[John Graham-Cumming] might not be the first person to thumb through an old book and find an IBM punched card inside. But he might be the first to actually track down the origin of the cards. Admittedly, there were clues. The book was a Portuguese book about computers from the 1970s. The cards also had a custom logo on them that belonged to a computer school at the time.
It is hard to remember, but there was a time when cards reigned supreme. Sometimes called Hollerith cards after Herman Hollerith, who introduced the cards to data processing, these cards used square holes to encode information. Reading a card is simple. There are 80 columns on a classic card. If a column has a single punch over a number, then that’s what that column represents. So if you had a card with a punch over the “1” followed by a punched out “5” in the next column and a “0” in the column after that, you were looking at 150. No punches, of course, was a space.
So, how did you get characters? The two blank regions above the numbers are the X and Y zones (or, sometimes, the 11 and 12 zones). The “0” row was also sometimes used as a zone punch. To interpret a column, you needed to know if you expected numbers or letters. An 11-punch with a digit indicated a negative number if you were expecting a number. But it could also mean a particular letter of the alphabet combined with one or more punches in the same column.
So where did [John’s] cards come from? Since he found the school that used the cards, he was able to locate a text book also used by the school. Inside, there were illustrations of exactly the real cards found in the book. They were exercises for students, and the second book detailed what all the fields actually meant. Mystery solved.
It was common in the old days for cards, tapes, and even disks to store data, and it was up to you to know what kind of data it was. That’s why FORTRAN had such an intricate FORMAT statement for I/O. We occasionally see new card readers. They rarely, though, work with standard cards.
Ha! I have seen only unpunched cards, and for some reason always presumed that they were used to represent binary code, 10 bits per byte/word. “Big numbers they knew back then” I thought… Luckily wikipedia gives some consolation that there were also binary formats in use, only arranged in a different way. 6*6 bits per word, 2 words per row.
Thumbs up for this one not being a YT video.
I remember there usually being a punched card in the box when I was bought shoes as a child. I guess it was early stock control.
Remember getting one with the electric bill.
Do not fold or spindle. Heh…had a teacher who would soak the card to slightly change it’s dimensions and trigger a manual entry. As a grade 5 student it made me giggle at the prospect of all customers doing the same thing. 😉
In High School, I took the last class teaching “programming” of the IBM 402 Accounting Machine. You programmed it with wires pushed into holes in a large (11″ x 17″ or so) metal-framed bakelite panel.
You got down and dirty with each column of the punch card and the punches in it. All relay logic, timing driven by the rotation of a motor driven card feed mechanism. The printer worked off the same motor and timing, and used typebars, one for each print column, and raised to the appropriate character, then impacted by a hammer to print it through a fabric ribbon.
It was large, grey, and very noisy. Pix here http://www.columbia.edu/cu/computinghistory/402.html
As I am old, I grew up with punched cards. Stores used them, turnpike tickets were punched, the DVM used them, and so on. My first work on large systems was a 360, interaction being by standing at a window and passing your deck to the white coated operator assistant (Acolyte). They would rest in a place of honor on a heavy-duty storage rack (the Nave) until the card reader (Confessional) was free, at which time the cards would be scanned to the job queue tape. Eventually, the Operator (Priest) would rewind the tape (anoint) and present it to the system (Alter). A couple hours later, in one of the bins outside the operations room (Cathederal) a stack of fanfold would appear with the card deck on top.
The worst were the routine job guys who had only half a dozen cards (account, JCL to start the job, and billing, IIRC), with a paperclip, in their sweaty hands. The same cards, day after day. They would routinely jam the reader.
The decks with several gross cards in boxes never did
Actual IBM card so probably running on their leased mainframe. Long story there. Special characters were 3 holes in a column, as in 0-3-8 IIRC was a left parenthesis in EBCDIC, ASCII had different codes. 401 accounting machines only interpreted EBCDIC. Forget what 038 interpreted as but could get confusing occasionally
Punched cards were handy for shopping lists, bookmarks, telephone notes, pretty much anything besides being used to transfer data to a computer!
They were (allegedly) sized to the dimensions of a dollar bill when Hollerith created them (the dollar has shrunk, physically as well as financially since that time). They fit neatly in a shirt pocket (behind the pocket protector) and wallet (when folded in half).
