERQs that I know of. (No,
> I am not counting Aaron Wohl's port of CPM because that ran entirely on
> the Z80, and it didn't use the main CPU at all.) The first OS -- called
> POS -- was written by Three Rivers. "POS" stood for "Piece Of Shit."
> The second OS was Accent. Accent probably was the coolest OS ever.
I don't think Accent was very popular in the UK; most UK PERQs ran
POS or PNX.
> The third OS was called QNX, and it came out of ICL Ltd. in England.
> I have never seen a PERQ running QNX, but I understand it to be a port
> of something called PNX which I understand was somewhat like UNIX.
> AFAIK, the PNX source code was written in the C programming language.
> I do not know whether the ICL folk wrote their own microcode, or whether
> they wrote a C compiler that targeted the Q-code instruction set.
I think you're confused here. (Although possibly I am). The ICL OS
was called PNX; QNX is something totally different, unless you had
something in mind different from the QNX currently used in embedded
systems. I believe (although I have no evidence) that the version of
PNX used on PERQs is based on 7th Edition UNIX. I don't think the
instruction set used is documented anywhere, and I don't think
it's Q-code. [I assume "PNX" to derive in the obvious way from "PERQ UNIX".]
> ICL went on to develop something called the Perq 3B (Perq II was just a
> new skin wrapped around the same old circuit boards), while the third
> company completed the design of what would have been the Perq 3A except
> they named it something else. I know nothing of the Perq 3B, but the
> Perq 3A was, TA DA!, A 68020 machine running real, honest-to-God, AT&T
> UNIX. (It was intended to be a whole lot more than that, but that's
> another story.)
There seems to be some dispute over what ought to get the PERQ 3a
label. The machine I call a PERQ 3a is what ICL called the
3300 Advanced Graphics Workstation:
http://www.chiark.greenend.org.uk/~pmaydell/hardware/carenath/
which I think is what you're describing as a 3B. It never got anywhere;
apparently ICL produced about 30 prototypes for development purposes
but canned the project before completion. The OS is PNX 300, a variant
of SysVR2; I'd love to have the sources to this but I don't think
they exist any more :-<
-- PMM
Foobar T. Clown
>
> Sun Microsystems was founded in 1982 and I'm thinking it was not
> the only outfit building Unix systems around the Stanford
> SUN design.
I can still feel the eye strain just THINKING about that 60Hz
interlaced display.
-- Foo!
Foobar T. Clown
>
> Thanks for the PERQ history.
Yer welcome. I just wish I could remember some dates to go with it.
Oh well, it was ALL 1980-something.
-- Foo!
Foobar T. Clown
>
> Foobar T. Clown <fu...@gazonk.del> wrote:
> > The control store was writable. It was orgainized as 4K words with
> > some really weird number of bits per word (numbers like 34 and 28
> > and 19 keep popping into my head).
>
> CPU registers (what the microcode sees; it can obviously fake any
> width register for user code) are 20 bits wide (24 bits on the rare
> 2-T4).
I never knew it by that name, but there was a time I could have told you
more about that 24-bit CPU than you'd ever have wanted to know. I am
the person who modified Accent so it could use the extra four bits of
physical address space, and I am the person who changed the microcode
for the multiply instructions to use the "multiply step" hardware.
> I think that memory is word addressed (ie location X+1 is 16 bits
> forward from location X) rather than byte addressed, but I could
> be wrong.
You are not wrong. The original PERQ could access up to 1M word (what
we would call 2M Bytes today) with its 20-bit physical address space.
The 24-bit PERQ could have addressed up to 16-M words, but AFAIK, no-
one ever built a memory board that big.
> I don't think you could write to the control store using the Z80;
> at least not with the PERQ 1. At boot up there was a small amount of
> ROM mapped into the control store, which contained test and
> bootstrapping code; at some point this code would flip a switch
> to permanently unmap the ROM in favour of RAM.
Oh! O.K. The I/O processor always was a complete mystery to me.
> > The third OS was called QNX, [...] I have never seen a PERQ running
> > QNX, but I understand it to be a port of something called PNX [...]
>
> I think you're confused here. (Although possibly I am). The ICL OS
> was called PNX; QNX is something totally different
Like I said, I've never seen a PERQ running it. You have. I probably
am the one who is confused.
> > ICL went on to develop something called the Perq 3B [...] while
> > the third company completed the design of what would have been the
> > Perq 3A except they named it something else.
>
> There seems to be some dispute over what ought to get the PERQ 3a
> label. The machine I call a PERQ 3a is [...] I think is what you're
> describing as a 3B.
O.K., then the one I called the 3A probably was the 3B. It never went
anywhere either except into the innards of a machine that was never
sold as a general purpose workstation.
-- Foo!
Foobar T. Clown
>
> Unisoft did many, many ports of V7 before the 68010 was released.
> Dual systems had a 68K S100 board that ran it. It also ran on Lisa.
I remember using a system that had a 68000 and Unisoft UNIX. The main
CPU and the memory were on a full-length ISA board in a genuine IBM
PC-AT. The IBM PC functioned as the I/O processor.
-- Foo!
Tony Duell
: Christopher Stacy wrote:
: >
: > [...] Simon Bowring ("Simon") writes:
: > Simon> Sorry if this is a FAQ.... What was the very first
: > commercially available DESKTOP Unix box?
: >
: > Three Rivers Corporation's PERQ was out in 1980 or 81, [...] It had
: > a bitmapped display, mouse, network, and 68000 based processor
: > running some kind of UNIX. It was a CMU thing.
: All true, except that it didn't use a 68000, and it didn't run UNIX (at
Well, apart from the PERQ AGW3300 (aka 3A, IIRC), which is a 68020-based
unix box. It's significantly different from other PERQs, so I generally
regard it as a PERQ in name only.
: There were two processors in the original PERQ I workstation; the main
: processor and the I/O processor. The I/O processor was a Z80, and it
: handled disk requests, two RS-232 ports, the keyboard, the mouse, and
Did the PERQ 1 have 2 RS232 ports? I've not got the docs to hand, but I
only remember one of them.
The I/O processor handled the floppy drive and on the PERQ 1, the head
positioning for the hard disk. The hard disk data went via a hardwired
interface based round an AMD2910 sequencer.
On the PERQ 1, the Z80 had 4K of RAM, and there was no way to download
your own code to it. The only software it ran was contained in EPROMs on
the I/O board. On the PERQ 2, there was 16K of RAM on the Z80 and a way
to download user programs to it. In fact much of the I/O software is
_not_ in EPROM on the PERQ 2 -- it's downloaded when the machine is booted.
: a GPIB bus. The I/O processor and all of its peripheral interfaces
: occupied one slot in the backplane (circuit boards were approximately
: 18 inches by maybe 12 inches.) The original I/O board was quickly
They were more square than that. It's actally a standard form-factor
board (and edge connectors) -- the boards in my Dylon GPIB magtape
controller are the same physical size as those in the PERQ.
: replaced by one which also contained a three megabit ethernet interface.
: IIRC, the so-called EIO board was a ten layer circuit board (ugly).
: The main CPU occupied another whole backplane slot all by itself. It
: was built mostly from 74LSxxx TTL. There was one LSI part: A micro-
The 4K CPU (PERQ 1) used a few small PROMs for things like state machines
and function decoding. The 16K CPU (PERQ 1a and all PERQ 2s) also has a
fair number of PALs on the CPU board. Along (in both cases) with a lot of
74S TTL.
: sequencer which (I think) was made by AMD, and there was a whole bunch
Yes, an AMD2910 sequencer. A rather poor choice as it turned out. This
chip would generate a 12 bit address, which was ideal for the 4K control
store in the PERQ 1. The problem came when the CPU was upgraded to 16K.
