Corrupted Word Files for Sale

On one hand, this is clever:

We offer a wide array of corrupted Word files that are guaranteed not to open on a Mac or PC. A corrupted file is a file that contains scrambled and unrecoverable data due to hardware or software failure. Files may become corrupted when something goes wrong while a file is being saved e.g. the program saving the file might crash. Files may also become corrupted when being sent via email. The perfect excuse to buy you that extra time!

This download includes a 2, 5, 10, 20, 30 and 40 page corrupted Word file. Use the appropriate file size to match each assignment. Who's to say your 10 page paper didn't get corrupted? Exactly! No one can! Its the perfect excuse to buy yourself extra time and not hand in a garbage paper.

Only $3.95. Cheap. Although for added verisimilitude, they should have an additional service where you send them a file -- a draft of your paper, for example -- and they corrupt it and send it back.

But on the other hand, it's services like these that will force professors to treat corrupted attachments as work not yet turned in, and harm innocent homework submitters.

EDITED TO ADD (6/9): Here's how to make a corrupted pdf file for free.

Posted on June 9, 2009 at 6:46 AM • 76 Comments

Comments

Da ScritchJune 9, 2009 7:20 AM

I hope they will refund you if it's openable with OpenOffice, wordfile, or any other program.

Because, okay, let's say you send your annual report to your boss and he's replying “can't open it”... But what if he foward the document to someone else, open it and read your “document”. And what kind of text is there inside ?

LollardfishJune 9, 2009 7:20 AM

The net effect of this type of thing is to turn syllabi into EULAs, rather than statements of expectations, assignments, and learning goals. I have colleagues with pages of legalistic CYA clauses on their syllabi in order to forstall complaints (or lawsuits), although I've resisted the trend mostly. Still, I do write:

"Work submitted electronically is not considered 'turned-in' until you have received a confirmation email from me."

KateJune 9, 2009 7:29 AM

This irritated me enough that I spent $3.95 several times to download different copies. They're using the same corrupted file. md5 detects it easily. I let all the professors I know exactly how to do it, and as soon as I get 3 minutes I'll write a Perl script to make it even simpler. Lame.

KieranJune 9, 2009 7:30 AM

Why not run email through a program that sanity-checks incoming attachments and bounces those that are empty or corrupted?

KateJune 9, 2009 7:33 AM

Oh hey can't edit comments. That should, of course read "I told all the professors I know exactly how to do it", and here's the hash signatures.


MD5 (10-page-word-file.doc) = f1aecc8c7a6af725a225a1a91e826c43
MD5 (2-page-word-file.doc) = 67b9feede4b1d9e76db593c923ec6bbb
MD5 (20-page-word-file.doc) = 1edf09e903b1877ee805e59c9780bcad
MD5 (30-page-word-file.doc) = d252d138aa5b34feb5b6091c03e383e2
MD5 (40-page-word-file.doc) = 74733afbde459b418d019f733e60f45c
MD5 (5-page-word-file.doc) = f103f8b607d84a0e49367c751e8a7d75

(At least until they catch on and change the files, I guess.)

cryptofishJune 9, 2009 7:42 AM

Because opening word file in notepad and typing or deleting some random garbage is so hard that people need to buy such files online.

BTW Kate you are so naive, really it's too much for me.

damonJune 9, 2009 7:46 AM

I've often been able to look at the contents of .doc files with a text reader when MS Word choked. There are a lot of nonsense characters, as you'd expect, but you can also see a lot of readable text. If your professor opened your fake corrupted paper this way, you'd be caught.

(Off-topic: It annoys me that my kids are required to hand in middle school science projects in PowerPoint format. Why not use a free format?)

Vincent GableJune 9, 2009 7:47 AM

Many schools already use some kind of plagiarism detection system, like turnitin.com, for paper submissions. So after any student has submitted one of these papers, subsequent attempts to uses them will be flagged as plagiarism and rejected anyway...

TequilaJune 9, 2009 7:55 AM

When I was in high school in the early 90s, I would use many blame-the-computer techniques like this to try to delay assignments. The last straw, for my teacher, was when I handed in a four-page paper that was just four copies of the first page. "Oops," I said, "my printer must have screwed it up!" She failed me anyway.

AnaJune 9, 2009 7:57 AM

Yet another good reason to ask students to turn in homework in PDF format. There are free converters for every platform, file size is smaller, less chance of passing on a virus, and the document will look the same to student and teacher. Students would also be free to produce the document using any software they choose.

bobJune 9, 2009 8:03 AM

Wow, times have changed. Back when I started college I was one of the very few to have a computer (RS mod 1, then PC clone when those came out). I had to get permission from instructors before I could use a PC for assigments (which usually involved printing some sample pages to show them what the output looked like).

These guys should wise up. They should have you enter a bunch of keywords and sprinkle them throughout the fake file so that examination with a hex editor would show plausible deniability (so to speak). Maybe have a cliff note tie-in.

Tom WelshJune 9, 2009 8:07 AM

While no corrupted file can be read without some sort of corrective work, I recall the case of a huge indexed data file that I was once asked to examine (and recover whatever I could). As luck would have it, I had just returned from a week-long training course on the structure of files and how to deal with exactly this kind of situation. Opening the file header with a special editor, I eventually flipped a single bit - and the entire file came back, as large as life and 100% legible.

Someone could come badly unstuck if they present a scrambled file as their own work - only to have the school turn it over to a programmer who fixes it and reveals it to be something entirely different.

Tom WelshJune 9, 2009 8:11 AM

"It annoys me that my kids are required to hand in middle school science projects in PowerPoint format. Why not use a free format?"

What's good for Microsoft is good for America, damon! 8-)

ScottJune 9, 2009 8:20 AM

@kate: try running it through linux `strings` and see what, if anything you can find.