IBM also had the 96 column card – roughly 1/3 the length of a 80 column card. Had 3 rows of data – same encoding as the 80, just smaller holes and three sets horizontally across the card.
These were used on the system/3, which was manufactured in Rochester, Minnesota.
When I left university in 1974, I had a room’s worth of long boxes containing programs written on cards. I eventually threw them away when the first home computers arrived and stored them.on magnetic tape. The same programs would fit in a USB key and hardly dent the memory but I sure remember getting in line to run my programs with other students. It was at University of Waterloo where Fortran Watfor and Watfiv were being developed. Good times!
Ah, memories of the hall full of card punch machines, with students toiling away in the wee hours. The unique sound. The characteristic smell…
Not good memories, mind, but memories nonetheless.
But those cards are so darned useful for so many other things. Shopping lists, bookmarks, shims… 3×5 index cards just aren’t right. My stash of them depleted in recent years, I tried to find a new box of cards. I assumed no more made, but expected NOS to be plentiful, as there was just so many of them around in the day. But nope. None to be had, except in ‘artisanal’ quantities at ludicrous prices. I guess they were good stock to simply recycle, so there are none remaining…
Sounds like a premise for a dystopian novel. After a monster solar flair wipes out digital memory and magnetic tape, mankind’s reboot thwarted by lack of new cards and paper tape to get the damn machines going..
They were used to make Christmas wreaths, of course.
I remember hearing a story about a company in the US needed to ship a punch card program deck some place in France. The France company was having a problem with their machine and needed a special diagnostic program for it. No problem, the US company punches the cards for this special program and ships it out. The shipment arrives, and the recipient loads and runs the program, and the program kacks out at a certain point. The US side finally has the deck double checked and they discover the program is missing a random card out of the middle. Weird and very annoying, they had tested it before shipment and it worked. Ok, they ship another card deck out. A couple weeks later the company gets it, loads it, and amazingly the program breaks again, but at a different spot. After troubleshooting they confirm that yes, another card is missing out of the deck. Very frustrating, but they ship ANOTHER copy of the program deck out, and have the recipient verify the card deck before they even run it. Welp, once again another random card is missing out of the middle of the deck. Ok, so there is some sabotage happening along the way. This time, the US company sends a technician hand-carrying the card deck out to the customer site. When the tech enters France, he has to declare this program deck. During the encounter the France customs inspector looks at the deck. pulls a card out of the middle and stuffs it in his pocket. That’s when all parties involved discover that France customs inspectors are allowed to take a small sample of any imports for personal use. In this case the inspector believed this to be a stack of blank paper and was taking cards out of any punch card decks he encountered and was using them for personal note cards.
I heard the same funny story with minor changes from my uncle who was graduating in computer science in the early 60. My father who was at that time a tech expert in IT hardware also told me a slightly different story.
Looks like this story was common at that time among ITProfessionals…
BTW custom inspectors may take samples for analysis or expert appraisal purposes only.
At that time the same uncle was traveling to Switzerland on a regular basis to test programs on a more powerfull system at the UPFL and once said to the custom inspector who was checking the punch card deck that “the more information on it, the less it weights”…
The customs officer look was priceless !
That’s eerily similar to a story Bob Cringley told about low yields with Intel wafers. They eventually send a detective along with the box of wafers, they eventually get to receiving at Intel, the guy pops open the box, riffles down the wafers counting them, then seals up the box and sends it on its way
Why didn’t they just pick up the telephone and tell the French people what the middle card should be punched and let them do it and insert it instead of sending an entire deck over and over? The story sounds like a lie. Punch cards by definition are not blank, even when unpunched. Stealing cardboard? No, the story is a lie.
The description in the article puzzles me. One digit per column? That can’t be right.
Grew up with my dad working in Fortran by writing out his program by hand. A secretary would type it in, producing a stack of cards with each card being a line of source code. Any mistypes gave a syntax error. A surgical scalpel was the tool to fix some of those
I remember my dad telling me that they had a machine where he worked that they (usually interns) could use to manually punch cards. He described it as something like a printing press, only they’d set pins instead of letters. When they lost power, which was apparently a regular occurrence, they’d pull the machine out and start punching cards. They’d also copy from greenbar to punchcard as a form of backup. When they officially moved to magnetic tape, he got to bring home boxes of cards. He turned them to mache and formed fire logs. They kept us warm for a couple of weeks.