There is no real way to extend the 2910. The result was a circuit
commonly known as the '2 bit kludge' -- pun totally intentional. This
generated the top 2 bits of the 14 bit address for the 16K control store
(a 2910 was used for the bottom 12 bits). But there could be no carry
from the low 12 bits to the top 2 bits (so code near the top of one 4K
block would wrap around to the bottom of that block, rather than
continuing in the next block). There were 3 types of jump -- short jumps
(changing the bottom 8 bits of the address only), long jumps (changing
the bottom 12 bits) and leaps (changing all 14 bits).
: of 70nS static RAM for the control store.
: The control store was writable. It was orgainized as 4K words with
: some really weird number of bits per word (numbers like 34 and 28 and
: 19 keep popping into my head). A later model expanded the control
No, 48 bits wide...
: store to 16 K words. There was some kind of interface which allowed
: the Z80 on the I/O processor to write to the control store at boot time,
No. The only device that can write to the control store is the main
(microprogrammed) CPU. The way it's done is a bit of a hack, and to
understand it requires a fair understanding of how the 2910 behaves.
Basically, the address to be changed is written to the S register in the
2910. The control store write instruction has to have particular values
in the condition and jump fields. At first, the condition input to the
2910 is held in one state, so that the 2910 outputs the contents of the S
register. The CPU puts the new data to be written onto the R bus and then
asserts the write line to the control store.
At the end of that microcycle, the instruction latches are _not_
reloaded. The same instruction tries to execute again, but this time,
most of the CPU is inhibited, so no data is updated, the control store is
not written to, and so on. But the condition input to the 2910 is now
held in the opposite state, and this time the contents of the normal
microcode program counter are output to the control store address lines.
At the end of this microcycle, the next instruction is fetched from the
control store, and the microcode continues.
So, how does the machine boot? There's a 512 word bootstrap ROM that
overlays the first 512 locations of the control store. It contains the
CPU diagnostics, and a bootstrap that tries to load microcode from the
PERQlink (16 bit parallel interface), floppy drive and hard drive (in
that order). The initial microcode is read in and stored in the control
store, and as soon as it is executed the bootstrap ROM is disabled until
the next hardware reset.
: I personally know of two different instruction sets that were
: implemented in microcode. The first one to be used was called the Q-
: code instruction set. It was more or less a straightforward implemen-
: tation of the UCSD P-Code virtual machine. (Most PERQ software was
: written in Pascal, and compiled with the UCSD compiler.) The other one
: I know of was the C-MU SPICE Lisp instruction set. There may have been
: a third, but more on that below.
There is a third one. Called C-code (normally), it was used with PNX,
ICL's port of Unix v.7
: The third OS was called QNX, and it came out of ICL Ltd. in England.
: I have never seen a PERQ running QNX, but I understand it to be a port
: of something called PNX which I understand was somewhat like UNIX.
AFAIK QNX and PNX are totally different. PNX is essentially PERQ unix,
and does have its own microcode instruction set.
-tony
Tony Duell
: In article <3C50D828...@gazonk.del>, Foobar T. Clown
: <fu...@gazonk.del> wrote:
: I used a PERQ system in the1982 to 1985 timeframe. It was a large
: brownish box, actually built into the desk. I think it used 8 inch
The PERQ1s have dark brown front and back covers (and beige sides). The
PERQ2s are a much lighter brown/beige all over.
: floppies, and had what looked like a 14 inch hard drive. I also
All PERQs other than the PERQ AGW3300 use 8" floppies. The PERQ 1s have the
floppy drive horizontal, the PERQ 2s have it vertical. A 14" (SA4000
series) hard disk pretty much identifies the unit as a PERQ 1 or 1a. The
PERQ 2T1 has an 8" Micropolis 1200 drive at the back, the PERQ 2T2 and
PERQ2T4 have 5.25" ST506-interfaced drives, also at the back.
: remember InTran (or possibly Three Rivers - it's been a long time!)
: sending us replacement boards when there was a hardware failure, and I
Real men repair their own PERQ boards to component level :-). Actually
they're quite simple when you get used to them.
: remember replacing the I/O board with the Z80 on it, and the hard drive
: controller board.
There is no separate hard disk controller board in any PERQ 1 or 2. The
hard disk controller is on the I/O baord. PERQ 2s have what's called a
'DIB' (Disk Interface board) that sits between the I/O board (the
interface to which is similar to that used by the 14" SA4000 drives) and
the 8" or 5.25" hard disk. And of course there are several PCBs on the
hard disk itself
: I've been trying to identify the model of this machine and the OS that
: it used for some time. It had a 4 button "puck" as the mouse, that
: you had to use on the graphics tablet. Each button was a different
The tablet was a Sumagraphics Bit Pad 1, GPIB version. It was the only
available pointing device on the PERQ 1s (there were others planned, but
AFAIK none were shipped). PERQ 2s have a thing called a Kriz mouse which
is 3 button, but which also has to work on its special tablet (about the
size of a modern mousmat). The Bit Pad 1 also works on the PERQ 2s, though
: color - red, green, blue and yellow as I recall. The software we used
: was mainly InTran's stuff, used to actually create forms and fonts for
: the Xerox printer. However, I do remember using the "shell" command
: from a menu that put you at a command prompt. I also remember a disk
: utility called "Scavenger", that was probably a defragger of some kind.
: While it was running, the "progress indicator" was a bird flying up and
It's normally claimed to be a Vulture :-)
Scavenger is used under POS and Accent, I think. I don't recognise the
'shell' command, so it was either Accent (which I've never used) or a custom
commadn in your application.
: down the screen. It also had a good chess game, and a rather
: outstanding pool game!
-tony
Jim
Bowring <sbowrin...@mpc-data.co.uk> wrote:
Well, I'm certainly no expert on the subject, but this is the first
time I've heard that story (Rick Rashid/Mach).
I *do* know that Dave Cutler was hired by Microsoft, however, and did
lead the design of Windows NT.
Jim
Ernest Cline
"Pete Fenelon" <pe...@fenelon.com> wrote in message
news:u537rk7...@corp.supernews.com...
>
> What was the Tandy thing with a 68k running Xenix and a Z80 doing the
> I/O? That was out very early in the 80s....
IIRC, I believe it was called the TRS-80 Model 4.
John F. Eldredge
Hash: SHA1
On 25 Jan 2002 10:54:31 -0800, Eric Smith
<eric-no-s...@brouhaha.com> wrote:
I used to own a 3B1. I remember that it had a 40 Megabyte hard
drive; if you installed the entire development system and
text-processing system (which I did), you still had 20 Megabytes
free. Nowadays, the on-line documentation alone would probably be
more than that. The last Linux distro I bought came on 5 CD's. The
machine was also surprisingly fast, given that it had only 1 Megabyte
of RAM. Of course, it was operating in text mode, not running
X-Windows. I finally sold the 3B1, a couple of years ago, to someone
who had a client who was still running his office off a 3B1, and
needed to have a spare machine to cannibalize for parts.
-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.1 Int. for non-commercial use
<http://www.pgpinternational.com>
iQA/AwUBPFI4ejMYPge5L34aEQKwqgCePqmVI4BxXQTMDCwWg2a53ZUj/R8An3Fx
SFRt9mhdoHVKMMAGpDjahmSi
=T/ja
-----END PGP SIGNATURE-----
--
John F. Eldredge -- new address jo...@jfeldredge.com
eldr...@earthlink.net, eldr...@poboxes.com still work
PGP key available from http://pgpkeys.mit.edu:11371
"There must be, not a balance of power, but a community of power;
not organized rivalries, but an organized common peace."
Woodrow Wilson
jmfb...@aol.com
"Foobar T. Clown" <fu...@gazonk.del> wrote:
<grin> What was it about that decade when everything seems to
run together?
/BAH
Subtract a hundred and four for e-mail.
Brian {Hamilton Kelly}
cgi...@sky.bus.com "Charlie Gibbs" writes:
> In article <GqGDy...@world.std.com> l...@TheWorld.com
> (Larry M Headlund) writes:
> >Sounds like the HP Integral. This was a portable (lugable) computer
> >about the size and weight of a portable sewing machine.
>
> I heard it described as a "lunchbox".
Nothing to do with Lindford Christie's, I sincerely hope?
--
Brian {Hamilton Kelly} b...@dsl.co.uk
"We have gone from a world of concentrated knowledge and wisdom to one of
distributed ignorance. And we know and understand less while being incr-
easingly capable." Prof. Peter Cochrane, formerly of BT Labs
Stephen Neal
daven...@sourplum.org.uk wrote in message ...
Didn't the 16032 (or was it 32016) 2nd processor and Cambridge Workstation
(the one surviving ABC) ran PanOs? (So called I believe after the
restaurant of the same name in Cambridge?) Was this related to Tripos? (I
thought Tripos was the precursor to the Amiga OS?)
May be all wrong!
Steve
>
>Dave
Pete Fenelon
> Didn't the 16032 (or was it 32016) 2nd processor and Cambridge Workstation
Same chip, IIRC -- Nat Semi renamed them because 32016 sounded more
impressive. :)
(Wasn't 32032 the version with a full 32-bit bus?)
pete
Pete Fenelon
> The last Linux distro I bought came on 5 CD's.
Sigh. And 99% of that is crudware. Dozens of "file managers", KDE/GTK
front ends to things that work perfectly well in a terminal, themes that
just make the machine look ugly... feature bloat. Punters buy the box
that seems to have the most megabytes of software per dollar.
(I think the SUSE distro we bought to install a couple of servers at
work was six CDs or a fairly full DVD. Shocking. My first Linux system
(0.12 with Poe's IGL and pretty much *everything* that had been
ported) sat comfortably in 30MBytes at the end of an old 85 meg IDE
disc. and had room to work in. Oh, and it ran in 4 meg!) -- which was of
course twice (or was it four times) the RAM of the VAX-11/750 that
could support 30 of us at once a few years before :/
Anne & Lynn Wheeler
> (I think the SUSE distro we bought to install a couple of servers at
> work was six CDs or a fairly full DVD. Shocking. My first Linux system
> (0.12 with Poe's IGL and pretty much *everything* that had been
> ported) sat comfortably in 30MBytes at the end of an old 85 meg IDE
> disc. and had room to work in. Oh, and it ran in 4 meg!) -- which was of
> course twice (or was it four times) the RAM of the VAX-11/750 that
> could support 30 of us at once a few years before :/
or CP/67 running in 768k real storage (104 4k pages for paging after
fixed kernel requirements), 75 users, mix-mode operation, interactive,
program development, test, apl modeling (sort of stuff frequently done
w/spreedsheets today) and various kinds of batch, 95 percentile
subsecond response time for trivial interactive operations (and I
believe possibly somewhat slower processor than 750?).
random refs:
http://www.garlic.com/~lynn/94.html#43 Bloat, elegance, simplicity and other irrelevant concepts
http://www.garlic.com/~lynn/94.html#55 How Do the Old Mainframes Compare to Today's Micros?
http://www.garlic.com/~lynn/98.html#12 S/360 operating systems geneaology
http://www.garlic.com/~lynn/2001e.html#45 VM/370 Resource Manager
http://www.garlic.com/~lynn/2001f.html#47 any 70's era supercomputers that ran as slow as today's supercomputers?
http://www.garlic.com/~lynn/2001h.html#26 TECO Critique
http://www.garlic.com/~lynn/2001l.html#6 mainframe question
--
Anne & Lynn Wheeler | ly...@garlic.com - http://www.garlic.com/~lynn/
Chris Hedley
> Sigh. And 99% of that is crudware. Dozens of "file managers", KDE/GTK
> front ends to things that work perfectly well in a terminal, themes that
> just make the machine look ugly... feature bloat. Punters buy the box
> that seems to have the most megabytes of software per dollar.
I quite like KDE... well, to qualify that, I dislike KDE less than
other GUI environments. I find that the graphical stuff has its uses,
although I agree that terminal-based applications are, by and large,
both better and easier to use (hence my email thingy I mentioned else-
where)
Chris.
Pete Fenelon
I'd rather someone banged all the script-kiddy heads together and got
them to build *one* decent application in each area rather than 50
pieces of halfbaked frippery.
KDE is fairly sound, deep down. If I use a Linux GUI, it's been KDE for
a long time. Gnome, with that abomination and themester's paradise
Enlightenment, seems to delight in wasting resources on cosmetics.
At least KDE is *relatively* economical out of the box.
I actually *liked* the NeXT GUI, because it felt good and wasn't
flashy, but it *ate* CPU and RAM on 'white hardware' and my NeXTstation
had a palatial 100Mbyte disc :P
pete
Dowe Keller
> John F. Eldredge <jo...@jfeldredge.com> wrote:
> > The last Linux distro I bought came on 5 CD's.
>
> Sigh. And 99% of that is crudware. Dozens of "file managers", KDE/GTK
> front ends to things that work perfectly well in a terminal, themes that
> just make the machine look ugly... feature bloat. Punters buy the box
> that seems to have the most megabytes of software per dollar.
Amen brother!
> (I think the SUSE distro we bought to install a couple of servers at
> work was six CDs or a fairly full DVD. Shocking. My first Linux system
> (0.12 with Poe's IGL and pretty much *everything* that had been
> ported) sat comfortably in 30MBytes at the end of an old 85 meg IDE
> disc. and had room to work in. Oh, and it ran in 4 meg!) -- which was of
> course twice (or was it four times) the RAM of the VAX-11/750 that
> could support 30 of us at once a few years before :/
My first distro I installed by ftp from Debian over a 33.6 kbps modem
connection, and ran just fine with X-Windows on my 120MHz 8MB box. My
present distro is SuSE, as setup default on the same 120MHz 5x86 with
24MB ram, it barely would be usable. I removed all those crap GUIs
and run it with no problem using fvwm as my wm, and lynx for my web
browser.
I imagine that a properly paired down Linux or mumbleBSD box with
enough serial ports would handle timesharing just fine. Nobody forces
users to run KDE and Netscrape!
OBFolklore:
On the other hand, V7 UNIX is so small, that I can run it in a virtual
pdp11 on my system and still get decent response.
An enlightening experience for someone who was weaned into UN*Xen via
Minix. I have learned that Minix is a V7 clone.
'Tis a pity that OSes seem to aquire cruft as they aquire power :-(.
--
do...@sierratel.com Home: http://www.sierratel.com/dowe
"generally, you are wrong. there might be specific platforms for which
it's true, though. for that matter, there might be chartreuse cows."
-John Gordon, comp.unix.programmer
Morten Reistad
Pete Fenelon <pe...@fenelon.com> wrote:
>John F. Eldredge <jo...@jfeldredge.com> wrote:
>> The last Linux distro I bought came on 5 CD's.
