[TUHS] Earliest UNIX Workstations?

[TUHS] Earliest UNIX Workstations?

[Paul Ruizendaal]

As a result of the recent discussion on this list I’m trying to understand the timeline of graphical computing on Unix, first of all in my preferred time slot ’75 -’85.

When it comes to Bell Labs I’m aware of the following:

- around 1975 the Labs worked on the Glance-G vector graphics terminal. This was TSS-516 based with no Unix overlap I think.
- around the same time the Labs seem to have used the 1973 Dec VT11 vector graphics terminal; at least the surviving LSX Unix source has a driver for it
- in 1976 there was the Terak 8510; this ran primarily USCD pascal, but it also ran LSX and/or MX (but maybe only much later)
- then it seems to jump 1981 and to the Blit.
- in 1984 there was MGR that was done at Bellcore

Outside of the labs (but on Unix), I have:

- I am not sure what graphics software ran on the SUN-1, but it must have been something
- Clem just mentioned the 1981 Tektronix Magnolia system
- Wikipedia says that X1 was 1984 and X11 was 1987; I’m not sure when it became Unix centered
- Sun’s NeWS arrived only in 1989, I think?

Outside of Unix, in the microcomputer world there was a lot of cheap(er) graphics hardware. Lot’s of stuff at 256 x 192 resolution, but up to 512 x 512 at the higher end. John Walker writes that the breakout product for Autodesk was Interact (the precursor to AutoCAD). Initially developed for S-100 bus systems it quickly moved to the PC. There was a lot of demand for CAD at a 5K price point that did not exist at a 50K price point.

[TUHS] Re: Earliest UNIX Workstations?

[Larry McVoy]

On Thu, Jan 26, 2023 at 04:58:11PM +0100, Paul Ruizendaal wrote:
> - I am not sure what graphics software ran on the SUN-1, but it must have been something

Probably SunView, that's the first one I saw.  It was

	SunView
	OpenLook (SunView ported to X11)
	CDE
	NeWS

I used either straight X11 or OpenLook but replacing the window manager
with twm and then ctwm.  Did the same thing at SGI, I hated all the
fancy desktops each vendor did.  In the days of 4MB or 8MB or even
32MB desktops, all that "fancy" used memory that I needed for actual
real work, like building the kernel.  Straight X11 with a simple window
manager gave me a lot of memory back.

I viewed that fancy stuff as stuff for people who didn't do real work.
A bit of a snob, I was :-)

[TUHS] Re: Earliest UNIX Workstations?

[Clem Cole]

[-- Attachment #1: Type: text/plain, Size: 3554 bytes --]

On Thu, Jan 26, 2023 at 10:58 AM Paul Ruizendaal <pnr@planet.nl> wrote:

>
> As a result of the recent discussion on this list I’m trying to understand
> the timeline of graphical computing on Unix, first of all in my preferred
> time slot ’75 -’85.
>
> When it comes to Bell Labs I’m aware of the following:
>
> - around 1975 the Labs worked on the Glance-G vector graphics terminal.
> This was TSS-516 based with no Unix overlap I think.
> - around the same time the Labs seem to have used the 1973 Dec VT11 vector
> graphics terminal; at least the surviving LSX Unix source has a driver for
> it
> - in 1976 there was the Terak 8510; this ran primarily USCD pascal, but it
> also ran LSX and/or MX (but maybe only much later)
>
 In the famous picture of Ken and Dennis you see a Tek display connected to
the 11/20.
Simply during that time there were a number of graphics systems from the
DVST (storage tubes) like Tek 4014 to Raster Systems like the GDPs we had
at CMU. There really are too many to list.



> - then it seems to jump 1981 and to the Blit.
> - in 1984 there was MGR that was done at Bellcore
>
> Outside of the labs (but on Unix), I have:
>
> - I am not sure what graphics software ran on the SUN-1, but it must have
> been something
>
Again - W was the windowing system for the Sun board, running on the V
kernel.  It was original envisioned as a very smart terminal to bigger
systems.  Remember it did not have an MMU to start with.  Andy added and
MMU and then eventually changed it to a 68010.  VLSI Tech was born and
eventual became Sun Micro Systems but that was a few years later.  I have
to believe W as moved to UNIX on the SUN Terminal and that would have been
what Chris Kent and folks started with for the microVax - but I do not know
for sure.




> - Clem just mentioned the 1981 Tektronix Magnolia system
>
1979/1980 actually -- Roger and I started that in summer of '79 and he
wrote that a year later when we go Tek money.  It was originall as 'g-job'
we were building for ourselves.  Our boss saw what were were doing and
Roger got $10K to do a proposal -- that document was the result.

I already had the basics of a compiler working by them (well sort of) and
the beginning of a Unix port on the test board. Jon Steinhart may be
remember some of this as they all visited us in the labs to see what we
were doing.

- Wikipedia says that X1 was 1984 and X11 was 1987; I’m not sure when it
> became Unix centered
> - Sun’s NeWS arrived only in 1989, I think?
>
> Outside of Unix, in the microcomputer world there was a lot of cheap(er)
> graphics hardware. Lot’s of stuff at 256 x 192 resolution, but up to 512 x
> 512 at the higher end. John Walker writes that the breakout product for
> Autodesk was Interact (the precursor to AutoCAD). Initially developed for
> S-100 bus systems it quickly moved to the PC. There was a lot of demand for
> CAD at a 5K price point that did not exist at a 50K price point.
>
Not completely true...  1-4K for BW was possible (expensive) but
available.  I  tend to believe that systems like E&S could do that. Many
raster systems went to 1K -- again is was about cost. I've forgotten the
resolution of the GDP2 but is was much higher -- it used a rather expensive
HP display.  The price of memory and price of the monitor tneded to
dominate. Also the processor was not cheap -- a GDP2 had a dedicated
PDP-11/20, but that was also try of things like GT40 and the
similar systems of the time.
ᐧ

[-- Attachment #2: Type: text/html, Size: 5416 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[emanuel stiebler]

On 2023-01-26 11:04, Larry McVoy wrote:
> [..] Straight X11 with a simple window
> manager gave me a lot of memory back.
> 
> I viewed that fancy stuff as stuff for people who didn't do real work.
> A bit of a snob, I was :-)

Xclock & many xterms, that what was on my desktop. Just replacing all 
the terminals, and to have one xterm with more lines for editing.

[TUHS] Re: Earliest UNIX Workstations?

[segaloco via TUHS]

Twm gets the job done for me too.  My only gripe is it doesn't support WM full screen hints that a lot of other WMs do so certain things don't full-screen properly, but others do.  In either case, not a huge problem, something if it mattered enough to me I'd just try to patch in.

- Matt G.

------- Original Message -------
On Thursday, January 26th, 2023 at 8:37 AM, emanuel stiebler <emu@e-bbes.com> wrote:


> On 2023-01-26 11:04, Larry McVoy wrote:
> 
> > [..] Straight X11 with a simple window
> > manager gave me a lot of memory back.
> > 
> > I viewed that fancy stuff as stuff for people who didn't do real work.
> > A bit of a snob, I was :-)
> 
> 
> Xclock & many xterms, that what was on my desktop. Just replacing all
> the terminals, and to have one xterm with more lines for editing.

[TUHS] Re: Earliest UNIX Workstations?

[Warner Losh]

[-- Attachment #1: Type: text/plain, Size: 3024 bytes --]

On Thu, Jan 26, 2023 at 8:58 AM Paul Ruizendaal <pnr@planet.nl> wrote:

>
> As a result of the recent discussion on this list I’m trying to understand
> the timeline of graphical computing on Unix, first of all in my preferred
> time slot ’75 -’85.
>
> When it comes to Bell Labs I’m aware of the following:
>
> - around 1975 the Labs worked on the Glance-G vector graphics terminal.
> This was TSS-516 based with no Unix overlap I think.
> - around the same time the Labs seem to have used the 1973 Dec VT11 vector
> graphics terminal; at least the surviving LSX Unix source has a driver for
> it
> - in 1976 there was the Terak 8510; this ran primarily USCD pascal, but it
> also ran LSX and/or MX (but maybe only much later)
> - then it seems to jump 1981 and to the Blit.
> - in 1984 there was MGR that was done at Bellcore
>
> Outside of the labs (but on Unix), I have:
>
> - I am not sure what graphics software ran on the SUN-1, but it must have
> been something
>

If this is the sun microsystem sun-1, the leaked sources online suggest
that these initially ran a V7 port by Unisoft. This switched to a 4.2BSD
port maybe before it went to customers as SunOS 1.0 if other leaked sources
can be believed.

If this is the Standford Unix Networked (?) sun, then I don't know.


> - Clem just mentioned the 1981 Tektronix Magnolia system
> - Wikipedia says that X1 was 1984 and X11 was 1987; I’m not sure when it
> became Unix centered
>

I believe very early. It ran first on the VS100, and I believe that those
machines at project athena were running Unix, but I'm not sure of the cut
over from Stanford V. Another thread posted the X announcement which was in
June 1984. There was also a pointer to a blog about pictures of the W
window system. None exist, it seems. The windowing system pictured in the
glossy marketing sheets for the VS100 were for VMS and VMS Windowing
System. I'd put money against the first X running on VMS. :)

We had X10R3 running on Sun 3/50s in our lab, though more often they ran
SunView since it was faster and more familiar to the admins that ran the
machines. This was in 86 or 87 I believe.


> - Sun’s NeWS arrived only in 1989, I think?
>

No. It had to be late 1987 or 1988 because I ran that on the Sun 3/60 that
the Hydrology department that I worked for ran. I didn't run it often, mind
you, and the 'generic terminal emulator for any termcap entry' terminal was
cool, but it was just a notch too weird for a daily driver.

Warner


> Outside of Unix, in the microcomputer world there was a lot of cheap(er)
> graphics hardware. Lot’s of stuff at 256 x 192 resolution, but up to 512 x
> 512 at the higher end. John Walker writes that the breakout product for
> Autodesk was Interact (the precursor to AutoCAD). Initially developed for
> S-100 bus systems it quickly moved to the PC. There was a lot of demand for
> CAD at a 5K price point that did not exist at a 50K price point.
>
>
>
>
>
>

[-- Attachment #2: Type: text/html, Size: 3898 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Jon Steinhart]

Paul Ruizendaal writes:
> As a result of the recent discussion on this list I’m trying to understand the timeline of graphical computing on Unix, first of all in my preferred time slot ’75 -’85.
> 
> When it comes to Bell Labs I’m aware of the following:
> 
> - around 1975 the Labs worked on the Glance-G vector graphics terminal. This was TSS-516 based with no Unix overlap I think.
> - around the same time the Labs seem to have used the 1973 Dec VT11 vector graphics terminal; at least the surviving LSX Unix source has a driver for it
> - in 1976 there was the Terak 8510; this ran primarily USCD pascal, but it also ran LSX and/or MX (but maybe only much later)
> - then it seems to jump 1981 and to the Blit.
> - in 1984 there was MGR that was done at Bellcore
> 
> Outside of the labs (but on Unix), I have:
> 
> - I am not sure what graphics software ran on the SUN-1, but it must have been something
> - Clem just mentioned the 1981 Tektronix Magnolia system
> - Wikipedia says that X1 was 1984 and X11 was 1987; I’m not sure when it became Unix centered
> - Sun’s NeWS arrived only in 1989, I think?
> 
> Outside of Unix, in the microcomputer world there was a lot of cheap(er) graphics hardware. Lot’s of stuff at 256 x 192 resolution, but up to 512 x 512 at the higher end. John Walker writes that the breakout product for Autodesk was Interact (the precursor to AutoCAD). Initially developed for S-100 bus systems it quickly moved to the PC. There was a lot of demand for CAD at a 5K price point that did not exist at a 50K price point.

