×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Resurrect Your Old Code With a DIY Punch Card Reader

timothy posted about a year and a half ago | from the convert-your-existing-recipe-cards dept.

Data Storage 100

First time accepted submitter mchnz writes "Need to read in some old punch cards? Have a hankering to return to yesteryear? I've combined an Arduino, the CHDK enhanced firmware for Canon cameras, and the Python Image Library to build a reader for standard IBM 80 column punch cards. You can see it in action in "Punch Card Reader — The Movie" or read more about it." This is an inspiring, intimidating project.

cancel ×
This is a preview of your comment

No Comment Title Entered

Anonymous Coward 1 minute ago

No Comment Entered

100 comments

I'm lazy (3, Interesting)

jmorris42 (1458) | about a year and a half ago | (#40780583)

I'm too lazy for all that lego building. If I needed to read punch cards I'd just find a scanner that would accept media that narrow and feed em through the ADF, feed that PDF into a script to pop apart the pages and then process the images.

Re:I'm lazy (0)

Anonymous Coward | about a year and a half ago | (#40781883)

Pretty much this.

I mean this thing is neat.. but scanners that can auto-feed are easy to come by.. and most can handle small documents (though the thickness might be an issue.. mainly comes down to whether the thing can easilly bend around the drum without jamming..)..

I was kind of hoping for some pin-based contraption. I mean image based is obviously the better way to do it, but still..

Re:I'm lazy (1)

icebike (68054) | about a year and a half ago | (#40781945)

Call around. There are still places with punch card readers installed, even if they don't turn them on very often. Lots of universities etc.
State of Florida. ;-)
They will probably read them in and email you the file for less than the cost of that crank shaft thingie.

Re:I'm lazy (1)

uglyduckling (103926) | about a year and a half ago | (#40784305)

Document scanners with straight paths are readily available. I just scanned a couple of thousand prints from 35mm photos through my Fujitsu Scansnap as I realised that low res reference images of all my old photos are good enough for me and relatives to look through, with the option of a high res negative scan for the ones we really want. I think my Scansnap would do upwards of 20 cards per minute, maybe 30' far faster than this solution.

Re:I'm lazy (2)

narcc (412956) | about a year and a half ago | (#40784359)

I was kind of hoping for some pin-based contraption. I mean image based is obviously the better way to do it, but still..

Pin-based contraption? Easy there Hollerith, it's not 1890 any more! Put down that abacus, we're not dropping pins through holes in to cups of mercury these days.

Check out the IBM 2501 -- it's optical, sweet-cheeks -- You can't do >500 cards/minute with one of your antique mechanical reader! Get with the times!

Re:I'm lazy (0)

Anonymous Coward | about a year and a half ago | (#40781949)

But that way doesn't allow for the Arduino ad!

That's cheating! (4, Interesting)

gallondr00nk (868673) | about a year and a half ago | (#40780773)

I was expecting something that mimicked the original way these cards were read. Anyone can take a photo of a punchcard :)

Re:That's cheating! (4, Insightful)

Chris Mattern (191822) | about a year and a half ago | (#40781419)

Digital electronic overtook mechanical decades ago--that's why punch cards aren't used any more in the first place. It therefore follows the smartest way to deal with any problem of this sort is to get it converted to electronic ones and zeros as quickly and simply as possible, no matter how dirty, and then process the digital data to get what you need. In this case, that's getting a digital photo of the punchcard, and then doing your work on that.

Anyone can take a photo of a punchcard

Indeed they can (don't forget to use your wooden table!). Then doing OCR on that photo to extract the data the punch card contained is a little more involved, however.

Re:That's cheating! (1)

na1led (1030470) | about a year and a half ago | (#40782553)

I can't imagine converting Punchcards would be a problem to anyone today. Sounds more like a fetish, or nostalgia to me.

Re:That's cheating! (2)

FitForTheSun (2651243) | about a year and a half ago | (#40789197)

Only in rare circumstances. I recently heard a program about a guy who was trying to solve the Pioneer Anomaly, and he had to go find old boxes of punch cards and convert that data, in order to determine the spacecraft's exact positions when it was still in the Solar System. But yeah, pretty rare.

Re:That's cheating! (1)

iluvcapra (782887) | about a year and a half ago | (#40786099)

Digital electronic overtook mechanical decades ago--that's why punch cards aren't used any more in the first place.

Agreed, though he would have maxed his geek points if he'd implemented his machine emulator with Minecraft Redstones.

Re:That's cheating! (3, Insightful)

adisakp (705706) | about a year and a half ago | (#40782081)

It's a cool project but it'd be more cool if it didn't require manual intervention (turning a crank per card). How hard would it be to add a servo controlled motor to turn that crank so the entire process could be automated?