>
>Sigh. And 99% of that is crudware. Dozens of "file managers", KDE/GTK
>(I think the SUSE distro we bought to install a couple of servers at
>work was six CDs or a fairly full DVD. Shocking. My first Linux system
>(0.12 with Poe's IGL and pretty much *everything* that had been
>ported) sat comfortably in 30MBytes at the end of an old 85 meg IDE
>disc. and had room to work in. Oh, and it ran in 4 meg!) -- which was of
>course twice (or was it four times) the RAM of the VAX-11/750 that
>could support 30 of us at once a few years before :/
4 meg on linux and 2 on VMS was too little to run any applications
comfortably. I tried moving around some financial real-time code
that was VERY tight in the 1987-1993 timeframe. It was designed
to run on late-seventies minis; and I re-implemented most of it
in modern environments (C, unix-style, but it ran fine on Prime,
VAXen and Linux). The tarball file is around 150k.
The test was to run a trading full day in real-time mode without more than
one second delay for any transaction; and that it could co-exist
with other uses of the machine.
Primos of the time could run in 2m, but needed 6m for a smooth run.
(midas file handling).
VMS could also run in 2m, but needed 8m for smooth running; probably
because I invoked RMS for the file handling.
Linux could sort of run in 2m, but really 4m was needed, and I only
got a smooth running system at 8m (I was unable to test 6m).
(gdbm for file handling).
This was a linux 'initial distribution' starting with 0.98 something
and ending with 1.09; upgrading further would have meant a reinstall
and the system had gotten cult value by then. I still have it in
storage.
This was in line with other experiences in the time period. 2m machines
gave lame performance. 16 meg was a luxury.
Then came windowing systems, networking and browsers; and whoosh;
the current laptop tends to page with 64m ram if I start all the
applications I lake to run at the same time.
- mrr
Jim
<a...@p850ug1.demon.co.uk> wrote:
> Jim (edu...@com.net) wrote:
> : In article <3C50D828...@gazonk.del>, Foobar T. Clown
> : <fu...@gazonk.del> wrote:
>
> : I used a PERQ system in the1982 to 1985 timeframe. It was a large
> : brownish box, actually built into the desk. I think it used 8 inch
>
> The PERQ1s have dark brown front and back covers (and beige sides). The
> PERQ2s are a much lighter brown/beige all over.
It *was* dark brown! Chocolate brown, I'd say.
> : remember InTran (or possibly Three Rivers - it's been a long time!)
> : sending us replacement boards when there was a hardware failure, and I
>
> Real men repair their own PERQ boards to component level :-). Actually
> they're quite simple when you get used to them.
Well, I do plenty of component-level repairs of boards these days, but
I wasn't up to it 18 years ago! :-)
>
> : remember replacing the I/O board with the Z80 on it, and the hard drive
> : controller board.
>
> There is no separate hard disk controller board in any PERQ 1 or 2. The
> hard disk controller is on the I/O baord. PERQ 2s have what's called a
> 'DIB' (Disk Interface board) that sits between the I/O board (the
> interface to which is similar to that used by the 14" SA4000 drives) and
> the 8" or 5.25" hard disk. And of course there are several PCBs on the
> hard disk itself
It's possible my memory is wrong. I just remember replacing a board
when we had hard drive problems. We did *not* replace the drive
itself, however.
> : While it was running, the "progress indicator" was a bird flying up and
>
> It's normally claimed to be a Vulture :-)
Yes! It *was* a vulture!
> Scavenger is used under POS and Accent, I think. I don't recognise the
> 'shell' command, so it was either Accent (which I've never used) or a custom
> commadn in your application.
It could easily have been a custom command. It was just a menu choice
somewhere, that when you clicked it, you were dumped into a command
line terminal. I don't recall now if you had to reboot or type
"exit" or something to get back to the main system.
I *do* remember being extremely impressed by the machine. The
software was positively 21st century stuff back then, and saved me
hundreds of hours of time. Also, I had never seen or heard of a Perq
(or Three Rivers) before then. I've also never seen or heard about
one since.
Until now!
Jim
Don Quixote
>VMS could also run in 2m, but needed 8m for smooth running; probably
> because I invoked RMS for the file handling.
You'll always need more once you invoke RMS.
Brian Inglis
<sbowrin...@mpc-data.co.uk> wrote:
>Sorry if this is a FAQ....
>
>What was the very first commercially available DESKTOP Unix box?
>
>I worked at the UK company Cifer during 1983-85 when we launched
>a 68000 based Unix system III desktop model. We did the port
>jointly with a company called root computers (I think, "root"
>anyway).
>
>Our early systems had a wopping 256kB of RAM. though we quickly
>went to 1Mb - the process-granular virtual memory meant that 256kB
>was just not enough!
>
>ISTR HP running adverts AFTER our product releases claiming they
>had the first desktop Unix systm - I think it had an integrated printer
>and may have been called the 9000 series (although I think they've
>subsequetly reused this designation)
When did OS/9 (Unix-alike) for the Motorola 6809 (e.g. RS Color Computer) become
available, and were there any predecessors for earlier 6800 CPUs? The 6800 used
an assembler language similar to the PDP-11 so someone could have ported an
early Unix version much earlier than the 68000 workstations mentioned.
--
Thanks. Take care, Brian Inglis Calgary, Alberta, Canada
Brian....@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca)
fake address use address above to reply
tos...@aol.com ab...@aol.com ab...@yahoo.com ab...@hotmail.com ab...@msn.com ab...@sprint.com ab...@earthlink.com ab...@cadvision.com ab...@ibsystems.com u...@ftc.gov
spam traps
Brian Inglis
<sbowrin...@mpc-data.co.uk> wrote:
>On Fri, 25 Jan 2002 08:51:17 -0700, Brian Inglis wrote:
>
>>From Salus "A Quarter Century of Unix" p133:
>>
>>Among the users who had helped himself was Haruhisa Ishida at the
>>University of Tokyo. "I was the first user and licensee of UNIX
>>in Japan," he told me. "It was in 1976 that I signed the
>>agreement with AT&T Bell Labs. The Unix was version 6, which I
>>implemented on an LSI-11 with only slight modification."
>>
>>The DEC LSI-11 was the desktop micro version of the PDP-11, so I
>>think we have a winner, unless someone has an earlier instance.
>
>That's certainly the best so far. Definately Unix and definately
>Desktop (but not sure about "commercial")!
I interpreted commercial as meaning desktop hardware bought from a vendor rather
than home brewed. Unix was commercial as it had to be licensed, but not exactly
a packaged OS that you could just load and boot, at least in that timeframe.
Pete Fenelon
I was talking Unix when I referred to the 750 - my first exposure to VMS
was on an 8200 some time later :)
pete
Pete Fenelon
> On the other hand, V7 UNIX is so small, that I can run it in a virtual
> pdp11 on my system and still get decent response.
Yeah, I've got V5 to V7, ITS, and a pile of PDP-8 and PDP-11 OSes on
here under Bob Supnik's excellent simulator :)
> An enlightening experience for someone who was weaned into UN*Xen via
> Minix. I have learned that Minix is a V7 clone.
Hmmmmm.... it's v7 *mostly* at the API level, but internally it's
Tanenbaum's message-passing architecture, and userland is rather
different. Feels a bit like working on a PDP-11 with separate I&D,
slightly. I used to run Minix on Ataris - remember it taking most
of an evening to rebuild the kernel. A whole meg of RAM, though,
and I think I gave it three quarters of my 20 Mbyte hard disc, when
I finally bought one (oh, running it from floppies *was* fun!). I
fell off around 1.5.10, IIRC, when I went Intel. Soon after, some
Finnish chap said he'd started hacking a new kernel.... :)
pete
William Hamblen
<Brian....@SystematicSw.ab.ca> wrote:
>When did OS/9 (Unix-alike) for the Motorola 6809 (e.g. RS Color Computer) become
>available, and were there any predecessors for earlier 6800 CPUs?
