Core Memory - down memory lane

[GreyGeek]

After I graduated from High School in 1959 I attended the Barnes School of Business in Denver, CO., to study "Data Processing" -- the IBM 540 Gangpunch, 402 Tabulator, IBM 704, etc. The ferrite core memory was the key storage device at the time:

After I graduated I went looking for a job in the field but no one would hire me because while I was 18 I looked like I was 14. So, I ended up going to college instead. Lucky for me. In grad school we programmed a Burroughs B2500 (IIRC) using Fortran IV written on punch tape using the ASR-33 paper punch keyboard.

[woodsmoke]

woodsmoke

[Qqmike]

I remember such things, too, starting college 1967, doing FORTRAN IV, on a mainframe IBM 360?, punched cards, etc. What I remember is that we were taught to write FORTRAN programs in such a way that they used minimal memory! as "minimal" as possible, that is ... how to re-write lines of code to requires less memory.

[ianp5a]

When I started we only had one byte of memory to work with. Don't laugh you could do a lot in one byte. These days you are spoilt. And even a megabyte looks small. But back then we were glad to get a whole byte. The days of millibytes were finally over.
And I remember when we got our first USB thumb drive. It was 'this' big. (holds arms out wide)

My first digital camera, that was a marvel of modern science. Of course it was nothing like today's slick devices. Mine didn't have a Megapixel. It had just one pixel! It's true! But you could do a lot with one pixel. You had to use your imagination a lot back in those days of course. And the SD card was about the size of a small carry-on suitcase.

[TWPonKubuntu]

Wow, You guys are OLD!
wait...
So am I, I just forget sometimes. Too many core dumps?

[Snowhog]

A good core dump is beneficial, even desireable. Firm and well organized. It's those loose core dumps you have to keep your eye on. Running on and on of meaningless data and no solid organization. Yeah, watch out for those.

Oh, wait. You're talking about computer (core) dumps? Never mind!

[TWPonKubuntu]

Exactly! There is reason a good core dump produces plenty of neatly folded paper.

Plus it gives you something to read while seated.

The experienced programmer has learned to call for a core dump about 15 minutes before each "rest" break.

[GreyGeek]

Millibytes? Luxury!

When I started programming we had to use our eight fingers to keep track of the value of one byte. Holding some fingers up and others curled down was hard to do!

[GreyGeek]

I remember such things, too, starting college 1967, doing FORTRAN IV, on a mainframe IBM 360?, punched cards, etc. What I remember is that we were taught to write FORTRAN programs in such a way that they used minimal memory! as "minimal" as possible, that is ... how to re-write lines of code to requires less memory.

The Dept of Physics, which was offering the Numerical Analysis course that involved Fortran IV programming, had a contract with a bank downtown to reserve 1,500 bytes of core to hold and run our programs, which were sent to them on the role of punched tape. We got back either a greenbar printout with the answer, or a program listing pointing to the error. Edit the code? We wished. We had to type the whole program in again to create another role of punched tape to send down to the bank. Some folks were able to complete only the first program, the solution to a quadratic equation using Fortran.

With about a month left in the semester the department got a Friden EC-132 Digital calculator.. It had an RPN type of input (much like the HP hand calculators) that allowed us to do the quadratic equation program in just a minute or so. That device was in such demand by the faculty that the only time we grad students could get access to it was after midnight on weekdays or six pm on weekends.

That was in 1968. I never took up programming again until 1974, IIRC, when I bought the EC-1 analog computer from Heathkit to use in my physics classes. It was great at illustrating solutions to integration problems. Then, in the summer of 1978 I bought the first Apple ][+ sold in the state of Nebraska and I re-learned to program using Apple BASIC and then Apple UCSD Pascal, which I loved. When I resigned in 1980 and began my computer consulting business I was using the Borland Turbo Pascal 3.02A with their database "Lunchbox", which was more powerful and a LOT faster than the UCSD Pascal. From there I learned and used probably a dozen more languages. My all time favorite was/is Forth, from this company. There is a link on that webpage leading to a tutorial about Forth, written by Leo Brodie, called "Starting Forth". It is the best, and funniest programming book I've ever read. I used it to write turn-key systems for autoparts stores, amortization tables for banks, GAAP inventory & POS systems for camera shops, fastener retailers, and applications I've forgotten about. The repository has pforth, gforth and yforth, but they are somewhat rudimentary. They are good enough, however, to be used to follow along with "Starting Forth".

[Snowhog]

If you only do it for yourself, give serious consideration to writing your memoirs. Simply from what you've divulged to us here, your life has been an interesting one!