Re:That's cheating! (1)

mirix (1649853) | about a year and a half ago | (#40783429)

Came in here to post this. Maybe I have a new project to make... electromechanical punch and punchcard reader. :)

Re:That's cheating! (0)

Anonymous Coward | about a year and a half ago | (#40788895)

They were optically read already during the WWII, before any software for a general purpose computer were written. A photo of a punch card is essentially the same thing.

OK, show of hands ... (3, Interesting)

gstoddart (321705) | about a year and a half ago | (#40780835)

OK, this is really cool ... but how many people still have decks of punch cards?

The closest I've been to them is a box we had of them we used for notes.

Though, given the level of technology pack-rats we likely have on Slashdot, I expect several people to say they still have some cool program or another tied up neatly waiting for just such a thing. :-P

Re:OK, show of hands ... (1)

jeffb (2.718) (1189693) | about a year and a half ago | (#40780981)

Probably a few; I'll bet I could even find them. I've also got a few programs that I dumped to paper tape through an ASR33 when I was getting ready to graduate high school. I've thought about fishing those out, building a rig to pull them past my EX-FC100 while it's in 1000FPS mode, and seeing whether a paper-tape reader can outrun a 9-track tape drive...

Re:OK, show of hands ... (3, Interesting)

Lumpy (12016) | about a year and a half ago | (#40781745)

I service a CNC machine at a machine shop that STILL uses paper tape. I am the only guy that will touch the old hardware so I get paid $150 an hour to tinker with the stuff. Hell they cant find a IT company that can handle DOS so I also pick up the other old machines.

I love it, 1 long afternoon/evening there and I go home with $1000 in my pocket.

Re:OK, show of hands ... (2)

plover (150551) | about a year and a half ago | (#40789235)

I service a CNC machine at a machine shop that STILL uses paper tape. I am the only guy that will touch the old hardware so I get paid $150 an hour to tinker with the stuff. Hell they cant find a IT company that can handle DOS so I also pick up the other old machines.

I love it, 1 long afternoon/evening there and I go home with $1000 in my pocket.

This bothers me. You could easily replace the paper tape reader with a computer based solution - all the paper tape readers we had on our CNC machines 25+ years ago were interfaced with ordinary DB-25 RS-232 connections. Back then I just stuck a BlackBox short-haul modem on them to run the wires back to the foreman's office, where the other modem was plugged into an original IBM PC. A simple GW-BASIC serial program allowed them to load and store their programs to a hard disk, which was backed up nightly. It wasn't rocket surgery back then, and I have to believe the state of the art has advanced a little bit in offering hardware and software choices to do this today.

  A Raspberry Pi and software from sourceforge would cost less than an hour of your time, and would open them up to lots of possibilities, not the least of which is insurance against something unfortunate happening to their only support person.

Re:OK, show of hands ... (1)

Lumpy (12016) | about a year and a half ago | (#40792691)

The owner of the shop understands the old system. so he keeps it working.

The Owner outranks a cobbled together solution, and I prefer the monthly $1000 bonus to my income.

Re:OK, show of hands ... (5, Interesting)

rasper99 (247555) | about a year and a half ago | (#40781017)

I have punch cards on my desk at work so I can remember back to a time when computers and software worked reliably. Why read the holes if the keypunch machine printed the characters above the columns?

I also have the heads and voice coil from a 185MB CDC removable disk drive (approx 15" long) in case I have to smack some young whippersnapper upside the head!

Re:OK, show of hands ... (2)

Teresita (982888) | about a year and a half ago | (#40781177)

You kids don't know how good you got it with your newfangled "punch cards". Why, back in my day we had to use cuniform, on clay tablets. The BC2AD abacus bug made your "Y2K" problem look like a glitch.

Re:OK, show of hands ... (2)

Gaygirlie (1657131) | about a year and a half ago | (#40781289)

You kids had it easy! Back in my day we scrawled our programs on the cave walls and had to relocate ourselves fully if we ran out of storage space. Not to mention that back then bugs actually meant bugs the size of your head!

Re:OK, show of hands ... (2)

icensnow (932196) | about a year and a half ago | (#40781669)

Walls? You had walls? Why, in my day, when we hadn't crawled out of the ocean yet, we had to position ourselves under the computer, blow an air bubble for 0 and make a little vortex for 1, and watch them rise into the reader. A large fish passing by could cause a transmission error that would make us start over from scratch.

Re:OK, show of hands ... (1)

Lumpy (12016) | about a year and a half ago | (#40781785)

Water? You had Water?

We had to float there in the solar system dust disc putting negative or positive charges on dust particles bu bending light rays from the sun with our thoughts.

Re:OK, show of hands ... (0)

Anonymous Coward | about a year and a half ago | (#40781863)

Pff! At least you had spacetime

Before all that mess was around... well we didn't do much of anything and we were THANKFUL

Re:OK, show of hands ... (0)

Anonymous Coward | about a year and a half ago | (#40782127)

Sadly I kept expecting this thread to get funny or something. What a waste of 45 seconds....

Re:OK, show of hands ... (1)

Anonymous Coward | about a year and a half ago | (#40785107)

I know, it was rising to a crescendo and then YOU showed up and brought it all down.

Way to go

Re:OK, show of hands ... (2)

Nimey (114278) | about a year and a half ago | (#40784489)

Luxury. We had to create the entire universe ourselves just to make an apple pie from scratch!

Re:OK, show of hands ... (2)

1u3hr (530656) | about a year and a half ago | (#40786819)

I have punch cards on my desk at work so I can remember back to a time when computers and software worked reliably.

Reliably? I have very unfond memories of my student years, the night before a project was due, standing in a queue in front of a bank of card readers waiting to run my Fortran or Cobol program. Then when the reader chewed up a card, going back to make a new card and waiting another 20 minutes to try to read it again. All the time the card readers were heating up and one by one stopping dead, so the queues got ever longer.

In second year we were granted the privilege of using a keyboard with attached line printer and entering the program with a line editor. And in third year we could use an actual CRT monitor and could use vi.

Card readers weren't quite Spock's "stone knives and bearskins", but close.

Re:OK, show of hands ... (1)

FitForTheSun (2651243) | about a year and a half ago | (#40789221)

/rolls eyes

You know, dude, my abacus has NEVER crashed, not once EVER*. Therefore, all modern computers are unreliable.

* except that time I dropped it

Re:OK, show of hands ... (2)

Antipater (2053064) | about a year and a half ago | (#40781067)

My doctoral thesis was a punch card program that could beat the world's best players at Tic-Tac-Toe. I still keep it around, next to my degree from Springfield Heights Institute of Technology.

Re:OK, show of hands ... (1)

Anonymous Coward | about a year and a half ago | (#40781435)

I solved Tic-Tac-Toe once. I found if you don't make mistakes, whoever goes first either wins, or the game is a draw.
It's not that hard a game though. :)

Re:OK, show of hands ... (1)

Anonymous Coward | about a year and a half ago | (#40781605)

That's only true of certain special cases of Tic Tac Toe.

In N-dimensional tic tac toe. it's not possible to foce a tie unless the size of each dimension is 1 more than the number of dimensions.

Re:OK, show of hands ... (0)

Anonymous Coward | about a year and a half ago | (#40782101)

Oh I didn't know that. I just did the simple single 3x3 grid one. Has someone solved it for multiple dimensions? I might try 3x3x3 for fun.

I was interested in the simple version years ago, as I built a mechanical machine to play it. I kinda cheated by having the machine always play first, and always in the top left corner, which simplified the game even more. :)

It was a really dumb machine, with 9 plates with holes in, in a stack. When you made your move, it moved the plate connected to the square , which would let water flow through whatever holes in the plates that lined up, and output water to the square that would indicate the machines move. Just combinations really, but it impressed some people as it always drew or won!

Re:OK, show of hands ... (0)

Anonymous Coward | about a year and a half ago | (#40782575)

There is a TRS-80 Assembly Language book (not the common one found via search engine) which had several applications and source code. One of the applications was for a Tic-Tac-Toe 3x3 program that would "learn" by weighing the choices it made in a complete game. It included an algorithm to turn the board into a trinary number, rotate and flip the board along all the symmetries, and store the result of its move in a binary tree by turn.
After about twenty games, it would wine or tie every time. I was impressed with the algorithm and the implementation. I did not have a TRS-80 at the time, so I implemented it for other 8-bit computers, with much slower results.
I should extrapolate the algorithm to use larger board sizes, and more dimensions.

Re:OK, show of hands ... (0)

Anonymous Coward | about a year and a half ago | (#40781547)

My college roommate had a copy of FLAG (FORTRAN Load-And-Go) [wikipedia.org] . Guess I should let him know about this.

Re:OK, show of hands ... (1)

Anonymous Coward | about a year and a half ago | (#40781705)

I've got some punch cards from about 25 years ago....(sunglasses on)....but they'll only work with an old Wang. yyyyYYYEEAAAAHHHHHH!!

Re:OK, show of hands ... (1)

techsimian (2555762) | about a year and a half ago | (#40786181)

I've got some punch cards from about 25 years ago....(sunglasses on)....but they'll only work with an old Wang. yyyyYYYEEAAAAHHHHHH!!

You've been waiting to whip that out for 25 years haven't you?

Re:OK, show of hands ... (1)

TBedsaul (95979) | about a year and a half ago | (#40781811)

I keep a pack of unpunched cards, just in case. I also have a very old Radio Shack desktop punchcard reader.

Been waiting for an excuse to break it out and save the day for the last 30 years. Any day now...

Re:OK, show of hands ... (1)

diodeus (96408) | about a year and a half ago | (#40781953)

In Grade 11 (1982) we were forced to use he punch-card machines in com-sci class for the first week. The deck would be sent over to the Board of Ed office and run on their machine overnight, then your printout would be sent back to your school the next day. It was a lesson in being accurate and checking your work, which was sadly lost on the bunch of us.

Then came the day where I found this drawer where all the "holes" went. This resulted in a big confetti fight in the teacher's parking lot.

Next week we moved onto using the TRS-80 Model One, with casettes. Oh joy!

Re:OK, show of hands ... (1)

plover (150551) | about a year and a half ago | (#40796973)

One cool thing about modern equipment is it makes Test Driven Development possible. Crank in a tiny change, and check it instantly. Your turnaround is a few seconds, and refactoring makes you improve as you go. Not that you shouldnt be smart, but you're leveraging the compiler to do syntax checking instead of an overnight drop-n-pray process.

Re:OK, show of hands ... (1, Interesting)

Anonymous Coward | about a year and a half ago | (#40783053)

My former employers signed a contract with the US EPA that included a promise to archive the data from the project. After a while they realized that the idiots who wrote the contract hadn't specified a retention period. So we called the EPA, who laughed and said "keep them forever, suckers!"

So I took the IBM punch cards that held the information and triple wrapped them in plastic and aluminum foil under an inert atmosphere, then took the resulting bricks and stacked them in a dank basement, with signs that explained this data had to be archived forever. They are still there to this day. The contract did not say the data had to be accessible, just that it had to be archived.

Re:OK, show of hands ... (0)

Anonymous Coward | about a year and a half ago | (#40788515)

Me. I've still got the card deck for the first Cobol program I wrote as a trainee programmer thirty years ago. I've also got a printout for the same program after all the corrections were made to my first attempt.

JCL (0)

Anonymous Coward | about a year and a half ago | (#40788585)

OK, this is really cool ... but how many people still have decks of punch cards?

Anyone who has to code JCL [wikipedia.org] on a daily basis?

Horribly inefficient (2)

Dynedain (141758) | about a year and a half ago | (#40780855)

Why not use a desktop scanner with a feed tray and process that? Eliminate the need for the fancy camera rig, arduino, legos, etc.

You could still do all the python processing, but it would be far mor efficient and less prone to bugs.

Re:Horribly inefficient (5, Funny)

Antipater (2053064) | about a year and a half ago | (#40781187)

I'm confused.

Why would you want to use fewer legos?

Re:Horribly inefficient (0)

Anonymous Coward | about a year and a half ago | (#40781337)

Why would you want to use fewer legos?
so you can use them for other things... duh...

Re:Horribly inefficient (0)

Anonymous Coward | about a year and a half ago | (#40781965)

Why would you want to use fewer legos?
so you can use them for other things... duh...

And this is why the LegoMaker will be the most important invention of the 21st century.

Re:Horribly inefficient (1)

mcgrew (92797) | about a year and a half ago | (#40782537)

Back when Hollerith cards were used, computers had little memory, and you had to write the tightest code possible. Fewer legos is like fewer commands. Back then, if you could write the same program I could write with a hundred bytes but use only fifty, you were twice as good as me (or more, because compuers were slower and sparce code runs faster).

Re:Horribly inefficient (1)

RabidReindeer (2625839) | about a year and a half ago | (#40781955)

Why not use a desktop scanner with a feed tray and process that? Eliminate the need for the fancy camera rig, arduino, legos, etc.

You could still do all the python processing, but it would be far mor efficient and less prone to bugs.

Camera rig? Actually, since I do have quite a few boxes of old punched cards, I've been fighting the impulse to build a Lego+arduino reader myself. Ironically, one of the biggest obstacles is that a standard Hollerith card isn't quite an even number of Legos in height.

On the other hand, my Arduino has 12 input channels, so I could hook up a discrete optical sensor to each one!

Re:Horribly inefficient (1)

gl4ss (559668) | about a year and a half ago | (#40783711)

this build uses a camera rig. essentially the arduino is there just to instruct the camera to take a picture and to release the card.
would be far more cooler arduino ad if the arduino did all the punch-card decoding and scanning.

with chdk and flash as card release(for the card release servo activation) I suppose you could do this whole thing without the arduino at all (card comes down, closes circuit which snaps the picture, flash activates and releases the card).

I don't get it (2)

slashmydots (2189826) | about a year and a half ago | (#40780931)

I'm 25 and I play a Sega Genesis emulator sometimes because those old games bring back fond memories and had such a different play style and difficulty. That's sort of the difference here. Who the hell had anything positive to say about punch cards back in the day that would make them want to experience it again? Fires, drops, manual collating, 1 missing card...everyone hated them! Nobody is going to say "oh wow, this is so fun just like back in the day" because it wasn't; it sucked.

