4.5Megabytes of data in 1955

62,500 punch cards

. <- 4.5 megabytes in 2019 [not to scale] :slight_smile:

I sure hope no one tipped that pile over :stuck_out_tongue:

And one bug in the program and the whole proces of punching and running overnight had been for nothing.
Programmers learned how to design and develop instead of trial and error using Cmd-R.

While I was at CERN, there was one programmer whose Fortran program was on 24,000 cards. That took 12 trays of 2000 cards each. Where I was, we had one of those CDC card readers that could read 1200 cards a minute - so it took him 20 minutes just to read the job into the computer.

I did try to persuade him that stuff he wasn’t changing could be compiled and the binary punched out, making many fewer cards, but he wasn’t having any. The computer also had four washing machine sized disk drives (removable packs of 50Mbytes each) that he could have stored the source on and done the editing at a terminal (rather than a card punch) but he wasn’t having that either.

Times change, eh?

Some people in 2019 does not even know what is that :frowning:

In the early days of REALbasic, the “R” key on my keyboard was totally worn, just because of Cmd+R. :slight_smile:

@Emile Schwarz — Dans les années 90 j’ai aidé mes parents à perforer des cartes pour le centre de calcul de Strasbourg :slight_smile: Je trouvais ça fun à l’époque, LOL

When the kids were young we took them to the technical museum in Munich and I pointed out some of the machines I’d used including punch cards of course.

On the Monday my daughter explained to her class: Daddy is so old the machines he used are in the museum!

@James Dooley — Just likeToutankhamon bones? LOL

Those card readers were very finicky about the cards they were reading, and jams were very common. In my final year at university, they replaced the old IBM card reader monstrosities with Documation card readers which were the size of a desktop printer, about five times faster and way more resistant to jams. They used an air stream to separate the cards in the feed hopper, and then a perforated belt with differential air pressure on opposite sides to move the cards. It was a miracle of simplicity, and so jam proof that we would deliberately mutilate cards to see if they would still go through. You could fold the cards into quarters then unfold them, and the reader would still read them. Then about a year later, punched cards were pretty much gone from the university (and everywhere else). That seems to be the way it goes. By the time the technology is perfected, it’s obsolete.

I found this video of a Documation card reader in operation: