• Please review our updated Terms and Rules here

In search of an IBM 5100 (c 1975)

PRChristy

New Member
Joined
Aug 29, 2022
Messages
1
’m hoping to find one that is physically in good shape (e.g., the keyboard “works”) but I don’t presently have dreams of making the original electronics work again (more later).

But first a little background. I guess I’m vintage as well. I learned about computers from a 1945 book “Giant Brains” that I read in junior-high, and went on to win a Science Fair in 1959 building a two bit working relay “computer.” I first wrote a real program in 1962 at an NSF high-school math program at UCLA (a number theory game). In the fall of 1962 my math buddy and I wrote a FORTRAN program to print card-stunt instructions — I went to a big high school, and through my father had access to an 1BM 7090 computer.

In college I switched from physics to very early computer science, and went to graduate school at Berkeley, following Butler Lampson who I had met as a fellow physics programmer of a PDP-1 at the Cambridge Electron Accelerator. Butler was part of a stellar project — Genie — that developed one of the first commercial time sharing systems on an adapted SDS 930 (Tymshare built their service on this system). Genie included people like L. Peter Deutch (the best programmer I ever met) and Charles Simonyi (later of Word fame) and Chuck Grant (founder of NorthStar Computers). After a couple of years I took at summer job at CSC working on the 1108 operating system that became their CSTS service and stayed there. Then a couple of years at UCSF Office of Medical Information Systems (another remarkable group of colleagues) and then off to DEC for a decade. I started at DEC as technical staff to the VP of software, and then worked in the semiconductor engineering group, and then was part of the project with Gene and Carl Amdahl’s Trilogy IBM/360 clone effort. That’s my “vintage” history (up to 1984) and the genesis of the interest in the IBM 5100.

IBM introduced the 5100 soon after I arrived at DEC and invoked what I was told was an already established reaction: Oh my God! IBM is coming after us! Abandon all hope! Sell the s stock. The 5100 was a luggable desktop computer (looked like a Wang computer) with a keyboard and small CRT display. It ran built in BASIC and APL interpreters. There was a nice cartridge tape drive, and it exuded IBM’s mechanical and design elegance. No wonder it was frightening.

But then Ken Olsen, DEC’s CEO, looked at it and scoffed “the font is too small! You can’t read it!” He was right for the market as a whole. You see Ken was older than the engineering corps and the only one with middle aged vision degradation. In the end DEC survived (although the eventual IBM PC was an amazing business).

In any case, that’s why I’m looking for a 5100. The current idea is to drive it with a PI emulator or at least a Pi BASIC and APL interpreters. I would love to hear feedback on all of this.

Peter.christy@gmail.com
 
"But then Ken Olsen, DEC’s CEO, looked at it and scoffed “the font is too small! You can’t read it!” He was right for the market as a whole. You see Ken was older than the engineering corps and the only one with middle aged vision degradation. "

I remember saying the same thing when I first saw the 5100 at NCC. But then I also said the same thing when I saw the prototype O1 spread out on a table at Sorcim.

I have terrible vision.
 
  • Like
Reactions: ACK
I'm surprised about complaints about the screen - the 5100 had support for an external screen in the back, through a BNC connector. It's the only way I still use one of my 5110's today (via a BNC to VGA adapter, but I also found a Samsung LCD that happens to also have a straight BNC connector input), since it's the internal CRT stopped working (so if you find one of these, don't give up on it if the screen doesn't fire up).

At least I've always referred to this just as a "BNC connector", maybe back in the day it was called something else?

Did not many people have that type of TV input back then? Or maybe IBM didn't do a very good job of advertising the capability? I know for lab or test range facilities, the ability to mirror an operator screen a long distance might be important (to show off real-time results or telemetry data to the brass or VIPs) -- I'm not sure how far that kind of coax could go, maybe 1000 ft?

Or I imagine in aftermarket, one could just rig up a kind of magnifying glass to sit in front of the 5" CRT :D


I've always been impressed with the Wang 2200. I'm not sure how capable the original 2200 was in 1973 (since they quickly moved onto "B", "T", "LVP" models in the subsequent year -- and I tend to hear more about those systems, than the original). But the Wang 2200 seemed to be a dumb terminal with a processor box that sat below it under the desk -- but still, a rather complete and functional PC, I think around $7500 originally. But (from the perspective of the 1970s) keeping the family name for the product I think was maybe their mistake.
 
  • Like
Reactions: ACK
Any screen smaller than 9" and meant for real work is tough on the eyes.. Even Young people will pay the price for it years later.
 
  • Like
Reactions: ACK
Well, studio monitors were composite-video input, but otherwise in the wild they were pretty rare at the time. And BNC was not always the rule--many still had UHF (PL-239) connectors.
I recall doing a simple sync combiner for the marketing group for our monochrome 8085 system so they could show the displays on a large-screen monitor.
 
  • Like
Reactions: ACK
The current idea is to drive it with a PI emulator or at least a Pi BASIC and APL interpreters.

The Executive ROS for the IBM 5100 has been extracted, but I stalled out in figuring out how it handles the Executive ROS to Language ROS transition (the PALM emulator code we have does run the Executive ROS binary, but then fails either in some Common or Language ROS CRC check, or in doing the actual hand-off to the language ROS).

We can drive keyboards inputs from a Pi-type device -- and that can be done somewhat un-intrusively by connecting to the pins on the backside of A1 board (i.e. don't have to remove keyboard).

But, if you find a 5100, be careful tinkering too much inside a "first of a kind" series of a system. Not just since they're old, but since the engineering of them might not be as refined as later models.

I always enjoy to hear stories of how the 5100 or 5110 were used, or how they were perceived in the 1970s.
 
  • Like
Reactions: ACK
BTW, I've been curious why both the Wang and the IBM 5100 chose 64x16 screens. Was this a punch card form factor in the past, or some kind of existing VT-xyz standard?
 
  • Like
Reactions: ACK
It's probably nothing more complicated than 16 * 64 = 1024.

If you want to spend an entire kilobyte on the video display without any waste, and if you have a display with standard 4:3 dimensions, it's pretty much the only reasonable choice.
 
  • Like
Reactions: ACK
Back
Top