OS-9 was a joint venture with Motorola and was born at the same time
as the MC6809. It amazed me they managed to sell OS-9 to Tandy, as
OS-9 was conceived as an industrial real time system. There wasn't
anything similar for the 6800. Technical Systems Consultant's Flex
operating system was popular for hobby 6800 systems. Motorola did
have a native disk development system that I know nothing about.
Pete Fenelon
> OS-9 was a joint venture with Motorola and was born at the same time
> as the MC6809. It amazed me they managed to sell OS-9 to Tandy, as
> OS-9 was conceived as an industrial real time system. There wasn't
> anything similar for the 6800. Technical Systems Consultant's Flex
> operating system was popular for hobby 6800 systems. Motorola did
> have a native disk development system that I know nothing about.
>
EXORMACS rings a bell - ISTR Motorola built some quite nice 6809-based
development systems running something called something very similar.
pete
Ben Clifford
> I recall torch having a Z80 based BBC 2nd processor in that time frame,
> but that would have been CP/M based
Was this compatible with a BBC hooked up to a z80 second CPU box?
We had one of those at a local computer club in Guildford, but we didn't
have any software to run on it (not even an OS).
--
Ben Clifford
http://www.hawaga.org.uk/ben/ GPG: 30F06950
webcam: http://barbarella.hawaga.org.uk/benc-cgi/watchers.cgi
Do not ever send e-mail to: phi...@hawaga.org.uk (seriously!)
Ben Clifford
> I can still feel the eye strain just THINKING about that 60Hz
> interlaced display.
Sounds like an Amiga - although it was 50Hz where I came from, I think.
Ben Clifford
> I had a large hand in the Z80 based
> original which came out in mid '82 (I can place this with precision having
> escaped university in '81).
The first computer I ever used - somewhere between 1982 and 1984, at my
father's office. I have a fuzzy recollection of playing snake on it, and
it being able to say "ow" out loud when you crashed into things - can
anyone confirm (or deny) that existed?
When it wasn't being used for such entertainment purposes, it had the
early 80s equivalent of a fax-modem, some giant box that would send out
telexes based on data retrieved from an EU/EC/EEC database.
Alan J. Wylie
> The third OS was called QNX, and it came out of ICL Ltd. in England.
> I have never seen a PERQ running QNX, but I understand it to be a
> port of something called PNX which I understand was somewhat like
> UNIX. AFAIK, the PNX source code was written in the C programming
> language. I do not know whether the ICL folk wrote their own
> microcode, or whether they wrote a C compiler that targeted the
> Q-code instruction set.
I can remember one of these being brought to my company for a demo,
and given to me to play with[1] sometime in the early 80's. We were
rather unsure about the wisdom of the naming of the O/S. IIRC,
it was a bit-mapped display - portrait format?, and if you ran
a Fortran compilation, part of the display memory was re-allocated
as memory for the compiler, causing pretty blinkenlights on the screen.
[1] i.e. break.
--
Alan J. Wylie http://www.glaramara.freeserve.co.uk/
"Perfection [in design] is achieved not when there is nothing left to add,
but rather when there is nothing left to take away."
Antoine de Saint-Exupery
Tony Duell
: In article <a2t0ql$2...@p850ug1.demon.co.uk>, Tony Duell
: <a...@p850ug1.demon.co.uk> wrote:
: > Jim (edu...@com.net) wrote:
: > : In article <3C50D828...@gazonk.del>, Foobar T. Clown
: > : <fu...@gazonk.del> wrote:
: >
: > : I used a PERQ system in the1982 to 1985 timeframe. It was a large
: > : brownish box, actually built into the desk. I think it used 8 inch
: >
: > The PERQ1s have dark brown front and back covers (and beige sides). The
: > PERQ2s are a much lighter brown/beige all over.
: It *was* dark brown! Chocolate brown, I'd say.
Then it was almost certainly a PERQ 1 or PERQ 1a. These are essentially
the same machine except that the PERQ 1 has the original 4K WCS CPU, the
1a has the 16K WCS CPU. Many PERQ 1s were upgraded to 1as by a CPU board
replacemnet.
: Well, I do plenty of component-level repairs of boards these days, but
: I wasn't up to it 18 years ago! :-)
Amazing :-). Most people (not me, I hasten to add) have gone the other way,
from doing compoennt level repair to just board swapping. Oh well...
: > the 8" or 5.25" hard disk. And of course there are several PCBs on the
: > hard disk itself
: It's possible my memory is wrong. I just remember replacing a board
: when we had hard drive problems. We did *not* replace the drive
: itself, however.
Back then, the PCBs on the drive mechanism itself (down the left hand
side of a PERQ 1) were not in any way 'keyed' to the HDA (Head Disk
Assembly). If you had disk drive problems then even board-swappers would
try replacing the PCBs on the drive mechanism first. The idea was to
allow the user to keep the same disks, so as to keep the data, if at all
possible.
I was wondering if you replaced the seek control board, or something, on
the disk drive. It's a simple enough thing to remove....
Incidentally, on the Micropolis 1200 drive used in the 2T1, there's a
PROM on the digital board that contains the bad track map, etc. If you
replace the board you have to pull the PROM from the socket on the old
board and put it into the new board.
: > : While it was running, the "progress indicator" was a bird flying up and
: >
: > It's normally claimed to be a Vulture :-)
: Yes! It *was* a vulture!
And off couse the 'busy bee' that flew around the screen when the machine
was doing certain operations.
: I *do* remember being extremely impressed by the machine. The
I'm still impressed by them. That's probably why I have 4 of the little
toys :-)
-tony
Dowe Keller
> Yeah, I've got V5 to V7, ITS, and a pile of PDP-8 and PDP-11 OSes on
> here under Bob Supnik's excellent simulator :)
I'm partial to tss8 and UNIX V7 myself, of course I might like ITS if
I could find somewhere to download the apps, as I have only the barest
of systems, but yea, simh is super neeto!
> > An enlightening experience for someone who was weaned into UN*Xen via
> > Minix. I have learned that Minix is a V7 clone.
>
> Hmmmmm.... it's v7 *mostly* at the API level, but internally it's
> Tanenbaum's message-passing architecture, and userland is rather
> different. Feels a bit like working on a PDP-11 with separate I&D,
> slightly. I used to run Minix on Ataris - remember it taking most
> of an evening to rebuild the kernel. A whole meg of RAM, though,
> and I think I gave it three quarters of my 20 Mbyte hard disc, when
> I finally bought one (oh, running it from floppies *was* fun!). I
> fell off around 1.5.10, IIRC, when I went Intel. Soon after, some
> Finnish chap said he'd started hacking a new kernel.... :)
In regards to the message-passing part, yea I knew it was a toy
microkernel system. I remember seeing some flamage between Tanenbaum
and Torvalds, in regards to the laters choice of a monolithic design.
but from the user level, it was very sparse and clean like V7, I think
the filesystem was similar.
BTW I currently have a 286 laptop that I run Minix on from floppies.
If your careful, it is suprising how much usability you can cram on a
couple of floppies (If only I could get the weird propriatery HD to
work, I'd really be set :-).
William Donzelli
> I was told that the machine was actually emulating an IBM mainframe.
Not emulating, but maybe talking to...
The 5100 could be used as a terminal to an S/3, and probably S/360 and
S/370.
I am not sure what was involved - I would have dig thru some docs that
I probably can't find anyway.
William Donzelli
Bernd Felsche
>In article <u561kli...@corp.supernews.com>,
>Pete Fenelon <pe...@fenelon.com> wrote:
>>John F. Eldredge <jo...@jfeldredge.com> wrote:
>>> The last Linux distro I bought came on 5 CD's.
