« Technology Transfer: A Contact Sport | Main | HPC and the Excluded Middle »

October 11, 2010


Feed You can follow this conversation by subscribing to the comment feed for this post.

Ole Phat Stu

I have programmed the Mercury, which, as its name suggests, used acoustic echoes in a mercury delay line as a serial memory!

And the Atlas in the basement at London University (UK) which paged from tape; GODDAMMIT!!! ;-)

BTW, I'm 66 now...


geeze I started in 1982 being 6 years old typing in basic on a 64K Tandy TRS-80


good old days operating the TFR BBS in the 403/604 back in 1988+

Whoo hoo!!


Gosh darn, well I 'member when you lost a bit if the cathode burned out.


Good grief! I *thought* I was a computing old timer, but compared to the rest of you, ha!, I'm not.

See I started using UNIX, actually Tandy Xenix, in 1985 as an end-user doing data entry while in college; I was primarily using a Decision Data 9610 keypunch. The Xenix system had the most rudimentary database for entering a few fields of info; the real processing took place on the IBM System/3 mainframes. It was the company's desire to move completely away from the IBMs, but there was no one there who could accomplish that.

In 1987 I taught myself programming and system administration on that Xenix system, and then took on the task of transitioning the company, a furniture store chain, from the IBM System/3s to a multi-user, multi-location *nix system. In the process I got to spend a lot of money, ~$25,000, on equipment--very fun when it's not YOUR money. :)

I replaced the Tandy 6000 with a custom built ALR, and installed SCO Xenix on it, along with filePro for database purposes. Added many Wyse 60 and 150 terminals, local and system printers, and modems. I had a dedicated phone line installed at home so that I could access the server 24/7 with my Wyse terminal.

As I worked on each segment of the programming, i.e., inventory, accounts payable, etc., we'd run parallel with the mainframes for a month or so to make sure there were no bugs, and then we'd go live. It was very cool.

I think it's worth noting that I was the only female I knew, or knew of, in this field at the time. Again, very cool. :)


1. ASR-33 – Owned one! Ex-wife probably still’s got it but I got the wok!
2. S/360 – 8 bits! – Try 24-bits if you were serious in ’64 (ICL 64Kb 1902 )
3. Ate cake off them when there was a birthday in lieu of plates at Dalgety’s
4. Had to do that to run diags on an ICL ME29 or dial-it-up to I.P.L. the 1900’s
5. Ahh! The old Hand-operated keypunch manufactured by British ICT (1960s) – I think my old boss might still have one?
6. Had one!
7. 1922! Berkeley for me - Who’d have thought it’s still earning some a living ;-)
8. OK – Got me there but 2900, S39, 1900, FDS-160
9. ICL! ICL!ICL! Oi! Oi! Oi!
10. Nothing to this!

Paul T. Lambert

I still have some ancient petaflop boxes lying around, along with an archaic 64K-core tablet from the days of yore when nanolithography was all the rage.


Guess I'm one 'o them young whippersnappers. All I have is memories of 6600 assembly, (dual) vax 780s and an old (defunct) 6600 core module :)

Joe Bottomlee

I worked on a computer for over 12 years before I knew it was a computer.

You see I was a Central Office Tech. for SW Bell in 1969. The telephone switching equipment was being changes over from an electro-mechanical system (SxS) to an electronic switching system (ESS). Twelve years later (1981), when I bought my first personal computer (TRS-80) is when, I discovered that I had been working on a computer all those years.

The software and data was stored on memory cards stored in modules called Program Store (PS). The memory cards were sheets of aluminum about 18 inches by 12 inches and had 64 by 16 bits on each card. The memory bits where little blocks of magnetic material that could only be change by over writing using a Memory Card Writer (MCW). The cards were physically remove from the PS and taken to the MCW. The RAM was stored in Call Store (CS) they were 8KB each and our switch had six of them in redundant mode, which mean the whole system was running on 24KB of RAM.

Joseph Senecal

I say you are a computer old timer if you worked on a computer that
1) had bits large enough to see with the naked eye (magnetic core)
2) Didn't have a boot ROM
3) Loaded programs from punched cards

I got started with the IBM 1130, a desk sized computer with 32k bytes, a 1 MB 14" hard drive, card reader and line printer.

Erik Dalhammar

"To me, you are a computer old timer if you worked on a computer that -
1) Used punched cards or paper tape for input media or
2) Had a real front panel (allowed you to input instructions and data via switches, view memory on lights and allowed you to start, stop and single-step the machine) or
3) Had 64 KB or less of magnetic core memory "

Then, I am a computer old timer:

I wrote my first professional real time programs 1967 in assembler on a SDS910.

It was equipped with:
1) 6 K 24 bit words = 18 K bytes core memory
2) Binary front panel
3) Paper tape reader and punch
4) One mag tape station
5) One 1000 lines/min drum line printer

The computer was used to control a high speed OCR document reader from Recognition Equipment Inc.
The system was located in Stockholm, Sweden, processing payment documents.

Program development was done on paper tape using a program called PseudoFlex and then running the two-pass assembler on the computer.
Operating system, application program and data all had to fit into the 6 K memory.


