Reading Punch Cards on Today's Hardware?

Gone Jackal asks: "I've run into a ludicrous situation. I'm a classicist at the University of Chicago, finishing off my undergraduate career with a bachelor's thesis. As part of this, I'm working on poetry and meter in Vergil. Imagine my joy when I found in the library catalogue a 20 volume set of books on just this topic, published between 1973 and 1985. Now imagine my horror when I open up the first volume and see that the entire work is on punch cards. What the heck do I do with punch cards? The CS department has laughed at me, and mainframe computing says they're not allowed to help since I'm not an administrator. Any suggestions on where to go, what to do to read these, or even find out what's on them? (The series title, in case you wanted to know, is 'Metrische Analysen zu Vergil', by Wilhelm Ott. It's in German)." Oh man...punch cards. I can understand Gone Jackal's frustration, however I'm more surprised that such works weren't converted to better media a long time ago.

Re:

[account_deleted]

Comment removed based on user account deletion

Re:Where the hell is your advisor?

[Alik]

I'm not whining, but one's advisor *is* the standard first resource, and he wasn't mentioned in your original question. Unfortunately, sounds like you're taking all the necessary steps.

Scanner

[affegott]

Hmmm... seems all you need is a scanned image on black background... then some clever software. I bet some of the CS profs at NIU could help... we are the mainframe capitol of the US of A.

Long live the mainframe!!!

j/k.

Peace out.

Re:Where the hell is your advisor?

[Gone Jackal]

ahem...yes, i would think so too, eh? My advisor's response was 'I have no idea; contact the CS department'. Oh well.

The problem is not obtaining the books; I have the books and the cards. The problem is finding a machine to read them, or something that will translate them. So far, I've emailed the classics bibliographer here at the U. of C., the author, several archaeologists I know who did something similar on punch-cards back in the 70's, and a few contacts on the Tübingen website for their text-processing program.

And, by the by, this seemed like the sort of strange problem Ask Slashdot would be interested in; this is far from my only resource. If you don't like the question, stop whining and post elsewhere.

I've got just the solution...err...

[Slynkie'sEvilTwin]

It just so happens that my Operating Systems class is codin' a virtual machine and OS that has a card reader, so send 'em on over! ;)

Re:Thinking it through

[Skapare]

Punch cards were encoded in Hollerith code. Digits had 1 hole (in rows 3-12 labeled 0-9). Upper case letters had 2 holes (one in rows 1-3 and one in rows 4-12). Lower case letters had 3 holes (two in rows 1-3 and one in rows 4-12). Other characters were coded in other ways. On the IBM 360/370 mainframes, the card reader or its controller converted them to BCD or EBCDIC before the data went over the byte channel to the channel process on the CPU.

Take a trip to Radio Shack!!

[Slynkie]

Surely, someone could hack together a quick perl script to analyze the scans from a CueCat, right?! *cough*...;)

duh.

[Zurk]

take em to these guys, get em translated. http://www.wirednews.com/wired/archive/7.03/punchc ards_pr.html

Do the cards have text at the top?

[BobGregg]

It's too bad there's no way to feed the cards automatically to a flatbed scanner. Most punch cards had a printed copy of the characters the punches represented across the top of each card. If you could feed them to a scanner in some automated way, then you could use OCR software and a simple script to strip out the text and assemble it all together fairly quickly.

Alternatively, there may be some way to either a) contact the original author (Wilhelm Ott), or b) find a copy of the work in a German library, and get an inter-library loan. When I entered the following search into Google...

"Metrische Analysen zu Vergil" "wilhelm ott"

...I was auto-forwarded to the following page on Amazon.de:

http://www.amazon.de/exec/obidos/external-sear ch/028-6615 460-4707748?tag=artvisitwww&keyword=Wilhelm+Ott&mo de=books [amazon.de]

From the looks of things, it may well be that some, or even all, of the volumes in this series are still in print. It looks like it would be expensive to spring for the entire series (several hundred US$), but if they're still in print, then it's likely there are copies available in a German library somewhere as well. If it's that important, you could try working with your advisor on getting an inter-library loan. As with most software tasks, it's better to find an existing solution (i.e. another copy of the books in print) than to create a new one (all that scanning work) yourself. :-)

Contact one of the old computer companies.