>>Sigh. And 99% of that is crudware. Dozens of "file managers", KDE/GTK
>[snip]
>>(I think the SUSE distro we bought to install a couple of servers at
>>work was six CDs or a fairly full DVD. Shocking. My first Linux system
>>(0.12 with Poe's IGL and pretty much *everything* that had been
>>ported) sat comfortably in 30MBytes at the end of an old 85 meg IDE
>>disc. and had room to work in. Oh, and it ran in 4 meg!) -- which was of
>>course twice (or was it four times) the RAM of the VAX-11/750 that
>>could support 30 of us at once a few years before :/
I'm pretty sure you don't actually need to install all of the SuSE
distro containing 2500 packages. :-) In fact, you can put together a
working system (sans X) in under 100MB. You have to work very hard
at it though - the "Minimal" system config provided is anything but
minimal. Emacs on a minimal system? You can then strip out about
20MB of modules you'll probably never use.
Relatively easy to get an installed system under 200MB if you only
install the necessary man pages instead of ALL the docs. I use SuSE
distro for firewall (packet filter and select proxies) so it's
important to not install anything that's not necessary for the
system to work.
A 500MB system even has headroom to allow a reasonable squid cache
for a small site.
You will need RAM though. 16MB is *really* pushing it for a 2.4.x
kernel unless you config and build your own. 32MB works for the
pre-built kernels. If you want to build your own kernel, allow 300MB
of drive space.
Having, in the distant past used various systems with as little as
256kBytes of RAM, 10MB of hard disk and 20 users, I'm well aware of
how the magnitudes have changed. In 1985, a small Unix system
weighed in with 50MB disk and 2MB RAM with a 10MHz 68010 for half a
dozen users. Computer have gotten 100 times faster and bigger; the
users seem to get less work done.
--
/"\ Bernd Felsche - Innovative Reckoning, Perth, Western Australia
\ / ASCII ribbon campaign | I'm a .signature virus!
X against HTML mail | Copy me into your ~/.signature
/ \ and postings | to help me spread!
Christopher Stacy
>>Christopher Stacy <cst...@swingandcircle.com> wrote in message news:<u665rm31...@swingandcircle.com>...
>> I was told that the machine was actually emulating an IBM mainframe.
William> Not emulating, but maybe talking to...
The 5100 was not a terminal -- it was the first portable personal computer.
What I heard was (back in 1975 when I used one) was that the APL
interpreter on the machine was executing the same instructions as
one of the mainframes.
Here's more information that I have learned: The 5100 "controller" was
called "PALM" (for Put All Logic in Microcode), with 16-bit microcodes
on a 16-bit bus, 16-bit general-purpose registers, and an 8-bit ALU.
Main memory is byte-addressble, parity-checked and up to 64 KB.
There was a bunch of read-only memory containing the diagnostic and IO
supervisory routines, virtual machine interpreters each for APL and BASIC.
The APL interpreter was based on APLSV and the APL microcode emulated
a subset of the S/360 instruction set.
Internally at IBM there was an earlier proof-of-concept computer system
called SCAMP ("Special Computer APL Machine Portable") that emulated the
IBM 1130 and ran APL\1130. The original machine is in the Smithsonian.
Hewlett-Packard produced a similar-looking computer a few years later
called the HP85, which had their BASIC in ROM, but was only 16KB, and
which also included a little thermal printer. I suspect sure that this
machine was emulating the the HP2100S (HP/2000 Time Shared Basic).
CBFalconer
>
> "Pete Fenelon" <pe...@fenelon.com> wrote in message
> news:u537rk7...@corp.supernews.com...
> >
> > What was the Tandy thing with a 68k running Xenix and a Z80 doing the
> > I/O? That was out very early in the 80s....
>
> IIRC, I believe it was called the TRS-80 Model 4.
No, the model 4 was a normal CP/M machine, to replace the peculiar
model 3. The 68k/z80/xenix box was the Model II, IIRC.
--
Chuck F (cbfal...@yahoo.com) (cbfal...@XXXXworldnet.att.net)
Available for consulting/temporary embedded and systems.
(Remove "XXXX" from reply address. yahoo works unmodified)
mailto:u...@ftc.gov (for spambots to harvest)
Simon Bowring
>The 6800 used
>an assembler language similar to the PDP-11 so someone could have ported an
>early Unix version much earlier than the 68000 workstations mentioned.
Doubtful - it'd be pretty hard to port unix to an 8-bit micro like the
6800 with it's 64kB address range!
Simon
Simon Bowring
>The last Linux distro I bought came on 5 CD's.
True, but at least you can still fit a minimal linux system
onto a single bootable floppy disk!
Simon
John F. Eldredge
Hash: SHA1
On Mon, 28 Jan 2002 09:04:06 GMT, CBFalconer <cbfal...@yahoo.com>
wrote:
>Ernest Cline wrote:
>>
>> "Pete Fenelon" <pe...@fenelon.com> wrote in message
>> news:u537rk7...@corp.supernews.com...
>> >
>> > What was the Tandy thing with a 68k running Xenix and a Z80
>> > doing the I/O? That was out very early in the 80s....
>>
>> IIRC, I believe it was called the TRS-80 Model 4.
>
>No, the model 4 was a normal CP/M machine, to replace the peculiar
>model 3. The 68k/z80/xenix box was the Model II, IIRC.
That is correct. My first UNIX job was software development, in
Business Basic, on a Tandy Model II. I remember that it had an
8-inch floppy drive, a drive size which I don't think you can even
get media for now.
-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.1 Int. for non-commercial use
<http://www.pgpinternational.com>
iQA/AwUBPFVEmTMYPge5L34aEQIT5QCgjtogZ9jSBoS5HBHVBzefM+R7XtEAoOVB
B9M5F4LSuFFSwrGwvV85uypW
=AFQu
-----END PGP SIGNATURE-----
--
John F. Eldredge -- new address jo...@jfeldredge.com
eldr...@earthlink.net, eldr...@poboxes.com still work
PGP key available from http://pgpkeys.mit.edu:11371
"There must be, not a balance of power, but a community of power;
not organized rivalries, but an organized common peace."
Woodrow Wilson
daven...@sourplum.org.uk
> daven...@sourplum.org.uk wrote in message ...
>>
>>The Acorn NS16032 2nd processor was later (we - the Cambridge Uni Comp
> Soc -
>>saw one late 84/early 85 IIRC) - this would have been capable of running
>>UNIX (1MB RAM, 20MB Winchester) but was running Tripos.
> Didn't the 16032 (or was it 32016) 2nd processor and Cambridge Workstation
> (the one surviving ABC) ran PanOs? (So called I believe after the
> restaurant of the same name in Cambridge?) Was this related to Tripos? (I
> thought Tripos was the precursor to the Amiga OS?)
It may well have run PanOS (whatever that was - I'm not familiar with it) - the
one I saw (which was a BBC B with 2nd proc box - not a packaged workstation)
definitely was running Tripos. In that era one of the common CST 3rd year
projects was to port Tripos to the latest hardware that the CS dept had access
to, which I beleive is how this machine acquired it. Amiga OS had some Tripos -
IIRC the file system.
Dave
daven...@sourplum.org.uk
> Stephen Neal <stephe...@nospam.as-directed.com> wrote:
>> Didn't the 16032 (or was it 32016) 2nd processor and Cambridge Workstation
> Same chip, IIRC -- Nat Semi renamed them because 32016 sounded more
> impressive. :)
Indeed. somewhere at home I still have the 16032 data sheet...
> (Wasn't 32032 the version with a full 32-bit bus?)
Yes - I vaguely remember an 8 bit bus version aswell (32008?)
Dave
daven...@sourplum.org.uk
>> I recall torch having a Z80 based BBC 2nd processor in that time frame,
>> but that would have been CP/M based
> Was this compatible with a BBC hooked up to a z80 second CPU box?