Phew, not an old timer yet, though VAX 780 did make me worry.
Only ever used punch cards to
1) smooth burrs off dental tools used under microscope when making thick film circuit prototypes (1985)
2) to build large sculptures in uni computer graphics lab (1990)

CPM is the oldest O/S I used. My first program was on a pet and my second in 6800.
And I only know the tune to a 1k modem (which is susceptible to neighbouring noise)
A few years in me yet ;-)

Louis Hembree

The CDC 6400 with right and left memory cores and drum drives. Punched tape, binary card decks, dropping the binary decks, developing mag tape, cleaning mag tapes, you know what negative zero is, punching output data to cards, 300MB disk drives the size of a washing machine, requesting that a disk pack be mounted. Univac’s which used ones complement in the input reader and twos complement in the system, or was it the other way around.

Patrick Mallory

I started with an IBM 1401 that had 16 K of core memory. We were programming with COBOL and the last instruction of the load deck was to overwrite the load space to give a little working storage. Source decks were precompiled into and intermediate (Autocoder like) intermediate deck that was compiled into the runtime machine code deck. In emergencies the code could be corrected in any one of these stages and you had to remember to go back and correct and recompile the source deck when there was time. Program to program communication relied on the 4 tape drives. remember blocking factors? When the airconditioner (the other half of the room) failed we had to shut down the whole system quickly.
Some years later I labored over a decision on whether to buy the PS2 with 30 or 30 Meg harddrive.
Appreciate the progress but now I know less (proportionately) about computers that I did forty years ago.


ok. i see: talking to kids about atari or commodore is a nothing compared to these here. :)

Chris Shaw

I first learned to program in 1956 on a Jet Propulsion Lab Burroughs machine that lacked index registers. To write a loop that summed a list of numbers, you'd clear the word holding the sum, set the addess in the add instruction to the word containing the first number in the list, execute the add instruction, add one to the address in the add instruction, test that address to see if it was greater than the address of the last number in the list, and branch back to the add instruction if it wasn't. The machine's words were stored in mercury delay-line memory. I'm still employed as a programmer.

Robert Reed

You might be an old-timer if...
You have memories of cutting and pasting code as an activity involving graph paper and scissors.

Charles R. Patton

I remember writing programs in GoTran, the predecessor to Fortran.

Yes, ASR-33s, but also used its equivalent, the "Flex-O-Writer"

Using a star-wheel paper tape reader (and wanting to cut it into little pieces for its unreliability.)
Later using a Ferranti 600 cps, stop on character, paper tape reader. What a wonderful piece of iron that would shake the equipment cabinet it was housed in. The pancake motor driven reels were capable of inflicting nasty wounds (no OSHA guards on that beast.)

Boxes full of paper tape chad, used to fill my business partner's brother's car as a wedding prank. Years later he said he was still finding pieces of chad!

Sid Schipper

I started programming in 1971 CDC Assembler Language on a CDC 6600. 60-bit words! I loved it. I moved over to the IBM world in 1975, BAL on early MVS systems. I used higher level languages like FORTRAN and PL/I also, but I was a die-hard assembler language freak, and to this day I still think IBM/360 Assembler Language is the simplest and easiest computer language ever invented. I've since programmed in just about every computer language there is from PASCAL to JAVA and I still long for those IBM assembler language days. It seemed the computer world was a lot simpler back then, nobody seemed concerned about stealing code or proprietary information, everybody freely shared whatever good code they had. Maybe I'm a sentimental old dinosaur but today's world seems so much more cutthroat, I hate to say it, but I long for the day when I can retire with my memories.

Keary Keller

I feel terribly young in that I started on an IBM Series/1 with it's teletype console interface, and a hard drive that would squeal because in 1991, you couldn't buy spares anymore. A good swift kick usually unfroze it.


1. My hand span let me type on ASR-33 with one hand. (The other was pointing at the code I had prewritten since we were only allowed 15 minutes of terminal time at a time).
2. Ditto on both items.
3. A bigger problem was dropping the deck, which is why you’d put sequence numbers in the normally unused columns past 72.
4. The PDP-8 required toggling in of the boot program. See http://www.rostenbach.com/pdp8i/pdp8i.htm the box labeled “RIM Loader”.
5. The 129 was the bees knees, no wasted cards, it buffered the input.
6. Used to record them on my cassette recorder to play them back at a science fair I entered with a project about artificial intelligence. I programmed the computer to play Battleship against you.
7. He forgot TSS-8 on the PDP-8.
8. 1401 (the line printer)
9. Also known as Snow White (IBM) and the Seven Dwarves (add Sperry and Hewlett-Packard).
10. Haven’t made it there, but bought some core memory and a silicon wafer from them. Also pestered them for information on the “Mona Lisa by Numbers” that I’ve been helping an art restorer in Italy work on.

I deny all of the above.

Larry Coates

Anyone remember building J-K Flip-flops using a pair of 6SN7 vaccum tubes, or the advanced version using 2n222a switching transistors?

You might be an old-timer if you remember when LSI chips with 4,000 transistors seemed impossible.


If somebody had a card deck that played anchors away on the 1403 line printer.


If you booted a PDP-8I with paper tape and programmed it with assembler and Focal.


If you built your own modem for Internet access with spare parts like a large black handset and POTS equipment.


Hmmm... All but one!

The comments to this entry are closed.

Dan Reed
Dan Reed

December 2023

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30

Twitter Updates

    follow me on Twitter
    AddThis Social Bookmark Button


    • Add to Technorati Favorites