The timeline for the GLANCE-G is off by a few years.  It might
have been as early as 1969 but I don't remember when the transition
from GLANCE-C (characters only) to GLANCE-G (graphics) occurred.
I'm absolutely sure that the G existed in 1972 since I worked on it then.
The only UNIX overlap was when the Ring was adapted for the PDP-11 so
that Ken could have a GLANCE-G for chess.

I have a vague memory of other graphics work at the labs; I remember
being in someone's lab that had modified big Crown audio power amps
for current feedback to drive deflection coils, but no other details.

[TUHS] Re: Earliest UNIX Workstations?

[Lars Brinkhoff]

Warner Losh wrote:
> If this is the Standford Unix Networked (?) sun, then I don't know.

Stanford University Network workstation.  I have seen some documents
about it on stacks.stanford.edu, but I don't remember about the
software.  In some versions, it's a more of a multi-head remote graphics
terminal, so maybe not Unix.

>  - Wikipedia says that X1 was 1984 and X11 was 1987; I’m not sure when it
>  became Unix centered
>
> I believe very early. It ran first on the VS100

Note that the VAXstation 100 is not a VAX, and not a standalone
computer.  It's a 68000-based graphics terminal that attaches to a VAX.
The VS100 has some firmware in ROM, and the host uploads additional
software.  There is such a software blob in X10R3.

> There was also a pointer to a blog about pictures of the W window
> system. None exist, it seems.

I have asked Asante, Reid, and Kent.  No luck so far.

[TUHS] Re: Earliest UNIX Workstations?

[Bakul Shah]

The V sources on bitsavers has a w command. But don't get excited!
It seems to be similar to unix's w, a variation on the who command.

I wonder if this mythical w is the same as V's VGTS as it seems to have
pretty much the same model. From
 https://dl.acm.org/doi/pdf/10.1145/357332.357334

 The ideal interface must take into account four fundamental principles:

 (1) The interface to application programs should be independent of
     particular physical devices or intervening networks.

 (2) The user should be allowed to perform multiple tasks
     simultaneously.

 (3) The command interaction discipline should be consistent and
     natural.

 (4) Response to user interaction should be fast.

 The first principle has led to work in virtual terminals (VTs) and
 deviceindependent graphics packages, the second to work in window
 systems, and the third to work in what has recently been called user
 interface management systems, the most common examples of which are
 command languages. Without adhering to the fourth principle, however,
 much of the other work is moot. In a distributed environment, in
 particular, the supporting network protocols cannot incur inordinate
 overhead.

From concluding remarks:
 To summarize the major attributes of the VGTS:
 - Instead of describing how to draw a picture, the application
   describes what is to be drawn. The user then specifies where the
   picture should be displayed.
 - Objects have a hierarchical structure. Hence, the VGTS supports
   structured display files rather than segmented display files.
 - The VGTS is portable to a range of relatively high-performance
   devices.
 - Applications can be distributed over multiple machines.
 - A single user can access several different applications
   simultaneously.
 - It performs well!

VGTS code is in the V system sources at bitsavers.

> On Jan 26, 2023, at 10:15 AM, Lars Brinkhoff <lars@nocrew.org> wrote:
> 
> Warner Losh wrote:
>> If this is the Standford Unix Networked (?) sun, then I don't know.
> 
> Stanford University Network workstation.  I have seen some documents
> about it on stacks.stanford.edu, but I don't remember about the
> software.  In some versions, it's a more of a multi-head remote graphics
> terminal, so maybe not Unix.
> 
>> - Wikipedia says that X1 was 1984 and X11 was 1987; I’m not sure when it
>> became Unix centered
>> 
>> I believe very early. It ran first on the VS100
> 
> Note that the VAXstation 100 is not a VAX, and not a standalone
> computer.  It's a 68000-based graphics terminal that attaches to a VAX.
> The VS100 has some firmware in ROM, and the host uploads additional
> software.  There is such a software blob in X10R3.
> 
>> There was also a pointer to a blog about pictures of the W window
>> system. None exist, it seems.
> 
> I have asked Asante, Reid, and Kent.  No luck so far.

[TUHS] Re: Earliest UNIX Workstations?

[Rob Pike]

[-- Attachment #1: Type: text/plain, Size: 1942 bytes --]

There was also all the work with the Three Rivers Graphic Wonder on the
PDP-11/45 at the University of Toronto Dynamic Graphics Project from 1974
onwards, as well as various film plotters, Versatec, Tennenhouse's frame
buffer, and so on.

-rob


On Fri, Jan 27, 2023 at 2:58 AM Paul Ruizendaal <pnr@planet.nl> wrote:

>
> As a result of the recent discussion on this list I’m trying to understand
> the timeline of graphical computing on Unix, first of all in my preferred
> time slot ’75 -’85.
>
> When it comes to Bell Labs I’m aware of the following:
>
> - around 1975 the Labs worked on the Glance-G vector graphics terminal.
> This was TSS-516 based with no Unix overlap I think.
> - around the same time the Labs seem to have used the 1973 Dec VT11 vector
> graphics terminal; at least the surviving LSX Unix source has a driver for
> it
> - in 1976 there was the Terak 8510; this ran primarily USCD pascal, but it
> also ran LSX and/or MX (but maybe only much later)
> - then it seems to jump 1981 and to the Blit.
> - in 1984 there was MGR that was done at Bellcore
>
> Outside of the labs (but on Unix), I have:
>
> - I am not sure what graphics software ran on the SUN-1, but it must have
> been something
> - Clem just mentioned the 1981 Tektronix Magnolia system
> - Wikipedia says that X1 was 1984 and X11 was 1987; I’m not sure when it
> became Unix centered
> - Sun’s NeWS arrived only in 1989, I think?
>
> Outside of Unix, in the microcomputer world there was a lot of cheap(er)
> graphics hardware. Lot’s of stuff at 256 x 192 resolution, but up to 512 x
> 512 at the higher end. John Walker writes that the breakout product for
> Autodesk was Interact (the precursor to AutoCAD). Initially developed for
> S-100 bus systems it quickly moved to the PC. There was a lot of demand for
> CAD at a 5K price point that did not exist at a 50K price point.
>
>
>
>
>
>

[-- Attachment #2: Type: text/html, Size: 2513 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Paul Ruizendaal]

> On 26 Jan 2023, at 17:29, Clem Cole <clemc@ccc.com> wrote:
> 
>> Outside of Unix, in the microcomputer world there was a lot of cheap(er) graphics hardware. Lot’s of stuff at 256 x 192 resolution, but up to 512 x 512 at the higher end. John Walker writes that the breakout product for Autodesk was Interact (the precursor to AutoCAD). Initially developed for S-100 bus systems it quickly moved to the PC. There was a lot of demand for CAD at a 5K price point that did not exist at a 50K price point.
> Not completely true...  1-4K for BW was possible (expensive) but available.  I  tend to believe that systems like E&S could do that. Many raster systems went to 1K -- again is was about cost. I've forgotten the resolution of the GDP2 but is was much higher -- it used a rather expensive HP display.  The price of memory and price of the monitor tneded to dominate. Also the processor was not cheap -- a GDP2 had a dedicated PDP-11/20, but that was also try of things like GT40 and the similar systems of the time.

I meant early micro/home computers. I think John Walker was comparing the typical late 70’s CAD (drawing) system, i.e a mini computer, a few graphics terminals and CAD software versus Interact running on a S-100 system with a high end graphics card, a digitiser board and a terminal. See for instance here:
https://www.3dcadworld.com/autocads-ancestor/

> - I am not sure what graphics software ran on the SUN-1, but it must have been something
> Again - W was the windowing system for the Sun board, running on the V kernel.  It was original envisioned as a very smart terminal to bigger systems.  Remember it did not have an MMU to start with.  Andy added and MMU and then eventually changed it to a 68010.  VLSI Tech was born and eventual became Sun Micro Systems but that was a few years later.  I have to believe W as moved to UNIX on the SUN Terminal and that would have been what Chris Kent and folks started with for the microVax - but I do not know for sure.



> Date: Thu, 26 Jan 2023 09:51:59 -0700
> From: Warner Losh <imp@bsdimp.com>

> 
>> - I am not sure what graphics software ran on the SUN-1, but it must have been something
> 
> If this is the sun microsystem sun-1, the leaked sources online suggest that these initially ran a V7 port by Unisoft. This switched to a 4.2BSD port maybe before it went to customers as SunOS 1.0 if other leaked sources can be believed.

I never really distinguished between the Stanford "SUN" and the Sun Microsystems "Sun-1”, oops. Taking Clem’s comment into account I could see that the SUN ran the V kernel and the W graphics system, and that the Sun-1 was using an early form of X.


> Bakul Shah bakul at iitbombay.org 
> Fri Jan 27 05:39:30 AEST 2023
> 
> I wonder if this mythical w is the same as V's VGTS as it seems to have
> pretty much the same model.
> From https://dl.acm.org/doi/pdf/10.1145/357332.357334


That is an interesting thought and paper.

The paper says:

"As noted in Section 3, the VGTS is only one component of the user interface software. The other components are the view manager, the exec server, and the executives. Together, the view manager and exec server constitute the user interface management.”

In this context an “executive” seems to be similar to a shell, the exec server appears to manage client connections and the view manager seems to be similar to a window manager in X. Maybe this whole was (later?) referred to as “W”?

This was the post announcing X:

From: rws@mit-bold (Robert W. Scheifler)
To: window@athena
Subject: window system X
Date: 19 Jun 1984 0907-EDT (Tuesday)

I've spent the last couple weeks writing a window
system for the VS100. I stole a fair amount of code
from W, surrounded it with an asynchronous rather
than a synchronous interface, and called it X. Overall
performance appears to be about twice that of W. The
code seems fairly solid at this point, although there are
still some deficiencies to be fixed up. 

We at LCS have stopped using W, and are now
actively building applications on X. Anyone else using
W should seriously consider switching. This is not the
ultimate window system, but I believe it is a good
starting point for experimentation. Right at the moment
there is a CLU (and an Argus) interface to X; a C
interface is in the works. The three existing
applications are a text editor (TED), an Argus I/O
interface, and a primitive window manager. There is
no documentation yet; anyone crazy enough to
volunteer? I may get around to it eventually. 

Anyone interested in seeing a demo can drop by
NE43-531, although you may want to call 3-1945
first. Anyone who wants the code can come by with a
tape. Anyone interested in hacking deficiencies, feel
free to get in touch.

[TUHS] Re: Earliest UNIX Workstations?

