rarest computers. was: RE: Xerox Alto Restoration + Emulation

From: Rick Bensene <rickb_at_bensene.com>
Date: Tue Aug 3 09:31:16 2004

Actually, there were some other machines between the 6132 and the XD88.
These machines
were the 4300-series, like the 4319. The 4300-Series machines I believe
used 68020's, IIRC.
I have a 4319 around here somewhere -- I think it still runs. These
also ran UTEK.

I worked at Tek during the days of ECS (Engineering Computing Systems).
There were a lot
of VERY talented people there.



Early on, sometime in the late 1970's, Tek developed
a very interesting machine called Magnolia. It was never productized to
my knowledge,
as there were all kinds of problems getting senior management of an
oscilloscope and
test equipment company to buy into the idea of making a computer. The
Magnolia used a
custom bit-slice CPU that was very fast for the time. It had an amazing
display processor also.
The machine was built into a pedastal that a 19" or so raster display
sat on top of. It used
solid-state main memory, and a Micropolis 8" hard disk mounted in the
bottom of the base.
It ran Smalltalk, with a completely in-house developed windowing desktop
environment, reminiscent of the Xerox PARC implementations. Tek made
quite a few of these machines that were used in-house to develop on. The
project was eventually cancelled. Don't know what happened to the
machines that were in existence at the time. I'd be surprized if any
still exist today.

Early on at Tek, there was a lot of PDP-8's running OS-8, and PDP-11's
running RT11 and RSX as
test system controllers, and even one system called the "DPO", which was
a digitizer
interfaced to a PDP-11 (not user which, probably a /35) that had a Tek
DVST terminal
hooked up to it. DPO stood for "Digital Processing Oscilloscope". They
made a dialect
of BASIC that had primitives in it for doing signal processing on
acquired signals.
It was a BIG piece of equipment, but could do some amazing signal
capture and analysis
for the time. I think that the DPO ran RT-11, if I remember correctly.

Later, Tek developed a microprocessor development system that utilized
an LSI-11 core.
It was called the 8250. The 8250 used an 8" Micropolis drive, and ran a
modified
AT&T System III Unix operating system. This was all done under the
microprocessor development
products group (MDP). These machines provided a basis by which Unix
gained a foothold
in the company.

Then, ECS was formed at the Tek Wilsonville plant. Their idea of what
they were building
was a workstation-class UNIX machine that'd blow the socks off of
everyting out there.
However, the politics of the situation was that management expected them
to build an instrument
controller for automated test environments. The result was the 6130,
which was rather weak
as a workstation (no display subsystem, only RS-232 serial to a
terminal), but worked
nicely as an instrument controller with its built-in GPIB (IEEE-488)
interface. Attempts
were made to develop a display subsystem for the machine...a big
daughter board that plugged
in on-top of the main CPU board. Prototypes were built, and mods to
UTEK were made
to support it, and even a primitive window system was developed.
However, Tek's instrument
mentality wouldn't let the project continue, and it was cancelled.

The 4132 follow-on to the 6130 was done mainly because of the
proliferation of SCSI.
The 6130 used ST-506 disks (though a SCSI add-on board was available),
which were starting
to become less popular. The 4132 replaced the ST-506 interface with an
integrated SCSI
controller, and replaced the floppy disc drive with a SCSI 1/4"
streaming cartridge tape drive.
This made OS loading a LOT faster. To load the OS on the 6130, you had
to load a big stack
of floppies! The 4132 didn't have a video system, just RS-232 ports for
console communications.

Then the 43xx workstations came out. These were true Unix workstations,
running a new
version of Utek, with a full X-Windows implementation and a decent
display processor.
They were quite capable machines for their time. The 4315 was marketed
as an "AI workstation"
with a smalltalk environment. The 4319 was a higher-end Unix
workstation, inteded for
CAD and software development.

The last hurrah for ECS was the XD88. There were MASSIVE arguments
internally about what
CPU to use. There was talk of the Z8000, but eventually the Motorola
88000 won out.
The XD88's were really powerful machines, designed chiefly for
electrical and mechanical
CAD design.

Sadly, Tektronix never really figured out that it had the capacity to
build nice computers.
The instrumentation and terminals mentality of the company kept the
thinking limited to these
areas. Sales folks had a hard time relating to the concept of a
computer workstation.
A lot of money and effort was expended to put Tektronix on the map in
the computer biz.
But, in the end, it was all for naught. ECS was disbanded. Some of the
folks left
to form other local computer companies, including BIIN among others.

All from memory, there may be some errors.

-Rick

Had Tek developed the Magnolia and started selling it as a real
workstation.

>
> That's interesting. I suppose my XD88 was the replacement for
> the 4132. It runs UTEK too, but of course on a newer 88k CPU.
> Unfortunately mine's only got the 8 bit display board in it,
> not the 24 bit. I believe that Tek sold hardly any of them,
> simply because they were way too expensive
> - very nice machines otherwise.
>
> cheers
>
> Jules
>
>
Received on Tue Aug 03 2004 - 09:31:16 BST