I suspect you had either the 'official' Acorn 2nd proc, or th etorch equiv,
but not both, attached to your BBC.
Dave
Jay Maynard
>KDE is fairly sound, deep down. If I use a Linux GUI, it's been KDE for
>a long time. Gnome, with that abomination and themester's paradise
>Enlightenment, seems to delight in wasting resources on cosmetics.
Amen to that. I gave up on GNOME - which I never particularly liked anyway
because of its political correctness - when I got tired of clicking on a
window to switch focus to it, waiting 10 seconds, and still having my typing
get directed at the old window - on a 4-way PII Xeon 450 with 512 MB of RAM
and hardware SCSI RAID!
I'm typing this into an SGI Indigo2 because I really, really like 4Dwm.
Jay Maynard
>On the other hand, V7 UNIX is so small, that I can run it in a virtual
>pdp11 on my system and still get decent response.
VM/370r6 is, too: I can run it under Hercules on my 500 MHz laptop and get
great response. Even if the laptop outperforms a 370/168, that still says a
lot about VM. (Tried it yet, Lynn?)
Jay Maynard
>The 68000 was introduced in 1979. Don't remember if there was a Unix
>for it before the 68010 was introduced.
Not on a desktop, but the original NCR Tower was a 68000. (The Tower XP,
which was my first "real" Unix box, was a 68010.) Wasn't the Charles River
box a 68000, too?...and I don't remember if that was a desktop or not.
Al Kossow
> > > I/O? That was out very early in the 80s....
Model 16
Heinz W. Wiggeshoff
> On 26 Jan 2002 20:36:40 -0800, Dowe Keller <do...@krikkit.127.0.0.1> wrote:
>>On the other hand, V7 UNIX is so small, that I can run it in a virtual
>>pdp11 on my system and still get decent response.
>
> VM/370r6 is, too:
The VM shop I last worked closed in May 93. Was r6 pre or post that?
Those version numbers get jumbled as time goes by. VM/CMS had just
been given enhanced file structure beyond F-name, F-type, F-mode but
I never got to use it.
Jay Maynard
wrote:
VM/370r6 was definitely pre that...it was late 70s/early 80s. You might be
thinking of VM/SP6, which would have been still around about then. VM/SP was
the system product (read: licensed expen$ive) version of VM/370, which was
public domain (and, therefore, can be downloaded today...see
http://www.cbttape.org for pointers).
Chris Hedley
> Amen to that. I gave up on GNOME - which I never particularly liked anyway
> because of its political correctness
Political correctness? I know of various reasons why I don't like Gnome,
but that one hadn't crossed my mind!
Chris.
Anne & Lynn Wheeler
> > VM/370r6 is, too:
>
> The VM shop I last worked closed in May 93. Was r6 pre or post that?
> Those version numbers get jumbled as time goes by. VM/CMS had just
> been given enhanced file structure beyond F-name, F-type, F-mode but
> I never got to use it.
VM/370 R6 was from the late '70s. It lived on a lot longer. It was
chosen as the basis for the 3090 service processor ... first was going
to be a 4331 and then upgraded to a pair of 4361 (for
redundancy). There was a development group working on it that was
larger than the original vm/370 cp&cms development group doing the
enhancements, applications and modifications to support the service
processor functions.
field service required a boot-strappable diagnostic process done in
the field starting with a "scope". The 3090 (& 3081 before it with
uc.5 service processor) wasn't directly "field scopable" ... while the
4361 was. The idea was field service could bootstrap dianostic by
scoping any problems with the 4361 (if necessary) ... and then using
the 4361 service processor functions (that had lots of probs into all
parts of the 3090) to diagnose the 3090. The original 4331 was
upgraded to pair of 4361s as service processors to make it even less
likely diagnosing 4361 was required.
random refs:
http://www.garlic.com/~lynn/95.html#3 What is an IBM 137/148 ???
http://www.garlic.com/~lynn/96.html#41 IBM 4361 CPU technology
http://www.garlic.com/~lynn/99.html#7 IBM S/360
http://www.garlic.com/~lynn/99.html#61 Living legends
http://www.garlic.com/~lynn/99.html#62 Living legends
http://www.garlic.com/~lynn/99.html#63 System/1 ?
http://www.garlic.com/~lynn/99.html#106 IBM Mainframe Model Numbers--then and now?
http://www.garlic.com/~lynn/99.html#108 IBM 9020 computers used by FAA (was Re: EPO stories (was: HELP IT'S HOT!!!!!))
http://www.garlic.com/~lynn/99.html#110 OS/360 names and error codes (was: Humorous and/or Interesting Opcodes)
http://www.garlic.com/~lynn/99.html#112 OS/360 names and error codes (was: Humorous and/or Interesting Opcodes)
http://www.garlic.com/~lynn/99.html#239 IBM UC info
http://www.garlic.com/~lynn/2000b.html#50 VM (not VMS or Virtual Machine, the IBM sort)
http://www.garlic.com/~lynn/2000b.html#51 VM (not VMS or Virtual Machine, the IBM sort)
http://www.garlic.com/~lynn/2000b.html#66 oddly portable machines
http://www.garlic.com/~lynn/2000b.html#69 oddly portable machines
http://www.garlic.com/~lynn/2000b.html#79 "Database" term ok for plain files?
http://www.garlic.com/~lynn/2000c.html#76 Is a VAX a mainframe?
http://www.garlic.com/~lynn/2000c.html#83 Is a VAX a mainframe?
http://www.garlic.com/~lynn/2001.html#54 FBA History Question (was: RE: What's the meaning of track overfl ow?)
http://www.garlic.com/~lynn/2001b.html#75 Z/90, S/390, 370/ESA (slightly off topic)
http://www.garlic.com/~lynn/2001b.html#83 Z/90, S/390, 370/ESA (slightly off topic)
http://www.garlic.com/~lynn/2001f.html#44 Golden Era of Compilers
http://www.garlic.com/~lynn/2001h.html#2 Alpha: an invitation to communicate
http://www.garlic.com/~lynn/2001j.html#13 Parity - why even or odd (was Re: Load Locked (was: IA64 running out of steam))
http://www.garlic.com/~lynn/2001l.html#61 MVS History (all parts)
http://www.garlic.com/~lynn/2001m.html#17 3270 protocol
http://www.garlic.com/~lynn/2001n.html#9 NCP
http://www.garlic.com/~lynn/2002.html#45 VM and/or Linux under OS/390?????
http://www.garlic.com/~lynn/2002.html#48 Microcode?
--
Anne & Lynn Wheeler | ly...@garlic.com - http://www.garlic.com/~lynn/
William Donzelli
> The 5100 was not a terminal -- it was the first portable personal computer.
It ran some sort of communications package - a terminal emulator, basically.
William Donzelli
Jay Maynard
The whole *point* of GNOME is political correctness: it was started as a
reaction to KDE, which was based on Qt, which used the politically incorrect
QPL instead of the politically correct GPV. Were it not for the FSF's
zealotry, GNOME wouldn't exist.
Brian Huntley
>
> Back in the 70's, I used a nifto neato glass terminal that could be
> switched to show ASCII control characters as they were received.
> Great for debugging intricate displays in text mode.
I debugged a NAPLPS library that way once. Almost went blind.
Volker-Craig 414, I think.
Chris Hedley
> reaction to KDE, which was based on Qt, which used the politically incorrect
> QPL instead of the politically correct GPV. Were it not for the FSF's
> zealotry, GNOME wouldn't exist.
Oh, I see what you mean. I was aware of that and thought at the time
it was bloody stupid (not only fanatically overzealous, but impatiently
shortsighted) A lot of what the FSF stands for is a Good Thing; sadly
IMHO it's severely compromised by a combination of "with us or against
us" attitudes toward its draconian licencing policies and a good measure
of NIH syndrome.
It wouldn't be so bad if it wasn't for the fact that my experience of
Gnome is something which can measure up to the worst of MICROS~1 for
bugginess, bloatedness and incompatible duplication of effort.
Chris.
Pete Fenelon
> Oh, I see what you mean. I was aware of that and thought at the time
> it was bloody stupid (not only fanatically overzealous, but impatiently
> shortsighted) A lot of what the FSF stands for is a Good Thing; sadly
> IMHO it's severely compromised by a combination of "with us or against
> us" attitudes toward its draconian licencing policies and a good measure
> of NIH syndrome.
GNOME is curious. It started off as what I perceived as a hairshirt
puritan approach to putting together a desktop (technically interesting,
mind) but then got stuck with the script kiddies and their vile window
manager on one side, and Miguel's attempt to match MS
feature-for-feature and bloat-for-bloat on the other. It's... not nice.
> It wouldn't be so bad if it wasn't for the fact that my experience of
> Gnome is something which can measure up to the worst of MICROS~1 for
> bugginess, bloatedness and incompatible duplication of effort.
I'd have to agree.
pete
Pete Fenelon
> I'm pretty sure you don't actually need to install all of the SuSE
> distro containing 2500 packages. :-) In fact, you can put together a
> working system (sans X) in under 100MB. You have to work very hard
> at it though - the "Minimal" system config provided is anything but
> minimal. Emacs on a minimal system? You can then strip out about
> 20MB of modules you'll probably never use.
Yeah, but you still have to put *most of* the CDs in to get an install.
I took a gigabyte off my development box and didn't feel any of it was
essential.
> Having, in the distant past used various systems with as little as
> 256kBytes of RAM, 10MB of hard disk and 20 users, I'm well aware of
> how the magnitudes have changed. In 1985, a small Unix system
> weighed in with 50MB disk and 2MB RAM with a 10MHz 68010 for half a
> dozen users. Computer have gotten 100 times faster and bigger; the
> users seem to get less work done.
Agree *strongly* with that. They spend too much time frigging with
fripperies.
Who was it that said "We like options. We like some of them so much we
don't let you change them."?
pete
Pete Fenelon
> In regards to the message-passing part, yea I knew it was a toy
> microkernel system. I remember seeing some flamage between Tanenbaum
> and Torvalds, in regards to the laters choice of a monolithic design.
>
Got *very* amusing when the lads from Bell Labs weighed in - on
Torvalds' side :)
pete
Ben Clifford
> I suspect you had either the 'official' Acorn 2nd proc,
I suspect that too.
I remember it having Acorn/BBC labelling.
It was also the right shade of creamy yellow with black round the back.
--
Ben Clifford
http://www.hawaga.org.uk/ben/ GPG: 30F06950
webcam: http://barbarella.hawaga.org.uk/benc-cgi/watchers.cgi
Do not ever send e-mail to: phi...@hawaga.org.uk (seriously!)
Joe Pfeiffer
there, but I find it works quite nicely on my 330MHz laptop, with
128MB of memory. I really like gtk and gtkmm for developing GUIs, so
Gnome sort of goes along with that...
--
Joseph J. Pfeiffer, Jr., Ph.D. Phone -- (505) 646-1605
Department of Computer Science FAX -- (505) 646-1002
New Mexico State University http://www.cs.nmsu.edu/~pfeiffer
Southwestern NM Regional Science and Engr Fair: http://www.nmsu.edu/~scifair
Eric Smith
> Hewlett-Packard produced a similar-looking computer a few years later
> called the HP85, which had their BASIC in ROM, but was only 16KB, and
> which also included a little thermal printer. I suspect sure that this
> machine was emulating the the HP2100S (HP/2000 Time Shared Basic).
The HP 2000 Time Shared BASIC systems used one or two HP-21xx minicomputers.
The dual processor systems used one as an I/O processor. The early systems
used a 2116 main (or only) processor and a 2114 I/O processor. Later
systems moved to the 2100, and the final systems used the 21MX E-series.
The HP-85 family uses a strange 8-bit microprocessor that is nothing like a
21xx. The BASIC was written from scratch for the 85. HP Journal covered
the development.
And "a few years later" was FIVE years after the IBM 5100. HP's desktop
BASIC machine that was contemporary with the IBM 5100 was the HP-9830.
It used a 16-bit processor built out of TTL, which was architecturally
similar to the 2116. It also ran a from-scratch BASIC interpreter, also
described in the HP Journal.
HP came up with quite a number of different BASIC interpreters from the
late 1960s through the 1980s. I doubt that I've heard of them all.
AFAIK, none of the following shared any actual code (except perhaps the
300 and 3000?):
Cupertino division:
HP-21xx single user BASIC
HP 2000 Time Shared BASIC - licensed from an educational institution?
HP 250 BASIC
HP 300 BASIC
HP 3000 BASIC
Fort Collins division:
HP-9830 BASIC
HP-9835 and -9845 BASIC ("BPC" NMOS 16-bit microprocessor)
HP-9816/26/36 "Rocky Mountain" BASIC (68000), also used on
HP 9000 series 200 and 300
HP-9000 series 500 BASIC ("Focus" 32-bit NMOS processor)
Corvallis division:
HP-85 family (includes -83, -86, -87, and -9815) and HP-75 handheld
HP-71B handheld (4-64 bit Saturn processor)
HP-9817 Integral PC (68000) "Technical BASIC"
Ken S. tells me that while Rocky Mountain BASIC was original a
standalone package (bare metal, RMB *was* the OS), there eventually were
versions of both Rocky Mountain BASIC and Technical BASIC for HP/UX on
the series 300. Perhaps the latter was derived from the HP-9817
version.
Eric Smith
> HP-9835 and -9845 BASIC ("BPC" NMOS 16-bit microprocessor)
Oops, BPC was bipolar, not NMOS. That's what the "B" stood for.
Eric Smith
> Doubtful - it'd be pretty hard to port unix to an 8-bit micro like the
> 6800 with it's 64kB address range!
Really tough to port Unix from the PDP-11 with its 64kB address
range to a 6800 with its 64kB address range?
Christopher Stacy
William> Christopher Stacy <cst...@theworld.com> wrote in message news:<ulmejy...@theworld.com>...
>> The 5100 was not a terminal -- it was the first portable personal computer.
William> It ran some sort of communications package - a terminal emulator, basically.
Yes, you can use a computer system as a terminal.
However, you are incorrect in the message where you asserted
that it was not a full-blown personal computer system that
emulated an IBM 360 and ran APL and BASIC.
(Yes, it also had a communications IO port.)
K.J. Seefried
<eric-no-s...@brouhaha.com> wrote:
>
>The 3B2 family runs Unix, but is NOT a "UnixPC". They use WE 32000
>processors.
>
Definately not the UnixPC, but most of the 3B2s were definately
"desktops" per the original question. As I recall, only the 3B2/1000
really stretched the definition of desktop.
Add a BLIT/5620/730 terminal (or several) and you've got a windowed
GUI with mouse. AT&T carpet-bombed Georgia Tech with this setup when
I was there circa 1985. Tres kewl, until I later got semi-exclusive
access to a Sun 3/160.
Ken
Eric Smith
> Am I the only person around here who likes gnome?
No, there are a lot of people who like it.
I like it both as a user, and as a programmer. I can write software
with GTK and GNOME *without* using C++.
[donning asbestos underwear]
If C++ is the answer, you're asking the wrong question. OO is fine, but
do it with a REAL language. Try Eiffel, Sather, Modula 3, Ada 95, or even
Java.