[SIGFPE]

Like IBM or Digital (Compaq). Compaq certainly have a bunch of people trying to keep alive one of each model of computer from their lines. I'd have thought a company could get a little bit of publicity out of a project like this!
--

Punch Card Fun

[BigBlockMopar]

Ya know, I love computer history. I've always thought it would be fun to have a punchcard reader connected to my notebook computer.

That way, I'll be able to read the cards and save the data to the 8" floppy drive I've already hacked on.

(Old 486SX notebook, don't worry.)

Thinking it through

[martyb]

Okay, even if you COULD get these punch cards read in, I'd expect you'd still have some language Issues:

1. Human language: German. If you can't read German, it's not going to be of much use, will it? Sure, you could use BabelFish but it will certainly lose something in the translation.
2. Programng language and environment. I would expect the punch cards most likely would contain both application(s) and data. Do you have access to a machine that could RUN these? Imagine the application(s) were designed to be run on a Commodore 64 - could you even find an environment where you could run these applications?
3. Character Set. Punch cards, IIRC, were generally encoded in EBCDIC. (At least they were when I used them in college.) Do you have access to a system that can even use this character set?

I'm not trying to disuade you from continuing to investigate this resource, but to point out there's more here than may be apparent at first glance.

This is indicative of the general class of problems that I've often run into when trying to locate information. When I've hit a rat hole, back up and THINK! Could someone else have run into this same problem, or am I the very first person who has run into this? That simple idea has helped me to stop what I had been doing, back up, and try another approach while making use of what I've learned so far.

1. I noticed others had posted here with some links that indicated these works have ISBN numbers. Search on those.
2. Hmmm ISBN? There's a good-sized library in Washington, DC. A search on google [google.com] for : "Library of Congress" took me to a web site [loc.gov] where I could search their collection in many different ways. Not only that, I found other works in the same subject area (Sobjects:):
   • Virgil--Versification.
   • Latin language--Metrics and rhythmics.

In short, when I hit a brick wall when searching for information, and I've got a headache from banging my head against that wall, it's a sign to me that maybe it's time to step back and try another approach. Hope this helps!

You might contact the author..

[MinusOne]

With a little research, I found out that he is still at the University of Tübingen in Germany - If the search I did is correct, I think his email address is :
[e-mail]

He may have his data in a more modern format, or be able to point you to a place where you can find it.

I don't read German, but the German Amazon.com has his books listed - here is a sample link:

http://www.amazon.de/exec/obidos/ASIN/3484600314 /qid=971051419/sr=1-6/028-4345424-2850943

Good luck!

Since all you have to do is read...

[Raetsel]

Why not set up a sheet-fed scanner with a black background and then run the cards through it. (Since they're old, and library materials, see if you can find one with a mostly straight feed path...)

Then a custom app (of your own -- or a willing CS major's -- design) can look for the presence or absence of contrast in specific locations. Punch cards have registration marks, so exact positioning on the scanner is not terribly an issue. Then you have the data in a machine-readable format, and conversion to something you can use should be the downhill stretch.

Just an idea, sorry I can't help with an implementation... I hope this gives you some inspiration.

Where the hell is your advisor?

[Alik]

If you're doing a thesis, you should have an advisor. This sort of thing is exactly what advisors are for --- you find who has the stuff you need and your advisor kicks ass until you get it. This is one of your basic perqs in a mentoring relationship. I ran into a not-quite-as-bad situation when doing my thesis in CS, and my advisor hassled department and campus admins to get the data I needed into our machines.

Failing that, search the web for places that have the necessary machine type and beg. Try the comp.* newsgroups. Ask the librarians --- their job is to get you your data. Move your damn ass instead of whining to Slashdot.