Re:I don't get it (3, Interesting)

gstoddart (321705) | about a year and a half ago | (#40781137)

Because, like anything, if you were there for it, you might have some nostalgia for it. It's something you did that most people can't even imagine having to do -- like programming in assembly, or walking to school (uphill, both ways).

I had a prof in university who used to wistfully tell me about toggling in the boot sequence with the buttons on the front of the machine. Hell, he once gave me the manual for the Winchester ST-506 hard-drive controller, and had me write the metal-up code to handle the HD -- about as bare metal as you can get as you shoved a byte into a register and wait for an interrupt to happen to respond to it. I still think that was kind of cool, but it's not something everybody wants to do. But, I gotta say, writing my own cat, ls, rm, mv etc in DOS going straight into the FAT filesystem, and knowing there wasn't a single line of code between me and the hardware I didn't write was fun and rewarding.

I bet most people using computers never had to deal with IRQ assignments for hardware to make them all work together. It was a pain in the ass, but we all fought through it.

And, finally, like so many of these technology projects "because I can" is sometimes all the reasoning you need. People do all sorts of stuff in their spare time, and one person's "shiny fun toy" is another person's "WTF would you do that for?".

Re:I don't get it (1)

mcgrew (92797) | about a year and a half ago | (#40783163)

It's something you did that most people can't even imagine having to do -- like programming in assembly, or walking to school

Why, thank you, young fellow! To tell the truth, I rather enjoyed learning assembly (but I've always enjoyed learning anyway). Assembling the machine code by hand since I had no assembler was a bit of a pain, though... I had to walk to school, but there were no hills. And no computers (I had a slide rule to cheat in math class with).

I bet most people using computers never had to deal with IRQ assignments for hardware to make them all work together. It was a pain in the ass, but we all fought through it.

I kind of miss flipping toggle switches in the old XT when I wanted to add hardware. Plug n play kind of took some of the fun out of it.

Re:I don't get it (1)

networkBoy (774728) | about a year and a half ago | (#40783653)

Plug and pray added lots of fun...
I remember having to use debug to step into ROM code on an ESDI controller to setup sector sparing (62|63|64) sectors per track indicating (2|1|0) spare sectors... Trade space now for reliability later? How much space?
Then you had to "format" the drive with it's sector map, only then was it ready for FDISK and a DOS format.
good times...
Still have that machine around somewhere for two old games... Need to port them to DosBox and toss the machine.
-nB

Re:I don't get it (1)

scsirob (246572) | about a year and a half ago | (#40787707)

C:\> DEBUG
- G=C800:5

How hard can it be?

Before MS-DOS, I owned a Grundy 8200 CP/M system. It had a two-board harddisk controller. After fixing the TTL logic I had to write my own formatting routines in Z80 assembly to get to use the 4 MB harddisk going. (Rotating Memory RMS-504)

Re:I don't get it (1)

networkBoy (774728) | about a year and a half ago | (#40790589)

Let's see...
When I was handed down the 286 with the Wren III 160 meg ESDI drives I was in 7th or 8th grade, so I was still learning some of the math involved... To me I remember hex numbers being magic strings that did things and loosely related to something called binary, also magic strings, that the computer understood. I knew they were number systems, but they were still disjointed from decimal. now after 12 years in the industry preceded by 10 to 15 years of a computer at home, all this and octal are equivalent, but at the time it was magic. Hence the nostalgia.

Re:I don't get it (2)

Samantha Wright (1324923) | about a year and a half ago | (#40781281)

I'm pretty sure the average veteran of punch-card computing was more tactilely aware than a young adult of today like you or me. Missing or dropping a card would've been more rare—and there actually were automated sorting machines, some of which had very clever was of mechanically implementing mergesort.

Not all nostalgia is exactly about the same kind of fun, though. Once you got past the nittiness and the grittiness of how the code was written (i.e. generally in a very limited compiled language, or assembler) the machines behind them were much simpler. It was possible in those days for a single person to be knowledgeable in CPU architecture and operating system design, and still be on the cutting edge in artificial intelligence. Because of that comparative simplicity, the opacity of the tools used to program the computers felt less like a chore and more like a game; c.f. the Story of Mel [catb.org] . The rules were different, code and UI style didn't exist, and the sophistication of a good hack reigned supreme.

As a result, every modest and large program for these older systems was like a little shining gem. A programmer felt proud if something they wrote was well-designed, or shared amongst computer users, or accomplished its work in a particularly clever and memory-saving manner. I think wanting to preserve those accomplishments and memories of one's glory days is at the root of all this.

(And they said computer history would never amount to a degree program—foo to them, say I!)

Re:I don't get it (1)

lgw (121541) | about a year and a half ago | (#40782835)

and there actually were automated sorting machines, some of which had very clever was of mechanically implementing mergesort.

That's radix sort [wikipedia.org] , not merge srot, BTW. For a while "order N sort" questions were all the rage in job interviews, and radix sort was my favorite piece of trivia, as it can be implemented mechanically with very strict memory limits.