[Bakul Shah]

On Jan 26, 2023, at 2:17 PM, Paul Ruizendaal <pnr@planet.nl> wrote:
> 
>> 
>> Bakul Shah bakul at iitbombay.org 
>> Fri Jan 27 05:39:30 AEST 2023
>> 
>> I wonder if this mythical w is the same as V's VGTS as it seems to have
>> pretty much the same model.
>> From https://dl.acm.org/doi/pdf/10.1145/357332.357334
> 
> 
> That is an interesting thought and paper.
> 
> The paper says:
> 
> "As noted in Section 3, the VGTS is only one component of the user interface software. The other components are the view manager, the exec server, and the executives. Together, the view manager and exec server constitute the user interface management.”
> 
> In this context an “executive” seems to be similar to a shell, the exec server appears to manage client connections and the view manager seems to be similar to a window manager in X. Maybe this whole was (later?) referred to as “W”?

For the curious the sources are available at http://www.bitsavers.org/bits/Stanford/VSystem_1983.zip

It just seems weird that
a) for a research OS such as V there were more than one such effort but
b) there are no papers about W or any mention of it in V related papers.
c) there is source code for VGTS as well as at least one paper about it.
d) VGTS seems to have had similar goals as X windows.
e) there seemed to be multiple applications written using VGTS.

I didn't spend much time searching so may have missed things. I have
asked a friend who had Cheriton as research advisor in '80s but most
likely it won't yield anything.

[TUHS] Re: Earliest UNIX Workstations?

[Paul Ruizendaal]

> On 26 Jan 2023, at 23:45, Bakul Shah <bakul@iitbombay.org> wrote:
> 
> On Jan 26, 2023, at 2:17 PM, Paul Ruizendaal <pnr@planet.nl> wrote:
>> 
>>> 
>>> Bakul Shah bakul at iitbombay.org 
>>> Fri Jan 27 05:39:30 AEST 2023
>>> 
>>> I wonder if this mythical w is the same as V's VGTS as it seems to have
>>> pretty much the same model.
>>> From https://dl.acm.org/doi/pdf/10.1145/357332.357334

And the answer is “no”:

From https://apps.hci.rwth-aachen.de/borchers-old/cs377a/materials/p79-scheifler.pdf :

"The name X derives from the lineage of the system. At Stanford University, Paul Asente and Brian Reid had begun work on the W window system [3] as an alternative to VGTS [13, 22] for the V system [5]. Both VGTS and W allow network-transparent access to the display, using the synchronous V communication mechanism. Both systems provide “text” windows for ASCII terminal emulation. VGTS provides graphics windows driven by fairly high-level object definitions from a structured display file; W provides graphics windows based on a simple display-list mechanism, with limited functionality. We acquired a UNIX- based version of W for the VSlOO (with synchronous communication over TCP [24] produced by Asente and Chris Kent at Digital’s Western Research Laboratory. From just a few days of experimentation, it was clear that a network- transparent hierarchical window system was desirable, but that restricting the system to any fixed set of application-specific modes was completely inadequate. It was also clear that, although synchronous communication was perhaps acceptable in the V system (owing to very fast networking primitives), it was completely inadequate in most other operating environments. X is our “reaction” to W.”

The reference [3] is "ASENTE, P. W reference manual. Internal document, Dept. Computer Science, Stanford Univ., Calif., 1984.”

The version of X discussed in the paper was apparently part of the 4.3BSD distribution tapes:

"The use of X has grown far beyond anything we had imagined. Digital has incorporated X into a commercial product, and other manufacturers are following suit. With the appearance of such products and the release of complete X sources on the Berkeley 4.3 UNIX distribution tapes, it is no longer feasible to track all X use and development.”

And I was wrong with:

> I never really distinguished between the Stanford "SUN" and the Sun Microsystems "Sun-1”, oops. Taking Clem’s comment into account I could see that the SUN ran the V kernel and the W graphics system, and that the Sun-1 was using an early form of X.

Whilst the part about SUN may be correct, the Sun-1 was apparently using MGR and SunWindows/SunView -- at least according to the interesting blog post here (discussed on TUHS a few months ago):

https://oldvcr.blogspot.com/2022/10/if-one-guis-not-enough-for-your-sparc.html

[TUHS] Re: Earliest UNIX Workstations?

[Lars Brinkhoff]

> Stanford University Network workstation.  I have seen some documents
> about it on stacks.stanford.edu, but I don't remember about the
> software.  In some versions, it's a more of a multi-head remote graphics
> terminal, so maybe not Unix.

In the 1980 document, it's described as a single 68000 serving up to 16
seats.  "only the basic Pup Telnet protocols need be implemented on the
workstation's MC68OOO processor.  The SUN terminals could then be
programmed to emulate currently available terminals, such as Datamedias,
Telerays, Tektronix 4014 graphics terminals, or III displays."
https://stacks.stanford.edu/file/druid:gg867qx3134/gg867qx3134.pdf

But by 1982 it looks much like a classic workstation.  "The SUN Memory
Management Unit has been designed to support a multitasking operating
system such as Bell Lab’s UNIX."
http://i.stanford.edu/pub/cstr/reports/csl/tr/82/229/CSL-TR-82-229.pdf

[TUHS] Re: Earliest UNIX Workstations?

[Paul Ruizendaal via TUHS]

[-- Attachment #1: Type: text/plain, Size: 1052 bytes --]


> On Jan 27, 2023, at 1:19 AM, Paul Ruizendaal <pnr@planet.nl> wrote:
> 
> The version of X discussed in the paper was apparently part of the 4.3BSD distribution tapes:
> 
> "The use of X has grown far beyond anything we had imagined. Digital has incorporated X into a commercial product, and other manufacturers are following suit. With the appearance of such products and the release of complete X sources on the Berkeley 4.3 UNIX distribution tapes, it is no longer feasible to track all X use and development.”
> 

This X is not on the TUHS Unix tree website, nor on the CSRG disks. It turns out that there is a directory “src/new” that is not included there. It is here:

http://www.retro11.de/ouxr/43bsd/usr/src/new/ <http://www.retro11.de/ouxr/43bsd/usr/src/new/>

The version of X included with 4.3BSD was X10. I assume this is the oldest surviving X Window source code.

Of course the source code for the Blit has survived, as has the source code of MGR. The source code for Sunwindows and NeWS is presumably lost?




[-- Attachment #2: Type: text/html, Size: 1663 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Warner Losh]

[-- Attachment #1: Type: text/plain, Size: 2265 bytes --]

On Fri, Jan 27, 2023 at 10:16 AM Paul Ruizendaal via TUHS <tuhs@tuhs.org>
wrote:

>
> On Jan 27, 2023, at 1:19 AM, Paul Ruizendaal <pnr@planet.nl> wrote:
>
> The version of X discussed in the paper was apparently part of the 4.3BSD
> distribution tapes:
>
> "The use of X has grown far beyond anything we had imagined. Digital has
> incorporated X into a commercial product, and other manufacturers are
> following suit. With the appearance of such products and the release of
> complete X sources on the Berkeley 4.3 UNIX distribution tapes, it is no
> longer feasible to track all X use and development.”
>
>
> This X is not on the TUHS Unix tree website, nor on the CSRG disks. It
> turns out that there is a directory “src/new” that is not included there.
> It is here:
>
> http://www.retro11.de/ouxr/43bsd/usr/src/new/
>
> The version of X included with 4.3BSD was X10. I assume this is the oldest
> surviving X Window source code.
>

There's X10R3 and X10R4 at https://www.x.org/archive/X10R3/ and
https://www.x.org/archive/X10R4/. On the FTP site, there's sym links for
R1, R2, R3, R4, R5, R6, R6.1, R6.3, R6.4, R6.5.1, R6.6 and R6.8 in the pub
directory as well, but they are dead links and correspond to the X11
releases that are also there, not X1, etc.

The X10R3 is from Feb 2, 1986. X10R4 is from December 2, 1986. The
retro11.de files are from June 1986, so
are no later than X10R4, and most likely either X10R3 or an internal
snapshot (I've not downloaded them both
to run a diff to see which).

Google searches for X9, X8, etc aren't at all helpful.

Of course the source code for the Blit has survived, as has the source code
> of MGR. The source code for Sunwindows and NeWS is presumably lost?
>

When I was a Solbroune, we started the OI toolkit with pdb, swm, uib, etc
because Sun refused to license the source code to SunView. Although I had
easy access to SunOS (which I wish I'd saved a copy of now), the SunView
code was never in the building. It was relatively easy to get SunOS sources
for a fee, but much harder for SunView. So I'm less than completely hopeful
here. And NeWS was a fringe thing with a significantly shorter product
life, so I'm even less hopeful there.

Warner

[-- Attachment #2: Type: text/html, Size: 3314 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Warner Losh]

[-- Attachment #1: Type: text/plain, Size: 2510 bytes --]

On Fri, Jan 27, 2023 at 10:36 AM Warner Losh <imp@bsdimp.com> wrote:

>
>
> On Fri, Jan 27, 2023 at 10:16 AM Paul Ruizendaal via TUHS <tuhs@tuhs.org>
> wrote:
>
>>
>> On Jan 27, 2023, at 1:19 AM, Paul Ruizendaal <pnr@planet.nl> wrote:
>>
>> The version of X discussed in the paper was apparently part of the 4.3BSD
>> distribution tapes:
>>
>> "The use of X has grown far beyond anything we had imagined. Digital has
>> incorporated X into a commercial product, and other manufacturers are
>> following suit. With the appearance of such products and the release of
>> complete X sources on the Berkeley 4.3 UNIX distribution tapes, it is no
>> longer feasible to track all X use and development.”
>>
>>
>> This X is not on the TUHS Unix tree website, nor on the CSRG disks. It
>> turns out that there is a directory “src/new” that is not included there.
>> It is here:
>>
>> http://www.retro11.de/ouxr/43bsd/usr/src/new/
>>
>> The version of X included with 4.3BSD was X10. I assume this is the
>> oldest surviving X Window source code.
>>
>
> There's X10R3 and X10R4 at https://www.x.org/archive/X10R3/ and
> https://www.x.org/archive/X10R4/. On the FTP site, there's sym links for
> R1, R2, R3, R4, R5, R6, R6.1, R6.3, R6.4, R6.5.1, R6.6 and R6.8 in the pub
> directory as well, but they are dead links and correspond to the X11
> releases that are also there, not X1, etc.
>
> The X10R3 is from Feb 2, 1986. X10R4 is from December 2, 1986. The
> retro11.de files are from June 1986, so
> are no later than X10R4, and most likely either X10R3 or an internal
> snapshot (I've not downloaded them both
> to run a diff to see which).
>
> Google searches for X9, X8, etc aren't at all helpful.
>

Also interesting to note is that X10 had clu bindings in the
CLUlib directory...


> Of course the source code for the Blit has survived, as has the source
>> code of MGR. The source code for Sunwindows and NeWS is presumably lost?
>>
>
> When I was a Solbroune, we started the OI toolkit with pdb, swm, uib, etc
> because Sun refused to license the source code to SunView. Although I had
> easy access to SunOS (which I wish I'd saved a copy of now), the SunView
> code was never in the building. It was relatively easy to get SunOS sources
> for a fee, but much harder for SunView. So I'm less than completely hopeful
> here. And NeWS was a fringe thing with a significantly shorter product
> life, so I'm even less hopeful there.
>
> Warner
>

[-- Attachment #2: Type: text/html, Size: 3970 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[josh]

[-- Attachment #1: Type: text/plain, Size: 381 bytes --]

On Friday, January 27, 2023, Paul Ruizendaal via TUHS <tuhs@tuhs.org> wrote:

>
> Of course the source code for the Blit has survived, as has the source
> code of MGR. The source code for Sunwindows and NeWS is presumably lost?
>

Arnold Robbins linked to this repository in a previous thread, it seems to
have some NeWS-related source:

https://github.com/IanDarwin/OpenLookCDROM

[-- Attachment #2: Type: text/html, Size: 725 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Rich Salz]

[-- Attachment #1: Type: text/plain, Size: 323 bytes --]

> Also interesting to note is that X10 had clu bindings in the
CLUlib directory...

I'd like to know what the first versions of X were written in; the notice
talks about CLU bindings exist and C coming soon. The app-side seems
reasonable, but what was the display engine written in?  (I took the CLU
class around 1980.)

>

[-- Attachment #2: Type: text/html, Size: 634 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Warner Losh]

[-- Attachment #1: Type: text/plain, Size: 575 bytes --]

On Fri, Jan 27, 2023 at 10:45 AM Rich Salz <rich.salz@gmail.com> wrote:

>
> > Also interesting to note is that X10 had clu bindings in the
> CLUlib directory...
>
> I'd like to know what the first versions of X were written in; the notice
> talks about CLU bindings exist and C coming soon. The app-side seems
> reasonable, but what was the display engine written in?  (I took the CLU
> class around 1980.)
>

I wrote  ~500-1000 lines of CLU in a computer language survey course in
1986...

Without the earlier versions' source, it's hard to answer this question...

Warner

[-- Attachment #2: Type: text/html, Size: 1121 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Lars Brinkhoff]

Warner Losh wrote:
> Rich Salz wrote:
>> I'd like to know what the first versions of X were written in
>
> Without the earlier versions' source, it's hard to answer this question...

V source code exists, right?  It seems likely W would have been written
in the same language as W.  And that early X would also be the same.
Another source of information would be to ask Bob Scheifler and Jim
Gettys.

[TUHS] Re: Earliest UNIX Workstations?

[Paul Ruizendaal]

> On 28 Jan 2023, at 10:14, Lars Brinkhoff <lars@nocrew.org> wrote:
> 
> Warner Losh wrote:
>> Rich Salz wrote:
>>> I'd like to know what the first versions of X were written in
>> 
>> Without the earlier versions' source, it's hard to answer this question...
> 
> V source code exists, right?  It seems likely W would have been written
> in the same language as W.  And that early X would also be the same.
> Another source of information would be to ask Bob Scheifler and Jim
> Gettys.

Whilst that is a reasonable assumption, I’m not sure it is true in this case. Bob Scheifler writes in 1986:

"We acquired a UNIX-based version of W for the VSlOO (with synchronous communication over TCP [24] produced by Asente and Chris Kent at Digital’s Western Research Laboratory.”

It does not say “C based”, but it is quite possible that the Unix port also meant moving to C.

Also, the work started in June 1984 and had gone to version 10, release 3 by February 1986. That is 12 versions in 20 months. Most likely X1-X10R2 are all snapshots done in rapid succession. The change notes for X10R3 read as describing a work still in progress:

http://www.retro11.de/ouxr/43bsd/usr/src/new/X/CHANGES

That “work-in-progress” feel also shows in the Xterm README:

"Xterm is in a reasonably usable state.  We are sick and tired of working
on it, but there are clearly major areas of improvement possible.  Do
not look to us to do more than integration work on other people's
improvement.  About 50% of it is the oldest existing code in the package
and needing major rewrite.  Our thanks to Bob McNamara for the 50% which
is solid."

The README for the X server itself (written in August 1985 it seems, http://www.retro11.de/ouxr/43bsd/usr/src/new/X/X/README) says:

"The server has been completely rewritten several times now, and I am reasonably
happy with it.  I have fine-tuned it specifically for the current (sub-optimal)
VAX compiler.  For other machines, faster code may be obtained in some cases
by changing sizes (e.g., to avoid indexing shifts on the 68000) or register
declarations.  Attempts to parameterize along these lines have only been made
for the byte-swapping code.”

So there were several rewrites from Summer 1984 till Summer 1985. In case the first version was in CLU, it would seem that the change-over to C happened in the very first months of the code base’s lifespan.

The next paragraph as to the state of the code base at this time is revealing:

Unfortunately, a great many invariants are not written down.  Hopefully you
will spend a few weeks understanding the code before you muck with it.  If
something seems easy to add or change, you probably forgot something important.
Almost everything depends on everything else.  It is almost impossible to
devise rigorous test cases.  Innocuous looking changes can have large
performance effects, so watch out.  If you muck with fundamental window
components, a good cross-check is to see how quickly you can manipulate a
window with, say, 100 non-adjacent subwindows.

After reading the above, Jon Steinhart’s post from 5 years ago about X is all the more interesting: https://minnie.tuhs.org/pipermail/tuhs/2017-September/012089.html

[TUHS] Re: Earliest UNIX Workstations?

[Warner Losh]

[-- Attachment #1: Type: text/plain, Size: 4131 bytes --]

On Sat, Jan 28, 2023, 4:05 AM Paul Ruizendaal <pnr@planet.nl> wrote:

>
> > On 28 Jan 2023, at 10:14, Lars Brinkhoff <lars@nocrew.org> wrote:
> >
> > Warner Losh wrote:
> >> Rich Salz wrote:
> >>> I'd like to know what the first versions of X were written in
> >>
> >> Without the earlier versions' source, it's hard to answer this
> question...
> >
> > V source code exists, right?  It seems likely W would have been written
> > in the same language as W.  And that early X would also be the same.
> > Another source of information would be to ask Bob Scheifler and Jim
> > Gettys.
>
> Whilst that is a reasonable assumption, I’m not sure it is true in this
> case. Bob Scheifler writes in 1986:
>
> "We acquired a UNIX-based version of W for the VSlOO (with synchronous
> communication over TCP [24] produced by Asente and Chris Kent at Digital’s
> Western Research Laboratory.”
>
> It does not say “C based”, but it is quite possible that the Unix port
> also meant moving to C.
>
> Also, the work started in June 1984 and had gone to version 10, release 3
> by February 1986. That is 12 versions in 20 months. Most likely X1-X10R2
> are all snapshots done in rapid succession.


X11 is the 11th version of the wire protocol. They bumped that number each
time there was a protocol change. It's not clear that all the early
versions were distributed beyond the local network. The Xlib book stated
something along these lines, but I can't find my copy to quote it or
refresh my recollection.


The change notes for X10R3 read as describing a work still in progress:
>
> http://www.retro11.de/ouxr/43bsd/usr/src/new/X/CHANGES
>
> That “work-in-progress” feel also shows in the Xterm README:
>
> "Xterm is in a reasonably usable state.  We are sick and tired of working
> on it, but there are clearly major areas of improvement possible.  Do
> not look to us to do more than integration work on other people's
> improvement.  About 50% of it is the oldest existing code in the package
> and needing major rewrite.  Our thanks to Bob McNamara for the 50% which
> is solid."
>

Rolling releases were quite common. They went out of style for a while, but
are back in vouge with CI....

The README for the X server itself (written in August 1985 it seems,
> http://www.retro11.de/ouxr/43bsd/usr/src/new/X/X/README) says:
>
> "The server has been completely rewritten several times now, and I am
> reasonably
> happy with it.  I have fine-tuned it specifically for the current
> (sub-optimal)
> VAX compiler.  For other machines, faster code may be obtained in some
> cases
> by changing sizes (e.g., to avoid indexing shifts on the 68000) or register
> declarations.  Attempts to parameterize along these lines have only been
> made
> for the byte-swapping code.”
>
> So there were several rewrites from Summer 1984 till Summer 1985. In case
> the first version was in CLU, it would seem that the change-over to C
> happened in the very first months of the code base’s lifespan.
>

Most likely the CLU library bindings in X10R3 are a hold over from other
software other departments were still using given the fast pace here...

The next paragraph as to the state of the code base at this time is
> revealing:
>
> Unfortunately, a great many invariants are not written down.  Hopefully you
> will spend a few weeks understanding the code before you muck with it.  If
> something seems easy to add or change, you probably forgot something
> important.
> Almost everything depends on everything else.  It is almost impossible to
> devise rigorous test cases.  Innocuous looking changes can have large
> performance effects, so watch out.  If you muck with fundamental window
> components, a good cross-check is to see how quickly you can manipulate a
> window with, say, 100 non-adjacent subwindows.
>
> After reading the above, Jon Steinhart’s post from 5 years ago about X is
> all the more interesting:
> https://minnie.tuhs.org/pipermail/tuhs/2017-September/012089.html


I'd forgotten about that...

Warner

>
>

[-- Attachment #2: Type: text/html, Size: 5979 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Lars Brinkhoff]

Paul Ruizendaal wrote:
>> V source code exists, right?  It seems likely W would have been
>> written in the same language as [V].  And that early X would also be
>> the same.  Another source of information would be to ask Bob
>> Scheifler and Jim Gettys.
>
> Whilst that is a reasonable assumption, I’m not sure it is true in
> this case. Bob Scheifler writes in 1986:
>
> "We acquired a UNIX-based version of W for the [VS1OO] (with
> synchronous communication over TCP [24] produced by Asente and Chris
> Kent at Digital’s Western Research Laboratory.”
>
> It does not say “C based”, but it is quite possible that the Unix port
> also meant moving to C.

I checked V, and it's written in C.

[TUHS] Re: Earliest UNIX Workstations?

[Lars Brinkhoff]

>>> I'd like to know what the first versions of X were written in
>> Without the earlier versions' source, it's hard to answer this question...
>
> V source code exists, right?  It seems likely W would have been written
> in the same language as W.  And that early X would also be the same.
> Another source of information would be to ask Bob Scheifler and Jim
> Gettys.

I asked Bob, and he says W was written in C.

[TUHS] Re: Earliest UNIX Workstations?

[Paul Ruizendaal]

> On 29 Jan 2023, at 07:48, Lars Brinkhoff <lars@nocrew.org> wrote:
> 
>>>> I'd like to know what the first versions of X were written in
>>> Without the earlier versions' source, it's hard to answer this question...
>> 
>> V source code exists, right?  It seems likely W would have been written
>> in the same language as W.  And that early X would also be the same.
>> Another source of information would be to ask Bob Scheifler and Jim
>> Gettys.
> 
> I asked Bob, and he says W was written in C.

Thank you!

In the meantime I have also found the thesis of William Nowicki, the author of VGTS:

https://apps.dtic.mil/sti/pdfs/ADA166935.pdf

It has a timeline for VGTS in its appendix C. In short, development begins in 1982 as a carve out of the display routines of a VLSI design package. It seems to have become usable in 1983 and development continued into 1984 (Nowicki graduated in March 1985).

This places the development of W in 1983 (before that VGTS did not exist, and by early ’84 a Unix version existed).

This https://apps.dtic.mil/sti/pdfs/ADA231239.pdf document from 1990 claims:

"The X Window System has a very alphabetical lineage. The family originated at Stanford University as the VGTS, or V system, a primitive networked graphics windowing system. Then Digital Electronic Corporation desired a more advanced version of V and worked with Stanford University to develop W. Because of the needs of a networking and windowing project sponsored by IBM at the Massachusetts Institute of Technology, MIT acquired the W system and greatly improved its networking capabilities.”

The above seems not quite accurate: besides the V / VGTS mixup, Scheifer writes that W was more a simplification rather than an extension of VGTS:

"VGTS provides graphics windows driven by fairly high-level object definitions from a structured display file; W provides graphics windows based on a simple display-list mechanism, with limited functionality. We acquired a UNIX-based version of W for the VSlOO (with synchronous communication over TCP produced by [Paul] Asente and Chris Kent at Digital’s Western Research Laboratory.”

However, the links with DEC that these paragraphs make are interesting. There is this blog post from Bryan Lunduke (https://lunduke.substack.com/p/w-the-window-system-before-x-that) makes a link between the W window system and DEC’s 1984 "VAXstation Display System Software”. It is possible that these two pieces of software are in fact closely related.

[TUHS] Re: Earliest UNIX Workstations?

[Jonathan Gray]

On Mon, Feb 06, 2023 at 06:01:49PM +1100, Jonathan Gray wrote:
> On Mon, Jan 30, 2023 at 12:20:52AM +0100, Paul Ruizendaal via TUHS wrote:
> > 
> > Herewith some interesting (somewhat) contemporary papers on early windowing systems:
> > 
> > 1. There was a conference in the UK early in 1985 discussing the state of window systems on Unix. Much interesting discussion and two talks by James Gosling, one about NeWS (then still called SunDew), and one about what seems to be SunWindows. It would seem then that these were developed almost in parallel.
> > 
> > http://www.chilton-computing.org.uk/inf/literature/books/wm/contents.htm
> 
> Another window system was Whitechapel Computer Works' Oriel.

further described in:
Dominic Sweetman - A Modular Window System for Unix
http://www.chilton-computing.org.uk/inf/literature/books/wm/p007.htm

Later there were MIPS machines, such as the R2000 based
Hitech 10 which ran UMIPS (4.3BSD) with a choice of X11 and NeWS.
http://web.archive.org/web/20030205212745/http://www.galactic.co.uk/iainf/mg1.html
http://www.umips.net/riscos/index.html

Dominic Sweetman and Nigel Stephens of Whitechapel would
later go on to be involved in Algorithmics, which was acquired by
MIPS Technologies in 2002 and became MIPS UK.
https://cgi.cse.unsw.edu.au/~cs3231/doc/R3000.pdf
https://www.linux-mips.org/wiki/Algorithmics

[TUHS] Re: Earliest UNIX Workstations?

[Jonathan Gray]

On Mon, Jan 30, 2023 at 12:20:52AM +0100, Paul Ruizendaal via TUHS wrote:
> 
> Herewith some interesting (somewhat) contemporary papers on early windowing systems:
> 
> 1. There was a conference in the UK early in 1985 discussing the state of window systems on Unix. Much interesting discussion and two talks by James Gosling, one about NeWS (then still called SunDew), and one about what seems to be SunWindows. It would seem then that these were developed almost in parallel.
> 
> http://www.chilton-computing.org.uk/inf/literature/books/wm/contents.htm

Another window system was Whitechapel Computer Works' Oriel.

"During 1985 major developments took place, mainly on the software front.
A factor of 6 improvement in graphics performance was obtained, and the
Oriel state-of-the-art window manager developed. The Newcastle
Connection and SUN's NFS have also been announced as products, available
on 42NIX, the Whitechapel release of Berkeley BSD 4.2."
http://www.chilton-computing.org.uk/acd/sus/perq_papers/perq_external/p003.htm

http://www.cpu-ns32k.net/Whitechapel.html
has a photo of the GUI running on a MG-1 with NS32016.

https://www.youtube.com/watch?v=JMQ9EvZLSos
"Booting 42nix 2.5 on a Whitechapel Computer Works MG-1"
from Tom Stepleton who also maintained
https://web.archive.org/web/20210625124716/http://mg-1.uk/

https://archive.org/details/byte-magazine-1985-02/1985_02_BYTE_10-02_Computing_and_the_Sciences/page/n378/mode/1up
has an article on Whitechapel Computer Works and the MG-1, before they
changed from Genix to 42nix
"The operating system is Genix, a Berkeley 4.1 UNIX customized by WCW to
support the MG-1's graphics abilities"

http://web.archive.org/web/20030205212745/http://www.galactic.co.uk/iainf/mg1.html
"The WCW MG-1 was launched in September 1984
..
The monitor was a large (17 inch) 1024x800 bit-mapped monochrome screen,
and there was hardware support for up to 16 mask-shift raster operations
at once, modelled after the 3RCC/ICL PERQ raster-ops.
...
It ran a kernel based window system called ORIEL (and a SunTools like
graphics library called ANGEL), which took full advantage of the
raster-op hardware."

[TUHS] Re: Earliest UNIX Workstations?

[Chris Hanson]

On Jan 29, 2023, at 9:23 PM, Jonathan Gray <jsg@jsg.id.au> wrote:
> 
> a different writeup of these is in:
> 
> https://www.tuhs.org/Archive/Documentation/AUUGN/AUUGN-V04.5.pdf
> September 1982, Vol IV No V
> July Usenix Abstracts

I see this also has a bit for Brad Cox's "The Object Oriented Pre-Compiler: Programming Smalltalk 80 Methods in the C Language" which continues to be relevant today though I believe that presentation still used the (quite ugly) original syntax for what became Objective-C.

  -- Chris

[TUHS] Re: Earliest UNIX Workstations?

[Marc Donner]

[-- Attachment #1: Type: text/plain, Size: 919 bytes --]

I was a grad student at CMU at the time.  Raj Reddy told me back then that
it was 1 MIPS, 1 million pixels, 1 megabyte of RAM.

On Mon, Jan 30, 2023, 21:04 Mary Ann Horton <mah@mhorton.net> wrote:

> On 1/25/23 17:01, Larry Stewart wrote:
>
> Jim Morris left PARC and went off to CMU and began talking up the idea of
> the "3M" workstation.  One MIPs, One Megabyte of RAM, and One Million
> Pixels.
>
> The way I heard it, a workstation needed to have Menus, a Mouse, and a 1
> MIPS processor. It also had a megabyte of RAM, at least a megabit LAN, a
> megapixel of display, and cost about a megapenny.
>
> Thanks,
>
>       *Mary Ann Horton* (she/her/ma'am)
>       maryannhorton.com
>
> "This is a great book" - Monica Helms
>
> "Brave and Important" - Laura L. Engel
>
>       Available on Amazon and bn.com!
>
> <https://www.amazon.com/Trailblazer-Lighting-Transgender-Equality-Corporate-ebook/dp/B0B8F2BR9B>
>
>
>

[-- Attachment #2: Type: text/html, Size: 2153 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Paul Ruizendaal]

> On 30 Jan 2023, at 06:23, Jonathan Gray <jsg@jsg.id.au> wrote:
> 
> On Mon, Jan 30, 2023 at 12:20:52AM +0100, Paul Ruizendaal via TUHS wrote:
>> 
>> 
>> Both the UK conference and the NeWS book mention a Unix kernel-based windowing system done at MIT in 1981 or 1982, “NU" or “NUnix”, by Jack Test. That one had not been mentioned before here and may have been the first graphical windowing work on Unix, preceding the Blit. Who remembers this one?
> 
> mentioned in ;login: Volume 7 Number 4, September 1982
> https://archive.org/details/login_september-1982/page/24/mode/2up
> Notes on the Boston USENIX and /usr/group Joint Meeting July 1982
> 
> "NUnix Window System Description, Jack A. Test
> Laboratory for Computer Science, Massachusetts Institute of Technology,
> Room 414, 545 Technology Square, Cambridge, Mass 02139

Appendix E in this report from the University of Illinois describes a port of NUnix to similar hardware. It also contains a walkthrough of a sample workstation session with 17 screen prints:

https://ntrs.nasa.gov/api/citations/19840008755/downloads/19840008755.pdf

Very informative.

[TUHS] Re: Earliest UNIX Workstations?

[Mary Ann Horton]

[-- Attachment #1: Type: text/plain, Size: 737 bytes --]

On 1/25/23 17:01, Larry Stewart wrote:
> Jim Morris left PARC and went off to CMU and began talking up the idea 
> of the "3M" workstation.  One MIPs, One Megabyte of RAM, and One 
> Million Pixels.
The way I heard it, a workstation needed to have Menus, a Mouse, and a 1 
MIPS processor. It also had a megabyte of RAM, at least a megabit LAN, a 
megapixel of display, and cost about a megapenny.

Thanks,

/Mary Ann Horton/ (she/her/ma'am)
maryannhorton.com <https://maryannhorton.com>

"This is a great book" - Monica Helms

"Brave and Important" - Laura L. Engel

       Available on Amazon and bn.com!

	<https://www.amazon.com/Trailblazer-Lighting-Transgender-Equality-Corporate-ebook/dp/B0B8F2BR9B>




[-- Attachment #2: Type: text/html, Size: 1877 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Lars Brinkhoff]

Paul Ruizendaal wrote:
> Both the UK conference and the NeWS book mention a Unix kernel-based
> windowing system done at MIT in 1981 or 1982, “NU" or “NUnix”, by Jack
> Test.

Sounds like Steve Ward's Nu Machine.

[TUHS] Re: Earliest UNIX Workstations?

[Jonathan Gray]

On Mon, Jan 30, 2023 at 04:23:53PM +1100, Jonathan Gray wrote:
> mentioned in ;login: Volume 7 Number 4, September 1982
> https://archive.org/details/login_september-1982/page/24/mode/2up
> Notes on the Boston USENIX and /usr/group Joint Meeting July 1982
> 
> "NUnix Window System Description

...

> 
> goes on to briefly cover:
> 
> Design of an Intelligent Bitmap Terminal
> Rich Fortier and Tony Lake
> Bolt Beranek and Newman Inc., 10 Moulton St., Cambridge, Mass 02238
> 
> The SUN Workstation
> Andreas Bechtolsheim
> Sun Microsystems, 2310 Walsh Ave., Sania Clara, CA 95051
> 
> Merging Bitmap Graphics and UNIX
> Rob Pike
> Bell Labs 2C-521, Murray Hill, NJ 07974
> 
> a different writeup of these is in:
> 
> https://www.tuhs.org/Archive/Documentation/AUUGN/AUUGN-V04.5.pdf
> September 1982, Vol IV No V
> July Usenix Abstracts

more from
https://www.usenix.org/legacy/publications/bibliography/byDate.html
though some of these are character based

      Author: Andreas Bechtolsheim
       Title: The SUN Workstation
       Pages: 61
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Summer 1982
    Location: Boston, MA
 Institution: Sun Microsystems
       Other: Abstract only

      Author: Richard Fortier
      Author: Anthony Lake
       Title: Design of an Intelligent Bitmap Terminal
       Pages: 51-60
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Summer 1982
    Location: Boston, MA
 Institution: Bolt Beranek and Newman Inc.