I remember using these cards long time ago. The line you typed in punches was printed in ascii characters at the bottom. No need to decode the punches as described in the article. But my Russian friends were used to cards with a bad quality of ink. That was wiped out in a day. So by visually scanning the card, they were able to decypher the holes in seconds.
There was also 96 column cards.
https://obsoletemedia.org/96-column-punched-card/
Those were common on System 3 (I used a System 3 Model 12 a LONG time ago). So much so that the 8 inch floppies we used as “cards” thought they were 96-column cards. Note that the card actually had 32 columns that were split into 3 sub-columns. That is column 1 was above column 33 which was above column 65 (unless I counted wrong).
When the holiday season rolled around, the keypunch department was always decorated with wreaths made of punched cards. Spray painted red and green and festooned with glitter, they added a delightfully festive touch.
Punched card chad was also a treat. Chads were the bits pasteboard punched out of cards by keypunch machines or the card punch machines. It collected in boxes inside the machines, the boxes needed to be emptied frequently. We used different colored cards for various applications, and so the chads were quite colorful. One of my co-workers dumped some into another co-workers car on his wedding day as part of the traditional drunken decoration fest. After that, every time he turned on the AC, chads would blow out.
Can’t say I miss them. Back then when I was in college I was glad they let us use the terminals of our PDP-8 since the card reader was out of order.
When I was a student our Fortran classroom assignments were submitted as punched cards. The TA handed out punched cards to each student to include those cards in the deck to route the finished program listing and output to the TA. TAs had access to terminals. The few terminals available to students were occupied 24-7.
If you were well versed, data collected in chemistry labs could be punched onto cards, and bundled with program cards to do the necessary calculations and graphs. Very helpful in analytical, instrumental and biochemistry enzymology labs. Personal hand-held calculators were expensive.
Some of my fellow graduate students wrote their dissertations on punched cards. After editing they ran the final deck and routed the output to a networked IBM Selectric. I paid the department secretary to type my handwritten dissertation on a Selectric. I was a terrible typist. My handwriting involved a lot of actual cutting and pasting and tape – the sticky kind.
I was lucky. When I started as a freshman at college, The seniors were the last to used the punched cards. Still had to learn Fortran though… Pascal was the ‘learning’ language. Growing my dad would bring home boxes of cards that we ‘built’ things with using tape and staplers for fun.
We used a very wide green marker to mark a diagonal across the top of the larger decks. It gave us a clue in the event that the deck was dropped.
I got my first real experience in typing on IBM 026 and 029 keypunch machines. Punching out programs in ALGOL for a Bendix G21 and then a Univac 1108.
Eventually I graduated to a 360/67, and we had to use punched cards for modifications to the operating system ( TSS/360). Our installation had lots of hard disks (and eventually large mounts of memory – several M of core), but no drums, so the IBM distribution needed to be tweaked.
I don’t think my wife let me keep any of the card decks I still had when we got married, 40 years ago.
Used punch cards from 1965 to 1978. IBM 1620, CDC 3200, 3150, 6400, IBM 360/40, 360/44, Burroughs 3500, 4700, IBM 360/20. Helped build tools (apps) to enable programmers to used no sync terminals for for programming. Retired two years ago. Now working for a startup that will change the the industry. Most fun that I have had since my first computer programming job in grad school back in 1970
I wrote my first programs (Fortran) on punched cards. Then worked my way through college (1973-1977) as an operator on our college main Frame (a Burroughs B6700, I was very lucky, as that was a very interesting machine). I was the guy you handed your cards to and who handed you back your output. We also had a card punch as an output machine. It occasionally caught fire, we kept an extinguisher nearby. I think I still have a few cards tucked away along with a 9 track mag tape and some 132 character width green and white striped paper.
The great part of my job was back when computer time was doled out in small increments to undergrads, I had all the computer time I wanted. Admittedly, much of playing Star Trek, but still…