Re:I don't get it (1)

bws111 (1216812) | about a year and a half ago | (#40781381)

He didn't create a punch, he created a reader. Maybe someone does have a box of old cards laying around - it would be pretty cool to be able to read them, wouldn't it? Sure, most of the cards are probably worthless, but I bet there are still a few gems of early examples of programming on them.

Something Positive (1)

AF_Cheddar_Head (1186601) | about a year and a half ago | (#40785025)

Had a work study job in the computer Lab at the University of Wisconsin. Went on more than one date with a pretty coed after helping her pick up and arrange the box of punch cards she dropped.

Now that is a positive you don't get with today's mass storage.

freelancer (-1)

Anonymous Coward | about a year and a half ago | (#40780935)

what Kathy replied I cannot believe that some one can earn $8648 in 4 weeks on the internet. Have you seen this site http://www.makecash16.com

Running the code? (0)

Anonymous Coward | about a year and a half ago | (#40781031)

I'm probably missing the point of this - is there anything that describes the system running the code, or is it just a way of converting punches to a listing?

All my best code is on paper tape (2)

ackthpt (218170) | about a year and a half ago | (#40781227)

You insensitive clod!

Re:All my best code is on paper tape (2)

PPH (736903) | about a year and a half ago | (#40785605)

Just like the British Colossus [wikipedia.org] . Interesting note: The CPU clock was taken from the input paper tape sprocket. So you could 'overclock' the system right up to the point at which the tape would break.

Amiga archiver (0)

Anonymous Coward | about a year and a half ago | (#40781279)

Reminds me of an Amiga floppy archiving setup someone did.

Wrong tech for the job. (0)

dave_leigh (67481) | about a year and a half ago | (#40781301)

This thing is glacial, whereas the original card readers read these things as fast as you could push them through. While physical contacts wear out, the best solution is simply some optical LEDs and photoreceptors, and a timer.

Re:Wrong tech for the job. (1)

Ungrounded Lightning (62228) | about a year and a half ago | (#40786501)

This thing is glacial, whereas the original card readers read these things as fast as you could push them through.

Actually the early electronic ones (the ones with relays hooked to the computers with tubes), ran the card about one per second. Still faster than this, of course. (The sorters ran a lot faster.)

Go all the way back to Hollerith and even a sorter was about the same speed:
  - Worker puts the card in the "press" and closes it (pushing pins through the holes and into the (pool of mercury?) back contact.)
  - Solenoid releases the spring-loaded door on the appropriate bin.
  - Worker opens the press, extracts the card, puts it into the bin, and closes the bin door.

Now get off my lawn!

I loved the confetti (1)

BetaDays (2355424) | about a year and a half ago | (#40781365)

I loved the confetti I would takes bags of the punched chads to the local football games and let the kids go wild throwing the stuff. They loved it.

Now, how many punchcards does it take.... (0)

Anonymous Coward | about a year and a half ago | (#40781373)

...for the youtube video ;)

Now do the same for hard drives. (0)

Anonymous Coward | about a year and a half ago | (#40781513)

I wouldn't mind a good, solid DIY hard disc platter reader.
Hard disc platters in almost all cases are completely intact in any hard drive failure. The exceptions are heads smashing in to platters pretty hard, which can do some rather nasty damage to them

I'm still annoyed hard drives won out in the end.
Having the reading mechanism on the device is so backwards.
Oh yeah, certainly now it makes sense since the stupid precision needed for the head densities we have now.
Admittedly this could be replicated with a cartridge system, but it would likely cost more than a typical reader would due to the precision needed in placing the platter and head correctly. Not that much more, but still, near-term profits > all.
But the fun part is the disc and cartridge system itself would likely be considerably cheaper than any hard drive you can find even on Amazon. Considerably cheaper drives and just a once-off occasional expense of a drive to read it. Worth it for not having to deal with data loss due to silly head mechanisms crapping out, or boards dying, or Seagate being rubbish as usual.
RIP ZIP and floppy drives.

Re:Now do the same for hard drives. (2)

Ungrounded Lightning (62228) | about a year and a half ago | (#40786621)

The earliest drives (RAMAC and the like) also had integral disks (with the spin motor integral inside the hub - with spare windings so you didn't lose your data if one burned out). But soon after that they went to removable packs, which was the way it worked for a while, with "washing machine" drives and "single platter in slot" drives.

But they had a lot of problems with contamination. And as the bits got closer together and the heads flew lower they were running into a serious issue.

The breakthrough came when IBM noticed that it had built less than twice as many packs as it had built drives. Turns out that the usual usage pattern was not to swap packs, but to do backups by copying to tape and to do "swaps" by copying tape to disk. Tape was far cheaper per bit than disk packs.

