8

There is a question here on retrocomputing if round punch card holes where mechanically stiffer:

Were round punchcard holes mechanically stiffer?

I wondered if the reason for the rectangular holes on IBM's 80-line punch card was not the readability of the punch cards themselves, but the durability of the punching tools used to make the holes.

A punching tool that should create rectangular holes only has to be sharpened on one side, like a chisel, and then it is sharp again. If you use four of these that are of triangular shape and combine them to a forming rectangle, you can create a nice, sharp, rectangular hole. And these four stamps could be sharpened very often, the longer the stamp tool, the more often that would be possible.

In the case of a round stamping die, the stamped piece is likely to be cut off first during manufacture, so that it is flat. In the next step, it gets an inner bore so that a hollow is created. The outsides thus become sharp for clean cuts, but these outer cutting edge wear out more quickly. Such a punch can then no longer be repaired with simple methods, but must be exchanged for a new one.

Is that perhaps the main reason why IBM used rectangular holes so that the stamping tools can be reground more often and last longer? A more detailed description of what the actual stamps looked like would be very revealing here. Macro Photos of the stamping tools would be super.

EDIT: In this video from university of Nottingham proof is given, that a lot of punching did wear out the punch tools: https://youtu.be/JafQYA7vV6s?t=114

15
  • I understand that the rationale for rectangular holes is closer packing, i.e., more data per square inch. As in this IBM history.
    – dave
    Aug 20, 2023 at 1:16
  • I doubt it, because if you make smaller round holes, you get the same density. And if you stagger the rows, you can increase the storage density with holes much more than with rectangles.
    – Coder
    Aug 20, 2023 at 1:20
  • 1
    @Coder staggering wasn't a solution. Going from 40 to 80 holes per row was already increasing the character rate more than mechanical devices could handle. Remember the punch card was defined and implemented at a time when there was no fast (not even slow) electronics. In addition staggering would have meant to double the needed amount of mechanical contacts wich in turn needed to be operated in an alternating pattern. A rather unwelcome complication. Not to mention that all tools (key punch) would have to operate on alternating height when creating/editing.
    – Raffzahn
    Aug 20, 2023 at 1:36
  • 1
    Where is the proof in that video? That one peroson long ago (before they started to punch codes on paper tape) said that this is what will happen is not a proof. A proof is if someone can give a figure of how often they had to change/sharpen the punches.
    – UncleBod
    Aug 20, 2023 at 15:34
  • 1
    @Coder No, the video says that "He said 'Look, if we are going to be punching holes coresponding to characters ...' ". He was (right or wrong) making an assumption from what he knew, but the video says nothing about anyone proving that the wear on the punches would be, or ever was, a problem. BTW, punched paper (and plymere) tapes were used paralell to punchcards, so the division was only that some media used round holes and some media used rectangual holes. (Might very well have been obscure uses of other hole types also).
    – UncleBod
    Aug 21, 2023 at 7:18

1 Answer 1

9

TL;DR: NO.

For one,

  • they are not cookie cutters but punches,

second

  • they are hardened steel and very durable and

third, if they degrade,

  • they will be simply exchanged - which includes both parts, punch and plate.

IBM punchcard holes are rectangular to keep competition at bay - as already described in answers to the cited question. 'Inventing' rectangular holes made the new form patentable (*1).

And no, there is no need for sharpening. I own keypunches that have been used for more than 30 years, stored for many decades and punch like on the first day. After all, why would they go dull? It's hardened steel against paper.

Also, hardened steel cannot be easily sharpened because it is quite brittle. It might also need to be rehardened, not a process in useful relation to the worth of that piece. They are throw away spare parts.


I would think the misconception here comes from the way you may assume them to work. Your description sounds much like a cookie cutter, a shaped blade pressed against a flat surface. But that's not the way (most) card punches work: They punch - that is a tool, called a punch, is pressed through the paper into a hole of the same shape and dimensions (*2). It works exactly like the good old paper punch we all have used uncountable times. Here a schematic from Wikipedia:

enter image description here

Do you remember having any of those ever sharpened? I don't. Cheap ones may go wobbly in their mechanics and thus lose function after a few decades of use, but not much more. I used one for many years which had been in that company since the early 1950s. Also, even the better ones have their cutting punches usually made from tempered steel, so less hard.

The principle used is the same as with a pair of scissors (when did you last sharpen those? (*3)) and called shearing.

Same goes for key punches as mentioned - but also for hand punches. Some I own date back to the 1940s and have been in good use at least until the late 1970s.


*1 - Back in those times patent offices did really have their own engineers check application for having a minimum invention level.

*2 - That is a tiny bit bigger to avoid getting stuck.

*3 - Not talking about you grandma's iron ones but modern hardened steel.

15
  • 2
    @Coder Oh, and then there the key punch - which you seem to see as a lesser stressed device. It isn't. Those typewriter like machines. Programmers did no use them. Programmers wrote on paper and let it be typed. That's data entry - you may have seen those pictures of halls filled with those machines and lots of young ladies typing. This is how all entry was done. Speed was between 10 and 20 cards a minute - yes, those gals were fast (also depends on data type). Programmers never typed themself. Bottom line, in real life those key punches were at least as much stressed as any gang punch.
    – Raffzahn
    Aug 20, 2023 at 2:11
  • 4
    @Raffzahn - this programmer (as a university student) punched his own cards on an 029.
    – dave
    Aug 20, 2023 at 2:34
  • 1
    @Raffzahn, "It's hardened steel against paper." - believe it or not, paper is a rather abrasive material. A personal hole-punch used for a few sheets a week may be serviceable for decades, but the cutting will degrade until fewer and fewer sheets can be accepted at once, more force must be applied (rapidly wearing the bearings of moving parts other than the cutting faces), and the cut becomes rougher.
    – Steve
    Aug 20, 2023 at 9:49
  • 1
    Minor point: While I agree no one is going to try to sharpen a card punch (among other issues, they are way too small and fiddly to bother sharpening), it is normal to sharpen hardened steel tools. If you use appropriate stones you can easily put a new edge on a hardened steel cutting tool. Aug 21, 2023 at 12:45
  • 1
    One thing not mentioned is that those things were built like tanks. If you stuck two treads and some motors on them you could drive them into battle and they'd survive. I'm talking of the ones I have experience with: 026 and 029. I never saw one that needed maintenance. Even the user-removable parts - like the cylinder you wrapped a template card around - couldn't possibly be broken by a squad of cocaine snorting orangutans. Maybe the reader/punches that were machine peripherals were less robust. The the ones you typed on couldn't be broken, didn't wear out - just kept on ticking.
    – davidbak
    Feb 14 at 21:23

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .