?

Log in

No account? Create an account

Previous Entry | Next Entry

For the first couple of decades of my life, computers as we know them today were exotic beasts that filled rooms, each requiring the care of a cadre of what were then called systems programmers. Therefore, in my single-digit years the closest thing to a computer that I laid my hands on was a typewriter-sized electromechanical calculator that did addition, subtraction, multiplication, and division. I had the privilege of using this captivating device when helping out with accounting at the small firm at which my mother and father worked.

I was an early fan of hand-held computing devices. In fact, I was in the last math class in my high school that was required to master a slide rule, of which I still have several. I also learned how to use an abacus, including not only addition and subtraction, but multiplication and division as well. Finally, I had the privilege of living through the advent of the electronic pocket calculator. My first pocket calculator was a TI SR-50, which put me firmly on the infix side of the ensuing infix/Polish religious wars.

But none of these qualified as “real computers”.

Unusually for an early 1970s rural-Oregon high school, mine offered computer programming courses. About the only thing I knew about computers were that they would be important in the future, so I signed up. Even more unusually for that time and place, we got to use a real computer, namely an IBM 360. This room-filling monster was located fourteen miles (23 kilometers) away at Chemeketa Community College. As far as I know, this was the closest computer to my home and school. Somehow my math teacher managed to wangle use of this machine on Tuesday and Thursday evenings, and he bussed us there and back.

This computer used punched cards and a state-of-the-art chain lineprinter. We were allowed to feed the card reader ourselves, but operating the lineprinter required special training. This machine's console had an attractive red button labeled EMERGENCY PULL. The computer's operator, who would later distinguish himself by creating a full-motion video on an Apple II, quite emphatically stated that this button should be pulled only in case of a bona fide emergency. He also gave us a simple definition of “emergency” that featured flames shooting out of the top of the computer. I never did see any flames anywhere near the computer, much less shooting out of its top, so I never had occasion to pull that button. But perhaps the manufacturers of certain incendiary laptops should have equipped each of them with an attractive red EMERGENCY PULL button.

Having provided us the necessary hardware training, the operator then gave us a sample card deck. We were to put our program at one specific spot in the deck, and our input data in another. Those of us wishing more information about how this worked were directed to an impressively large JCL manual.

The language of the class was FORTRAN, except that FORTRAN was deemed to difficult an initial language for our tender high-school minds. They therefore warmed us up with assembly language. Not IBM's celebrated Basic Assembly Language (BAL), but a simulated assembly language featuring base-10 arithmetic. After a couple of sessions with the simulated assembly, we moved up to FORTRAN, and even used PL/1 for one of our assignments. There were no error messages: There were instead error numbers that you looked up in a thick printed manual located in the same bookcase containing the JCL manual.

I was surprised by the computer's limitations, especially the 6-to-7 digit limits for single-precision floating point. After all, even my TI SR-50 pocket calculator did ten digits! That said, the computer could also do alphabetic characters (but only upper case) and a few symbols—though the exclamation point was notably missing. The state-of-the-art 029 keypunches were happy to punch an exclamation mark, but alas! It printed as “0” (zero) on the lineprinter.

I must confess that I was not impressed with the computer. In addition to its arithmetic limitations, its memory was quite small. Most of our assignments were small exercises in arithmetic that I could complete much more quickly using paper and pencil. In retrospect, this is not too surprising, given that my early laissez-faire programming methodology invariably resulted in interminable debugging sessions. However, it was quite clear that computers were becoming increasingly important, and I therefore resolved to take the class again the following year.

So, the last time I walked out of that machine room in Spring of 1974, I fully expected to walk back the following Fall. Little did I know that it would be almost 30 years before I would once again write code for an IBM mainframe. Nor did I suspect that it would be more than 15 years before work started on the operating system that was to be running on that 30-years-hence mainframe.

My limited foresight notwithstanding, somewhere in Finland a small boy was growing up.

Comments

(Anonymous)
Sep. 5th, 2018 08:10 pm (UTC)
Details?
But *which model* of IBM 360?

And was the simulated assembly language Knuth's MIX? UCLA used a MIX simulator on a 360/91 to introduce assembly language concepts back then.
paulmck
Sep. 5th, 2018 09:18 pm (UTC)
Re: Details?
I don't recall the exact model, but I believe that it was somewhere in the 30s or 40s. Nothing anywhere near as fancy as as a 360/91!

And the simulated assembly language was most definitely not MIX, though I code up part of a MIX simulator a couple of years later. (More on that later.)

I believe that the simulated assembly language was home-brew, and I have not come across it since. It had the usual load/store/arithmetic opcodes, and there were exactly 100 memory locations. I do not recall whether or not the instructions took up any of those memory locations. Nor do I recall the width of each memory location.

Lost to the mists of time! :-)