So they realized they could solve the contamination problem by having a sealed unit. The first version had the platters, heads, and cooling air filters in a sealed uint but the motors and actuators were external. The "packs", with heads and such, were still swappable. It also moved the heads to a parking area and let them land on the disk, rather than having a complex mechanism to lift them off the surface when shutting down. This was the "Winchester" - which became a generic term for head-landing disk drives and then for disk drives in general.

But electronics was getting very cheap and the mechanical connection to the external machinery was still problematic. So two generations later they went for an integral sealed unit with motors, actuators, and an electronic board. Thus was the current paradigm born. And with economy of scale and even tighter tolerances it has stayed that way until now (when semiconductors are finally looking to push disk drive technology into end-of-life).

FLL kids could do this (0)

Anonymous Coward | about a year and a half ago | (#40781753)

Guys, this isn't so fantastic. There are first lego league teams who could do this just using a few mindstorms bricks and light sensors -- to boot they would surely not be doing a hand cranked feeder. I'm just not seeing the novelty here -- sure if it were some high rate reader or something but what's the point.

My favorite part about this video. (4, Insightful)

Guano_Jim (157555) | about a year and a half ago | (#40782365)

When I hit that video the first time, the first couple of comments on that video aren't "cool!" "nice job!" or anything resembling constructive criticism. It's all "this is the wrong tech for the job" "seems like a hell of a lot of effort just to read what's already on the top of the card," etc.

Haters gonna hate, I guess. But what ever happened to just enjoying a hack for a hack's sake?

I think it's clever. Who cares how much time the guy spent, what technology he chose, as long as he enjoyed doing it.

Re:My favorite part about this video. (0)

Anonymous Coward | about a year and a half ago | (#40787477)

In the nineties, forums were something where you could net a handful of comments, but to reach them you had to be looking for them. These days Jerko Johnson Jr. is watching something and POOM! a link pops up his "related" stream or inbox that shouldn't have...suddenly he starts commenting where he has no business to even be, because no involved searching is needed to enter the home of new content.

Propriety is over: Unlike the internet forums of old, the untrained masses easily find LOTS of things now shared by others because of all these one-stop-for-EVERYTHING services: facebook, youtube, reddit, etc. Free membership makes it pretty hard to just ignore things when you don't need to reregister nor pay nor watch daily maximums for comments. Webring / shrine guestbooks were popular in the nineties --there were no PHP forums. Worse, the unspoken netiquette required that you only said good things...you didn't have a comments page per content page, and you wouldn't expect daily updates, so talking about specific content was unheard of in guestbooks. You'd come across content written months or years above and praise the site, but you'd go on your merry way because sites weren't huge and you could read them whole in a few minutes. There were no handles, so you just filled some text fields that expected your real name. It feels like writing a semi-intimate message on someone's dorm room whiteboard. All that is lost on a place where you pretty much cause text to exist without any context and accountability.

Imagine writing atheist messages in a church bulletin board while everyone is still there without expecting some reprimand.

Wrong Title (1)

drewco (1631735) | about a year and a half ago | (#40782531)

The title of the post should be "Resurrect Your Old Code With a DIY Rube Goldberg Machine Punch Card Reader" It's neat, but why the hell would you use a camera + image processing software to actually read the punches when there is an arduino right there?

What's the point (1)

na1led (1030470) | about a year and a half ago | (#40782661)

If you're doing this to learn something, do it the old fashion way. This method requires new technology, and thousands of lines of program code, to extract a few dozen lines of old code.

Re:What's the point (1)

Mal-2 (675116) | about a year and a half ago | (#40784153)

It's hardware that he already had though. If a card reader could be had at a local garage sale and refurbished in a weekend, he probably would have done exactly that -- but it can't, for most people. The code was probably fun, as was building it from Lego pieces.

Strange device that makes the puch cards move? (0)

Anonymous Coward | about a year and a half ago | (#40783111)

Are the plans for the strange device on the right hand side of the video available, that's turning some kind of cranks to make the next card move, after the previous one was read? That seems to be some complex thing, but no explanations :(

Kinda Slow (3, Funny)

twmcneil (942300) | about a year and a half ago | (#40783655)

Compared to the real thing, this is a wee bit slow. But of course I threw out all my punch cards years ago. They warped so bad in about 12 months time that they would jam the reader.

Quick way to win friends, jam the reader with about 15 people behind you.

Now get off my lawn!

The correct term is PUNCHED card (1)

Al Kossow (460144) | about a year and a half ago | (#40783975)

Look at the title of books and documentation from the period when they were in use. They
ALWAYS refer to them as PUNCHED cards.