      Author: Rob Pike
       Title: Merging Bitmap Graphics and UNIX
       Pages: 61
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Summer 1982
    Location: Boston, MA
 Institution: Bell Labs, Murray Hill
       Other: Abstract only

      Author: Jack A. Test
       Title: NUnix Window System Description
       Pages: 45-50
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Summer 1982
    Location: Boston, MA
 Institution: MIT

      Author: David Mankins
      Author: Daniel Franklin
       Title: A Simple Window Management Facility for the UNIX Timesharing System
       Pages: 203-228
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Summer 1983
    Location: Toronto, Ont.
 Institution: Bolt Beranek and Newman, Inc.
https://archive.org/details/login_sept83_issue/page/n23/mode/2up (abstract)

      Author: Michael Shantz
       Title: Graphics Standards for Personal Workstations
       Pages: 257-259
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Winter 1983
    Location: San Diego, CA
 Institution: Sun Microsystems, Inc.
https://archive.org/details/1983-proceedings-unicom-san-diego/page/257/mode/2up

      Author: Steven R. Evans
       Title: Windows with 4.2BSD
       Pages: 260
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Winter 1983
    Location: San Diego, CA
 Institution: Sun Microsystems, Inc.
       Other: Abstract only
https://archive.org/details/1983-proceedings-unicom-san-diego/page/259/mode/2up

      Author: Rex McDowell
       Title: A UNIX-Based Color Graphics Workstation
       Pages: 115-122
   Publisher: USENIX
 Proceedings: USENIX UniForum Conference Proceedings
        Date: January 17-20, 1984
    Location: Washington D.C.
 Institution: Metheus Corporation
https://archive.org/details/1984-proceedings-uni-forum-dc/page/116/mode/2up

      Author: Robert J.K. Jacob
       Title: User-Level Window Manager for UNIX
       Pages: 123-134
   Publisher: USENIX
 Proceedings: USENIX UniForum Conference Proceedings
        Date: January 17-20, 1984
    Location: Washington D.C.
 Institution: Naval Research Laboratory
https://www.cs.tufts.edu/~jacob/papers/uniforum.pdf
https://archive.org/details/1984-proceedings-uni-forum-dc/page/122/mode/2up

      Author: Peter E. Collins
       Title: WINDX - Windows for the UNIX Environment
       Pages: 159-165
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Summer 1984
    Location: Salt Lake City, UT
 Institution: Ithaca Intersystems, Inc.
https://archive.org/details/1984-proceedings-summer-salt-lake-city/page/158/mode/2up

      Author: Chris Torek
      Author: Mark Weiser
       Title: The Maryland Window System
       Pages: 166-172
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Summer 1984
    Location: Salt Lake City, UT
 Institution: University of Maryland
https://archive.org/details/1984-proceedings-summer-salt-lake-city/page/166/mode/2up
https://archive.org/download/CDROM_March92/CDROM_March92.cdr/UNIX_C%2FLANGUAGS%2FC%2FMW.TAR
https://archive.org/details/CDROM_March92
derived in part from Gosling emacs
mw.tar includes source and windows.nr -me paper

[TUHS] Re: Earliest UNIX Workstations?

[Paul Ruizendaal]

> On 30 Jan 2023, at 06:23, Jonathan Gray <jsg@jsg.id.au> wrote:
> 
> mentioned in ;login: Volume 7 Number 4, September 1982
> https://archive.org/details/login_september-1982/page/24/mode/2up
> Notes on the Boston USENIX and /usr/group Joint Meeting July 1982
> 
> "NUnix Window System Description”, Jack A. Test

Thank you for those finds!

The Usenix bibliography gives:

      Author: Jack A. Test
       Title: NUnix Window System Description
       Pages: 45-50
   Publisher: USENIX
 Proceedings: USENIX Conference Proceedings
        Date: Summer 1982
    Location: Boston, MA
 Institution: MIT

My google fu does not drag up those particular proceedings, or those 5 pages. Anybody got access to that?

[TUHS] Re: Earliest UNIX Workstations?

[Jonathan Gray]

On Mon, Jan 30, 2023 at 12:20:52AM +0100, Paul Ruizendaal via TUHS wrote:
> 
> 3. Next is the 1989 NeWS book that has a nice overview and history of windowing systems in its chapter 3:
> 
> http://bitsavers.trailing-edge.com/pdf/sun/NeWS/The_NeWS_Book_1989.pdf
> 
> Both the UK conference and the NeWS book mention a Unix kernel-based windowing system done at MIT in 1981 or 1982, “NU" or “NUnix”, by Jack Test. That one had not been mentioned before here and may have been the first graphical windowing work on Unix, preceding the Blit. Who remembers this one?

mentioned in ;login: Volume 7 Number 4, September 1982
https://archive.org/details/login_september-1982/page/24/mode/2up
Notes on the Boston USENIX and /usr/group Joint Meeting July 1982

"NUnix Window System Description

Jack A. Test
Laboratory for Computer Science, Massachusetts Institute of Technology,
Room 414, 545 Technology Square, Cambridge, Mass 02139

The NUnix Window System is a set of software that provides a user a
basic window management mechanism on a high resolution display.
It was developed for use with the MIT Real Time Systems Group NU
Personal Computer, a 68000-based machine which uses a 820x1024 point
raster-scan display, keyboard, and mouse for the user interface.
The NU machines are being used for developing a multi-font editing
system and drawing facility, in several circuit design projects,
and in the development of new operating system concepts.

The user may create multiple overlapping rectangular windows on the
display. Each is associated with an independent UNIX teletype device
and a display device. A window may have up to eight independent and
changeable font maps. The windows are controlled with ioctl calls
and special signals. These allow such actions as creating a new
window, drawing on it, selecting or changing the fonts associated
with it, reading the state of the mouse, obtaining and/or changing
the state of the window, etc. Each window belongs to a process-group
to which a signal is sent whenever there is a change to the attributes
of the window. There is a window-manager program which makes use
of the mouse device to allow the user to select functions from a set of
displayed menus. The user also has access to the display bitmap and a
special graphics routine library.

The NUnix Window System is implemented in a set of device-driver
routines in the UNIX V7 kernel. Most of the window driver code is
machine independent with the exception of two low-level routines
for driving the raster display and keyboard devices respectively.
The NUnix Window System provides a basic window management mechanism
that: (1) is transparent to the vast majority of user programs, (2)
provides a clean user interface without the addition of any new
system calls, and (3) allows user processes to manage their windows
independently and with minimal kernel-imposed limitations.

The code for the NUnix Window System is available from MIT if you
have a UNIX V7 license. The speaker has submitted a paper for the
Conference Proceedings."

goes on to briefly cover:

Design of an Intelligent Bitmap Terminal
Rich Fortier and Tony Lake
Bolt Beranek and Newman Inc., 10 Moulton St., Cambridge, Mass 02238

The SUN Workstation
Andreas Bechtolsheim
Sun Microsystems, 2310 Walsh Ave., Sania Clara, CA 95051

Merging Bitmap Graphics and UNIX
Rob Pike
Bell Labs 2C-521, Murray Hill, NJ 07974

a different writeup of these is in:

https://www.tuhs.org/Archive/Documentation/AUUGN/AUUGN-V04.5.pdf
September 1982, Vol IV No V
July Usenix Abstracts

[TUHS] Re: Earliest UNIX Workstations?

[Jonathan Gray]

On Mon, Jan 30, 2023 at 12:20:52AM +0100, Paul Ruizendaal via TUHS wrote:
> 
> 3. Next is the 1989 NeWS book that has a nice overview and history of windowing systems in its chapter 3:
> 
> http://bitsavers.trailing-edge.com/pdf/sun/NeWS/The_NeWS_Book_1989.pdf
> 
> Both the UK conference and the NeWS book mention a Unix kernel-based windowing system done at MIT in 1981 or 1982, “NU" or “NUnix”, by Jack Test. That one had not been mentioned before here and may have been the first graphical windowing work on Unix, preceding the Blit. Who remembers this one?

"In April 1981, after the initial breadboard Nu Machines were
constructed at M.I.T. and the UNIX operating system was ported to
it, the design was licensed to Western Digital Corporation to be
re-engineered and marketed as a commercial product. By late 1982,
Western Digital had completed the redesign but decided not to enter
the workstation business. Texas Instruments acquired the Nu
Machine/NuBus technology including the original development group
located in Irvine, California."
pg 1

"The second category of software is the Nu Machine Operating System.
This is a port of Bell Lab’s UNIX Operating System. The Nu Machine
Operating System is derived from UNIX, Seventh Edition and contains,
in addition to the normal UNIX commands and kernel, enhancements
from the Berkeley version of UNIX and a proprietary window system
for the Nu Machine video display subsystem. Of course, the Nu Machine
Operating System will not run on Nu Machines that do not contain a
68010 processor board."
pg 35

https://bitsavers.org/pdf/lmi/LMI_Docs/HARDWARE_1.pdf

also mentioned on page 20 of
http://www.bitsavers.org/pdf/perq/RD_Davis/Davis-PERQ_Workstations_Nov03.pdf

"    MIT Researchers were working on two systems which used the concepts
of the Alto using very different technology. This first was known as the Lisp
Machine project; the MIT Lisp Machine had all the features associated with
a workstation, but was much more expensive. There were two spinoffs of
this project which were less costly lisp machines produced by Symbolics and
LMI.

    The other MIT project which evolved from knowlege of the Alto was
the Nu machine; this was to be a workstation using a Motorolla 68000
microprocessor as its CPU. In various attempts to turn the Nu machine into
a commercially viable product, MIT teamed up with Exxon Enterprises,
then Heath/Schlumberger, and then Western Digital. All of these attempts
failed. The only thing that resulted from this was a deal between MIT
and Western Digital in which Western Digital was to make a Lisp processor
board which would plug into the bus of the Nu machine.

    After some officials from Prime Computer took a tour of MIT and saw
the Nu machine, Lisp Machine, and the Alto, it is runmored that the idea
for the Apollo workstations was formed. The head of engineering at Prime
Computer, William Poduska, started Apollo. Apollo had the added benefit
of much more financial backing than Three Rivers had. Apollo introduced
its Domain processor."

[TUHS] Re: Earliest UNIX Workstations?

[Paul Ruizendaal via TUHS]

Herewith some interesting (somewhat) contemporary papers on early windowing systems:

1. There was a conference in the UK early in 1985 discussing the state of window systems on Unix. Much interesting discussion and two talks by James Gosling, one about NeWS (then still called SunDew), and one about what seems to be SunWindows. It would seem then that these were developed almost in parallel.

http://www.chilton-computing.org.uk/inf/literature/books/wm/contents.htm

2. Then there is a 1986 paper by James Gettys, discussing the 18 month journey towards X10. In particular it focuses on the constraints that Unix set on the design of the X system.

https://www.tech-insider.org/unix/research/acrobat/860201-b.pdf

3. Next is the 1989 NeWS book that has a nice overview and history of windowing systems in its chapter 3:

http://bitsavers.trailing-edge.com/pdf/sun/NeWS/The_NeWS_Book_1989.pdf

Both the UK conference and the NeWS book mention a Unix kernel-based windowing system done at MIT in 1981 or 1982, “NU" or “NUnix”, by Jack Test. That one had not been mentioned before here and may have been the first graphical windowing work on Unix, preceding the Blit. Who remembers this one?

4. Finally, an undated paper by Stephen Uhler discussing the design of MGR is here:

https://sau.homeip.net/papers/arch.pdf

I’ve not included Rob Pike’s papers, as I assume they are well known on this list.

Some of the above papers may be worthy of stable archiving.

[TUHS] Re: Earliest UNIX Workstations?

[Jon Steinhart]

John Cowan writes:
> On Wed, Jan 25, 2023 at 8:47 PM Chris Hanson <cmhanson@eschatologist.net>
> wrote:
>
>
> > > * What do I really mean by workstation? Ex.gr. If an installation had a
> > PDP-11 with a single terminal and operator, is it not a workstation? Is it
> > the integration of display into the system that differentiates?
> >
> > I think "a graphical system intended to be used by a professional to use
> > in their work" is a good starting point for a definition. I should check at
> > home tonight how "A History of Personal Workstations" defines it.
> >
>
> WP says the Terak 8510/a was the first graphical workstation; it came out
> in 1976-77 and ran the UCSD p-System.  I had never heard of it before. The
> first *personal* workstation (non-graphical) was probably the IBM 1620 (aka
> the CADET system, "Can't Add, Doesn't Even Try") from 1959.

I think that WP is not correct here, the Tektronix 4051 beat it to market
by a year.

[TUHS] Re: Earliest UNIX Workstations?

[Jon Forrest]

On 1/25/2023 11:20 PM, John Cowan wrote:

> WP says the Terak 8510/a was the first graphical workstation; it came 
> out in 1976-77 and ran the UCSD p-System.  

It also ran standard RT-11. UC Santa Barbara had a room full of them
that I used to (badly) teach PDP-11 assembler in the early 1980s.
Each student had their own 8" floppy disk which held the OS and
enough utilities to edit and debug programs.

Jon

[TUHS] Re: Earliest UNIX Workstations?

[John Foust via TUHS]

At 01:20 AM 1/26/2023, John Cowan wrote:
>WP says the Terak 8510/a was the first graphical workstation; it came out in 1976-77 and ran the UCSD p-System.  I had never heard of it before. 

I have a dozen or so Teraks (a PDP-11/03 based system) as well as 
many floppies and other inherited items and notebooks from one 
of the Terak founders.  This may seem like a lot but there's another 
guy who might still have a larger collection.

Mini-Unix is described here:

http://www.tavi.co.uk/unixhistory/mini-unix.html

Sixth edition, no MMU.  The Bell memo there is dated January 1977.  

There was a Mini-Unix for the Terak described here in May 1979 but 
I don't think I have a copy.  See page 14...  

https://conservancy.umn.edu/bitstream/handle/11299/159028/UCC_Special%20_Issue_May_1979.pdf?sequence=1&isAllowed=y

Terak floppies are described here:  

http://www.60bits.net/msu/mycomp/terak/termedia.htm

A memo there says they got their copy in April 1980.

There's no indication that this Mini-Unix can *use* the Terak's mono 
bitmapped display, short of writing your own routines.  Pinning "first"
on computers is always a tricky process.

- John

[TUHS] Re: Earliest UNIX Workstations?

[Clem Cole]

[-- Attachment #1: Type: text/plain, Size: 1519 bytes --]

Check out: Magnolia - The Tektronix Unix Personal Computer September 1980
<https://streaklinks.com/BXrcuEsysvYb2V7QEQCkESzw/https%3A%2F%2Farchive.org%2Fdetails%2Fbitsavers_tektronixmliaplan_4739361>
This redates - Sun or Apollo, but not the Perq - a.k.a. the 'Pascalto'

Magonila was done in Tek Labs and it eventually birthed Tek
Workstation Division which was a complete failure, although did Tek
eventually commercialized the Magnolia as the Tek 4404 -- a Smalltalk
system (which you can so a web search).  One of the Busines Schools (HBS I
think) wrote a study about it called 'Why Skunks Don't Work."
ᐧ

On Wed, Jan 25, 2023 at 7:32 PM Joseph Holsten <joseph@josephholsten.com>
wrote:

> It seems like there are bountiful articles able the decline and fall of
> the UNIX workstation, but I’ve had a hard time finding narrative about
> workstations prior to the Stanford SUN workstation.
>
> * was the SUN-1 the first commercially successful product? What are the
> “it depends” edge cases?
> * were there common recipes for proto-workstations within academic or
> industrial research? What did those look like, who was involved?
> * What do I really mean by workstation? Ex.gr. If an installation had a
> PDP-11 with a single terminal and operator, is it not a workstation? Is it
> the integration of display into the system that differentiates?
>
> --
> Joseph Holsten
> http://josephholsten.com
> mailto:joseph@josephholsten.com
> tel:+1-360-927-7234 <(360)%20927-7234>
>

[-- Attachment #2: Type: text/html, Size: 3253 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[arnold]

We had some at Georgia Tech when I was in grad school. The Accent OS for it
from CMU (I think) had a lot of ideas and techniques that showed up
later in Mach.  I once referred to Mach as "Accent running on a vax".

IIRC they had portrait shaped fairly large monochrome bitmapped displays.
But I may be misremembering.

At the time I felt like anything that wasn't Unix / written in C
wasn't worth messing with.

Arnold

Marc Donner <marc.donner@gmail.com> wrote:

> PERQ
>
> There was some talk of making Unix run on it, but POS (PERQ-OS) was written
> in Pascal and there was no reasonable way to port over all of the existing
> stuff without rewriting it all.  I had a PERQ in my office for a while … it
> put out so much heat and noise that my officemates lobbied to have me
> evicted.
>
> On Wed, Jan 25, 2023 at 8:01 PM Larry Stewart <stewart@serissa.com> wrote:
>
> > There was quite a lot of early activity of course!
> > Essentially all of computer science academia was aware of the Xerox Alto,
> > although it wasn't a commercial product and wasn't Unix.
> >
> > Jim Morris left PARC and went off to CMU and began talking up the idea of
> > the "3M" workstation.  One MIPs, One Megabyte of RAM, and One Million
> > Pixels.
> >
> > One of the early commercial attempts was the Three Rivers PERC (or PERQ?)
> > from the Pittsburgh startup.  There were Unix adjacent systems as well,
> > such as Apollo Domain.
> >
> > Of course then Sun got started, and MIPS, and the IBM RT and VaxStations
> > so by the mid '80s it was quite crowded.
> >
> > There is a whole other arc about the graphics workstations, with SGI,
> > Ardent, Stellar, Stardent, and so on.
> >
> > Also, before graphics became affordable, there were various clustered
> > character generator based systems like Convergent Technologies.
> >
> > -L
> >
> > On Jan 25, 2023, at 7:31 PM, Joseph Holsten <joseph@josephholsten.com>
> > wrote:
> >
> > It seems like there are bountiful articles able the decline and fall of
> > the UNIX workstation, but I’ve had a hard time finding narrative about
> > workstations prior to the Stanford SUN workstation.
> >
> >
> > * was the SUN-1 the first commercially successful product? What are the
> > “it depends” edge cases?
> > * were there common recipes for proto-workstations within academic or
> > industrial research? What did those look like, who was involved?
> > * What do I really mean by workstation? Ex.gr. If an installation had a
> > PDP-11 with a single terminal and operator, is it not a workstation? Is it
> > the integration of display into the system that differentiates?
> >
> > --
> > Joseph Holsten
> > http://josephholsten.com
> > mailto:joseph@josephholsten.com
> > tel:+1-360-927-7234 <(360)%20927-7234>
> >
> > --
> =====
> nygeek.net
> mindthegapdialogs.com/home <https://www.mindthegapdialogs.com/home>

[TUHS] Re: Earliest UNIX Workstations?

[Marc Donner]

[-- Attachment #1: Type: text/plain, Size: 2331 bytes --]

PERQ

There was some talk of making Unix run on it, but POS (PERQ-OS) was written
in Pascal and there was no reasonable way to port over all of the existing
stuff without rewriting it all.  I had a PERQ in my office for a while … it
put out so much heat and noise that my officemates lobbied to have me
evicted.

On Wed, Jan 25, 2023 at 8:01 PM Larry Stewart <stewart@serissa.com> wrote:

> There was quite a lot of early activity of course!
> Essentially all of computer science academia was aware of the Xerox Alto,
> although it wasn't a commercial product and wasn't Unix.
>
> Jim Morris left PARC and went off to CMU and began talking up the idea of
> the "3M" workstation.  One MIPs, One Megabyte of RAM, and One Million
> Pixels.
>
> One of the early commercial attempts was the Three Rivers PERC (or PERQ?)
> from the Pittsburgh startup.  There were Unix adjacent systems as well,
> such as Apollo Domain.
>
> Of course then Sun got started, and MIPS, and the IBM RT and VaxStations
> so by the mid '80s it was quite crowded.
>
> There is a whole other arc about the graphics workstations, with SGI,
> Ardent, Stellar, Stardent, and so on.
>
> Also, before graphics became affordable, there were various clustered
> character generator based systems like Convergent Technologies.
>
> -L
>
> On Jan 25, 2023, at 7:31 PM, Joseph Holsten <joseph@josephholsten.com>
> wrote:
>
> It seems like there are bountiful articles able the decline and fall of
> the UNIX workstation, but I’ve had a hard time finding narrative about
> workstations prior to the Stanford SUN workstation.
>
>
> * was the SUN-1 the first commercially successful product? What are the
> “it depends” edge cases?
> * were there common recipes for proto-workstations within academic or
> industrial research? What did those look like, who was involved?
> * What do I really mean by workstation? Ex.gr. If an installation had a
> PDP-11 with a single terminal and operator, is it not a workstation? Is it
> the integration of display into the system that differentiates?
>
> --
> Joseph Holsten
> http://josephholsten.com
> mailto:joseph@josephholsten.com
> tel:+1-360-927-7234 <(360)%20927-7234>
>
> --
=====
nygeek.net
mindthegapdialogs.com/home <https://www.mindthegapdialogs.com/home>

[-- Attachment #2: Type: text/html, Size: 4314 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Paul Ruizendaal via TUHS]

>> * What do I really mean by workstation? Ex.gr. If an installation had a 
>> PDP-11 with a single terminal and operator, is it not a workstation? Is 
>> it the integration of display into the system that differentiates?
> 
> I remember people calling something a workstation,
> if it has the four "M"
> 
> at least 1 MByte memory
> at least 1 megapixel display
> at least 1 mbit/s network
> can't remember the fourth(was there a fourth?)

I remember it as:

at least 1 MByte memory
at least 1 megapixel display
at least 1 MIPS
cost at most 1 mega penny (10K, maybe 35K in today’s money)

That matches with Wikipedia, for whatever that is worth: https://en.wikipedia.org/wiki/3M_computer
but note that it talks about 3M not 4M.

With hindsight, not adding in networking speed looks strange -- but maybe the world had already settled on LAN speeds above 1Mb/s by 1980 (Ethernet, ARCNet)

[TUHS] Re: Earliest UNIX Workstations?

