7 ms·
Whoa, that was delightful to watch, especially after learning about the diagonal stripes from your comment. But do the diagonal stripes actually work? If you d
by q7xvh97o2pDhNrh 3y ago
Whoa, that was delightful to watch, especially after learning about the diagonal stripes from your comment.
But do the diagonal stripes actually work? If you dropped the deck and then re-assembled it with a single pair of cards swapped, would you really be able to catch that based on the stripe?
Listening to the machine work was incredibly interesting, too. I wonder how much stress it puts on the cards to be run through the machine each time. Did they have to make "fresh" copies once the cards got worn down by going through the machine too many times?
- jfoutz 3y agoCards are from before my time. The blank cards I've seen had printing on one side, and usually some sort of title area. so you could get orientation correct. The punched cards I've seen were all hand numbered. The stripe, I think, was mostly a quick check. if it looks like a line, they're probably in order. if there's a spot that looks weird, you can check that range quickly. I don't think the failure mode was like someone grabbing your deck and throwing it in the air, it was more they got dropped, stayed mostly stacked, but you could quickly file the strays in the right order. and be pretty confident the cards that stayed stacked were right. if you've got a blurry spot of the line, you can quickly check that section. so, yes? I think the stripes worked? Again, before my time, this is lore handed down from professors. I think if you secretly swapped 2 cards, and they didn't notice, they would hunt you down and beat you up. IIRC, the cards did wear down, but there were duplicators, that would read old cards and punch fresh ones. Obviously, if something was torn or damaged, you'd have to fix that by hand. I think I've seen one once, but I don't recall exactly where. edit- https://en.wikipedia.org/wiki/IBM_519#514 https://en.wikipedia.org/wiki/IBM_519#514 <- that's not what I remember, but they did exist.
- Animats 3y ago> I wonder how much stress it puts on the cards to be run through the machine each time. Less than almost every other card reader ever made, which is why that Documation unit took over the industry. The hard problem is pulling exactly one card off the bottom of the stack. Most previous readers involved some kind of pusher which had a ledge slightly less than one card thick. The pusher pushed the card through a gate slightly more than one card thick. The IBM 1402 card reader, and all the earlier IBM electromechanical card machines, used that technology.[1] Pushing cards through a gate works only if cards are within tolerance and in good condition. Otherwise, card jams are frequent. Card data processing operations had dehumidifiers, because high humidity would make the cards thicker. Documation made a breakthrough. There's a crosswise air jet that separates the bottom few cards in the stack before picking. The video shows that. The picking device itself is a vacuum picker. It grabs the bottom card, rotates, and starts the card into the rollers. The card is always pulled by a face, never pushed by an edge. Card thickness does not matter much. It just works. That air jet trick is still used in some printers and printing presses. It's one of those clever tricks of mechanical engineering that yields a simple, reliable device. [1] https://en.wikipedia.org/wiki/IBM_1402 https://en.wikipedia.org/wiki/IBM_1402
- sbierwagen 3y agoAs the counterpart to printing presses, some automated book scanners also work that way-- puffer to separate the pages, vacuum paddle to flip them: https://www.youtube.com/watch?v=GmGntnfohrA https://www.youtube.com/watch?v=GmGntnfohrA