Overengineered (0)

Anonymous Coward | about a year and a half ago | (#40784277)

It's just 12 rows of holes or non-holes... if you're already going Arduino... why not just... have a row of LEDs and photosensors...?

punch cards at school, octal bool loaders... (0)

Anonymous Coward | about a year and a half ago | (#40786261)

My CompSci 101 class had 200 students, and there were 4 terminals, with a long line for each machine, back in 1981. But I had become friends with Bob, who was an IBM service tech who supported, among other tools, punch machines. There were a dozen of them, and no line. No one but Bob and I used them, and we wrote 12, 18, 30 line programs in PL/1 and were out of there. He taught me how to correct typos on a card by reading it in and repunching it with a correction added in column 41, or wherever.

Before school I worked at a newspaper switching from hot type to cold type - cast molten lead type to printed-on-photo-paper type for offset use. The printers needed their boot-loader loaded in with octal switches on the front of the machine. The computers (PDP-11s, two of them ganged to handle tall the terminals) were boot-loaded with 8-bit mylar tape, plastic so it could be loaded once or twice a day, every day, for the life of the system.

I once spent the night lhying on the floor with a multi-meter on the phone with a tech in New Hampster, trying to troubleshoot an OCR, used for loading stories typed by reporters that didn't rate a terminal, mostly used by editors and type-setters doing ads.

Wonderful mix of 19th century and late 20th century technology! There was a fire extinguisher on one wall, surrounded with pots of molten lead, a glass ball filled with carbon Tetrachloride, a terrible solvent, poisonous, carcinoginic, but also a fire extinguisher! I bet it's still there on the wall. They can't tear the building down, or repurpose it. it's completely contaminated with lead, among other deadly substances. I'm 62 now, amazed that I made it this far.

Punch-cards. My friend Steve worked as a machine room operator, loading decks of cards, on summer break from school. He remembers a guy who came to the window each evening, with a huge box of cards, a complex program, larger each time it ran.

Then one night, he was bringing his huge deck to the machine room, and dropped it - without sequence numbers!!!! Months of work, gone in an instant. Steve never saw him again.

Suicide rates (0)

Anonymous Coward | about a year and a half ago | (#40786807)

Suicide rates to increase on cmpletion of the project

One terabyte in boxcar-fulls of punched cards... (2)

jabberw0k (62554) | about a year and a half ago | (#40787059)

A now-common one-terabyte disk drive, represented on punched cards, would occupy a cube approximately 15.3 metres (50 feet) on a side... similar to a decently sized five-story building, and would have a mass of over 34,000,000 kilograms.

If loaded into standard U.S. railroad boxcars, that one terabyte would physically fit inside 17 boxcars but because of weight would have to be divided amongst 631 boxcars (each boxcar being rated at 60 tons, or about 1.5 terabytes' worth of punched cards). That works out to a freight train over seven miles long.

Re:One terabyte in boxcar-fulls of punched cards.. (0)

Anonymous Coward | about a year and a half ago | (#40788811)

A now-common one-terabyte disk drive, represented on punched cards, would occupy a cube approximately 15.3 metres (50 feet) on a side... similar to a decently sized five-story building, and would have a mass of over 34,000,000 kilograms. If loaded into standard U.S. railroad boxcars, that one terabyte would physically fit inside 17 boxcars but because of weight would have to be divided amongst 631 boxcars (each boxcar being rated at 60 tons, or about 1.5 terabytes' worth of punched cards). That works out to a freight train over seven miles long.

I think you meant "about 1.8 gigabytes'" there (34,000,000kg / 1024 GB in a TB * 1.8 = 59,765.625 kg).
</pedantic>
PS: Don't get me started on binary equivalent [wikipedia.org] of the long scale vs short scale [wikipedia.org] thing, or such peculiarities as long [wikipedia.org] , short [wikipedia.org] , or metric [wikipedia.org] tons.

Of all the things I lost I miss... (0)

Anonymous Coward | about a year and a half ago | (#40789509)

Of all the technology changes over the last 50 years the one that I miss is synchronous error reporting. It used to be that when the cpu encountered an error, the exception generated could be traced to a specific instruction in the stream. Debugging was a lot simpler than when instruction pipelining came along -- you knew an error had occurred but precisely where was a bit fuzzy. My current Win7 systems go beyond that, the error reporting mechanisms simply ignore wide swaths of conditions -- so the reliability reporting always looks wonderful. So we can pretend that we are in the best of all possible worlds...

Slowest punched-card reader ever (1)

Theovon (109752) | about a year and a half ago | (#40789947)

I don't want to pick on this too much. There are plenty of cool things that this developer did that are intesting in their own right. For instance, the image processing. But it looks like this machine is mostly manual. Doesn't it seem obvious to make it crank out the cards automatically? And couldn't the pictures be taken more quickly? And aren't there more efficient ways of detecting the holes in the cards? It's an interesting machine, but it's definitely not an efficient solution for this specific problem.

Check for New Comments
Slashdot Account

Need an Account?

Forgot your password?

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>
Sign up for Slashdot Newsletters
Create a Slashdot Account

Loading...