[Jaap Akkerhuis via TUHS]

[-- Attachment #1: Type: text/plain, Size: 58 bytes --]



> can't remember the fourth(was there a fourth?)

Mips


[-- Attachment #2: Message signed with OpenPGP --]
[-- Type: application/pgp-signature, Size: 267 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Rob Pike]

[-- Attachment #1: Type: text/plain, Size: 611 bytes --]

at least 1 million words of marketing?

-rob


On Thu, Jan 26, 2023 at 8:54 PM emanuel stiebler <emu@e-bbes.com> wrote:

> On 2023-01-25 19:31, Joseph Holsten wrote:
>
> > * What do I really mean by workstation? Ex.gr. If an installation had a
> > PDP-11 with a single terminal and operator, is it not a workstation? Is
> > it the integration of display into the system that differentiates?
>
> I remember people calling something a workstation,
> if it has the four "M"
>
> at least 1 MByte memory
> at least 1 megapixel display
> at least 1 mbit/s network
> can't remember the fourth(was there a fourth?)
>
>

[-- Attachment #2: Type: text/html, Size: 1239 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[emanuel stiebler]

On 2023-01-25 19:31, Joseph Holsten wrote:

> * What do I really mean by workstation? Ex.gr. If an installation had a 
> PDP-11 with a single terminal and operator, is it not a workstation? Is 
> it the integration of display into the system that differentiates?

I remember people calling something a workstation,
if it has the four "M"

at least 1 MByte memory
at least 1 megapixel display
at least 1 mbit/s network
can't remember the fourth(was there a fourth?)

[TUHS] Re: Earliest UNIX Workstations?

[Dave Horsfall]

[-- Attachment #1: Type: text/plain, Size: 571 bytes --]

On Thu, 26 Jan 2023, John Cowan wrote:

> WP says the Terak 8510/a was the first graphical workstation; it came 
> out in 1976-77 and ran the UCSD p-System.  I had never heard of it 
> before. The first *personal* workstation (non-graphical) was probably 
> the IBM 1620 (aka the CADET system, "Can't Add, Doesn't Even Try") from 
> 1959.

Mmm...  Would a GT-40 connected to a PDP-11/40 count as a workstation?  I 
started using that around 1975 (along with serial terminals, of course), 
in my first job at Uni of NSW (after graduating from there).

-- Dave

[TUHS] Re: Earliest UNIX Workstations?

[John Cowan]

[-- Attachment #1: Type: text/plain, Size: 813 bytes --]

On Wed, Jan 25, 2023 at 8:47 PM Chris Hanson <cmhanson@eschatologist.net>
wrote:


> > * What do I really mean by workstation? Ex.gr. If an installation had a
> PDP-11 with a single terminal and operator, is it not a workstation? Is it
> the integration of display into the system that differentiates?
>
> I think "a graphical system intended to be used by a professional to use
> in their work" is a good starting point for a definition. I should check at
> home tonight how "A History of Personal Workstations" defines it.
>

WP says the Terak 8510/a was the first graphical workstation; it came out
in 1976-77 and ran the UCSD p-System.  I had never heard of it before. The
first *personal* workstation (non-graphical) was probably the IBM 1620 (aka
the CADET system, "Can't Add, Doesn't Even Try") from 1959.

[-- Attachment #2: Type: text/html, Size: 1399 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Chris Hanson]

On Jan 25, 2023, at 4:31 PM, Joseph Holsten <joseph@josephholsten.com> wrote:
> 
> It seems like there are bountiful articles able the decline and fall of the UNIX workstation, but I’ve had a hard time finding narrative about workstations prior to the Stanford SUN workstation.
> 
> * was the SUN-1 the first commercially successful product? What are the “it depends” edge cases?

Nope, Apollo and the Three Rivers PERQ both predate Sun. And while one can certainly question PERQ's commercial success I don't think one can question Apollo's—they were done in by HP's purchase, not by market forces.

Apollo Aegis wasn't actually UNIX, but was very UNIX-like since it was designed by people who had worked at Prime on PrimOS and on MULTICS before that. And later releases provided more specifically UNIX-style environments for users as an alternative to Aegis.

The Three Rivers PERQ had several different environments available, including multiple UNIX implementations.

> * were there common recipes for proto-workstations within academic or industrial research? What did those look like, who was involved?

They mostly looked like the Xerox Alto, once it made waves in the research community.

> * What do I really mean by workstation? Ex.gr. If an installation had a PDP-11 with a single terminal and operator, is it not a workstation? Is it the integration of display into the system that differentiates? 

I think "a graphical system intended to be used by a professional to use in their work" is a good starting point for a definition. I should check at home tonight how "A History of Personal Workstations" defines it. Note that I wouldn't necessarily require a workstation to be primarily used by a single user; lots of 1970s CAD systems supported multiple user workstations on a single computer.

  -- Chris

[TUHS] Re: Earliest UNIX Workstations?

[Jon Steinhart]

Luther Johnson writes:
> I think what made 'workstations' different was:
> 
> 1) You had your own computer, you were not just using a terminal connected to a
> computer, that was time-shared with a bunch of other users

While it wasn't UNIX but a stupidly lost market opportunity
the Tektronix 4051 was one of the first workstations.

[TUHS] Re: Earliest UNIX Workstations?

[Tom Lyon]

[-- Attachment #1: Type: text/plain, Size: 1119 bytes --]

I believe the term "workstation" came from the integrated interactive
engineering design terminals/systems.  Can't remember details, but
companies like ComputerVision had had them before UNIX and 68000 were a
thing,

On Wed, Jan 25, 2023 at 7:31 PM Joseph Holsten <joseph@josephholsten.com>
wrote:

> It seems like there are bountiful articles able the decline and fall of
> the UNIX workstation, but I’ve had a hard time finding narrative about
> workstations prior to the Stanford SUN workstation.
>
> * was the SUN-1 the first commercially successful product? What are the
> “it depends” edge cases?
> * were there common recipes for proto-workstations within academic or
> industrial research? What did those look like, who was involved?
> * What do I really mean by workstation? Ex.gr. If an installation had a
> PDP-11 with a single terminal and operator, is it not a workstation? Is it
> the integration of display into the system that differentiates?
>
> --
> Joseph Holsten
> http://josephholsten.com
> mailto:joseph@josephholsten.com
> tel:+1-360-927-7234 <(360)%20927-7234>
>

[-- Attachment #2: Type: text/html, Size: 2220 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Luther Johnson]

[-- Attachment #1: Type: text/plain, Size: 3108 bytes --]

I think what made 'workstations' different was:

1) You had your own computer, you were not just using a terminal 
connected to a computer, that was time-shared with a bunch of other users

2) A very capable large bit-mapped display, hence the variation on this 
term, "graphic workstation"

3) Networking to shared filesystems - in time sharing, you're all on one 
computer, going to the same files on the same disk(s), in a network of 
workstations, you're all going to a network with filesystems mounted on 
disks on machines all over the network. Some workstations had disks and 
filesystems, some did not - hence another variation, "disk-less 
workstation", which booted from the network. "Servers" were big machines 
with big disks, really the same kinds of machines, basically, but 
configured to be effective at hosting resources and people were 
discouraged from logging in and camping out on them.

That's how I remember it, the machines and capabilities were also 
accompanied by the different modes of resource use that they enabled.

On 01/25/2023 05:51 PM, segaloco via TUHS wrote:
> Nothing says a PDP-11 has to be in a rack with peripherals right? If 
> one could devise up emulated peripherals that plug into the backplane 
> or otherwise tear them down to fit in much less space, a PDP-11 could 
> probably be made to inhabit similar desk real-estate as a workstation, 
> especially some of the smaller LSI models. There's also the MicroVAXen 
> but the SUN-1 beats those to the market.
>
> All in all, I would wager workstation has never been a well regulated 
> term and, especially once PCs and other micros got better, the 
> delineation between a workstation and a consumer PC has just gotten 
> blurrier and blurrier. For instance, I would use the term workstation 
> to apply to my Raspberry Pi, someone else would probably chuckle at 
> the thought while sitting at their modern POWER9 system. It fits all 
> the needs of my non-day-job computing, so workstation enough for me.
>
> - Matt G.
> ------- Original Message -------
> On Wednesday, January 25th, 2023 at 4:31 PM, Joseph Holsten 
> <joseph@josephholsten.com> wrote:
>
>> It seems like there are bountiful articles able the decline and fall 
>> of the UNIX workstation, but I’ve had a hard time finding narrative 
>> about workstations prior to the Stanford SUN workstation.
>>
>> * was the SUN-1 the first commercially successful product? What are 
>> the “it depends” edge cases?
>> * were there common recipes for proto-workstations within academic or 
>> industrial research? What did those look like, who was involved?
>> * What do I really mean by workstation? Ex.gr. If an installation had 
>> a PDP-11 with a single terminal and operator, is it not a 
>> workstation? Is it the integration of display into the system that 
>> differentiates?
>>
>> --
>> Joseph Holsten
>> http://josephholsten.com <http://josephholsten.com/>
>> mailto:joseph@josephholsten.com <mailto:joseph@josephholsten.com>
>> tel:+1-360-927-7234 <tel:%28360%29%20927-7234>
>


[-- Attachment #2: Type: text/html, Size: 6311 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[Larry Stewart]

[-- Attachment #1: Type: text/html, Size: 3429 bytes --]

[TUHS] Re: Earliest UNIX Workstations?

[segaloco via TUHS]

[-- Attachment #1: Type: text/plain, Size: 1826 bytes --]

Nothing says a PDP-11 has to be in a rack with peripherals right? If one could devise up emulated peripherals that plug into the backplane or otherwise tear them down to fit in much less space, a PDP-11 could probably be made to inhabit similar desk real-estate as a workstation, especially some of the smaller LSI models. There's also the MicroVAXen but the SUN-1 beats those to the market.

All in all, I would wager workstation has never been a well regulated term and, especially once PCs and other micros got better, the delineation between a workstation and a consumer PC has just gotten blurrier and blurrier. For instance, I would use the term workstation to apply to my Raspberry Pi, someone else would probably chuckle at the thought while sitting at their modern POWER9 system. It fits all the needs of my non-day-job computing, so workstation enough for me.

- Matt G.
------- Original Message -------
On Wednesday, January 25th, 2023 at 4:31 PM, Joseph Holsten <joseph@josephholsten.com> wrote:

> It seems like there are bountiful articles able the decline and fall of the UNIX workstation, but I’ve had a hard time finding narrative about workstations prior to the Stanford SUN workstation.
>
> * was the SUN-1 the first commercially successful product? What are the “it depends” edge cases?
> * were there common recipes for proto-workstations within academic or industrial research? What did those look like, who was involved?
> * What do I really mean by workstation? Ex.gr. If an installation had a PDP-11 with a single terminal and operator, is it not a workstation? Is it the integration of display into the system that differentiates?
>
> --
> Joseph Holsten
> [http://josephholsten.com](http://josephholsten.com/)
>
> mailto:joseph@josephholsten.com
>
> tel:[+1-360-927-7234](tel:(360)%20927-7234)

[-- Attachment #2: Type: text/html, Size: 3508 bytes --]