Jon Elson wrote:

Michael A. Terrell wrote:

No one here has ever played with any Data General computers? How
about Prime? National Semiconductor? Metrodata? Intel Multibus? How
about VME/VXI?



Sure, I started fooling around with PDP-8's, and worked quite a bit with
a PDP-5, a discrete transistor 12-bit machine essentially identical with
the PDP-8 (no suffix) except for physical size, the -5 was about twice
the size. I also did a bit of work with the LINC, a computer designed
at MIT and built by grad students at a summer session there. Some 50+
were built and then shipped back to the home institutions of those
students, and were used for many years. Again, 12-bit word width, but
ones-complement arithmetic, discrete transistors, XY oscilloscope
display and block-oriented 3/4" magnetic tapes. That instruction set
was later incorporated into DEC's LINC-8 and PDP-12 systems. All that
was at Washington University in St. Louis. I also did some learning on
IBM 360's there. Later I worked on PDP-11s, Vaxes and Alpha systems there.

I worked with Data General Novas at University of Missouri at Rolla.

I built a multiprocessor system with the National Semi 16032 (later
renamed the 32016). It ran up to 7 of these computers with a 2 MB
common memory and using a Vax 11/780 for I/O. About the time I got it
really running well, the faster microvaxes came along, and it was
obsolete. It is still in my basement for sentimental reasons. It used
the extended Multibus-I standard for connection to the VAX and memory
boards, as well as power.

We had a Pr1me (that's a numeral one in the middle, not a letter "i")
here but I never worked with it. I did work with an Interdata 7/32
which was used to run an insane 3rd party disk controller for our
PDP-11s. it enabled us to run Calcomp 40 MB 14" hard drives on our PDP.
The instruction set looked a lot like the IBM 360. I think Pr1me was an
offshoot of Interdata.


We had a Prime, with an 'I' at Microdyne. A very odd system that
handled the MRP, Accounting, sales, personnel & inventory on about 50
terminals.

The first time I heard of Prime was Wane Green, a publisher of
Amateur Radio magazine 73, and Byte magazine in its early days. he was
sold a prime to keep his subscription information, and he went into
great deal about all of the system's flaws. Like the fact that the
system couldn't reuse drive space from deleted files.


I worked on a MicroData microprogrammable minicomputer at Rolla. I even
worked on a Vacuuum-tube computer, the Bendix G-15 at Wash U, but we
were never able to get it running. The magnetic drum was badly scratched.

The first computer I built for myself had an 8008 CPU, but too little
memory to do anything worthwhile. I then built an S-100 system with
8080, and then moved up to a Z-80. I had all sorts of stuff interfaced
to that, including a Honeywell 14.5" drum printer that was bigger than a
chest freezer, 9-track mag tape, a 12" XY display with light pen, and a
10 MB Memorex hard drive. I built a 32-bit AMD bit slice computer on
two huge wire wrap boards, but writing the microcode for it was a huge
project. I did get it running, and had a bunch of test programs, but it
was light years from becoming a usable general-purpose computer. I made
a copy of a Nat Semi 32016 CPU and built a working Genix system, but it
was insanely slow. many things were 5 times slower than my Z-80! I did
use it for at least a year, though. This must have been around 1986,
when I got the opportunity to buy a MicroVax CPU board from a broker,
and get the rest of the needed stuff from 3rd party vendors.

Jon


--
You can't have a sense of humor, if you have no sense!