A genuine file that happened to get corrupted would still contain some text appropriate to the assigned topic.

Maybe something like
$strings corrupted.doc | grep -P '[\w\s,.]{30,}'
would be good, to find entire sentences.

Petey BJune 9, 2009 8:31 AM

Really? How hard is it to open your rough copy in notepad, change the encoding and remove some characters from the header, and save it.

Anderer GregorJune 9, 2009 8:35 AM

Why, if the student can not proof that the file was *not* corrupted before the deadline, I see no reason to assume it turned it. I mean, it is their problem, not mine.

(and personally, I make it clear that I consider .doc or .ppt as "not turned in" anyways ...)

aikimarkJune 9, 2009 8:44 AM

@cryptofish

Since we don't know the method the vendor uses to render these Word documents un-openable, there is no guarantee that text file editing would preserve this desired state. It would probably be worth some educator's time to create a script to look for particular patterns in these files that are not found in 'normal' Word documents. That would catch the subset of students that successfully alter the hash while preserving the un-openable status.

I applaud Kate's efforts to thwart this scheme. So I must disagree with your 'naive' assessment. Her efforts would catch those students who follow the vendor's instructions and simply rename the file before submitting their 'work'.

Shachar ShemeshJune 9, 2009 8:45 AM

"All electronic submissions must contain SHA-256 of submitted document as part of the email's text".

That pretty much kills this particular cheat.

Shachar

Aaron ToponceJune 9, 2009 8:45 AM

And how do you turn in a printed copy of a corrupted file? Exactly. You don't. As a current college student, I have yet for a professor to ask for an electronic copy.

Alan KaminskyJune 9, 2009 8:55 AM

"It's services like these that will force professors to treat corrupted attachments as work not yet turned in, and harm innocent homework submitters."

Not likely.

I'm a professor of computer science. I require my students to turn in all assignments as email attachments. I only accept PDF files for reports, plain text source files for programs. I do not accept an assignment until I verify that I can read the PDF document or compile the source files. If I do not accept an assignment, no matter what the reason ("corrupted" files, compilation errors), the student has to submit the assignment again. If I have not received a successful submission by the deadline, the student gets a grade of zero for the assignment. I make my students well aware of my policy.

I've never had a student turn in a "corrupted" file in order to get more time for an assignment.

uberdilligaffJune 9, 2009 8:56 AM

This service is just stupid. Every file (including corrupted, binary, or any other) can be immediately opened and inspected with a hex editor. Word docs carry metadata that tells Word how to interpret and format the file contents -- if corrupted, then Word will not be able to open the doc. Unless the file is actually encrypted (not just corrupted), it is trivial (as others have observed) to use readily available and free or inexpensive tools to display the textual content and see the gist of what the supposedly "corrupted Word doc" actually contains.

billJune 9, 2009 8:57 AM

@Kate "This irritated me enough that I spent $3.95 several [x6] times "

They annoyed you, so you *gave* them $23.70?!

I once paid a lone (and terrible) violinist in a restaurant to stop, he quickly moved to another couple, who did the same.

There's a lesson in predicable irrationality here, but damned if I know what.

RandyJune 9, 2009 8:59 AM

@damon

Consider yourself lucky.

My children complain that I don't have the *latest* version of MS Office installed on my computers at home. Somehow the different versions cause their PPT and DOC files to display differently. Different fonts or backgrounds or some such.

As far as the corrupted file goes. I'm guessing the professor would simply have to say, "OK. You do make backups...right? No? Then I'm guessing that the next time you will."

Randy

GradSchoolJune 9, 2009 9:11 AM

@ Aaron: "And how do you turn in a printed copy of a corrupted file? Exactly. You don't. As a current college student, I have yet for a professor to ask for an electronic copy."

We have various modes of submission here: 1) hard-copy to School Office/drop box; 2) electronic submission system; 3) email.

Each approach has it's advantages/disadvantages.

JRRJune 9, 2009 9:16 AM

I read a story once about a guy who was getting pressured in meetings daily to get a data conversion done; his was the first step and another team would take the data tape and run it through their code. The other team's leader was going along with pressuring the first guy, saying he was holding up their team.

The first guy's team was working hard but needed another few weeks. He gave them a blank tape, saying it had the data on it. Now the pressure was on team #2 - turned out they weren't anywhere near ready for the data anyway, and it was 2 months before they even TRIED to read the tape. Then they came back to team #1 (which by then had actually finished) saying the tape was unreadable, so they got a new "copy" of the data tape.

ax0nJune 9, 2009 9:31 AM

My mother -- a college English professor -- would often sequester my "forensics skills" (usually little more than strings [filename]) to get the contents from a student's corrupted assignment. Then the whole Writing Center would start sending people my way. That got old quickly, however I met some cute students that way. More often than not, students were positively delighted to get a good chunk of the text back, often requiring only formatting and re-typing what little was missing.

Now that I no longer work at the college, my mother's entire department uses a freeware MS Word recovery tool that I pointed out to them. I'm guessing whatever can be recovered from these "corrupted" files would me more than enough to dispel students' assertion that "the floppy drive ate my homework"

Impossibly StupidJune 9, 2009 9:32 AM

I love the investigative work in the comments here, and I'm going to agree that the charge of plagiarism against a student for using the supplied files is *far* worse than the risk of other students with legitimately corrupt files being harmed.

Another policy to potentially adopt if this becomes a serious issue is to only allow resubmission via diff files. Anything too large is a substantial revision and should be downgraded accordingly (e.g., 10% off for a 10% difference).

Andre LePlumeJune 9, 2009 9:33 AM

This is an outrage. The very idea of using such a thing as an excuse for sloppy or nonexistent thought is a #`%${%&`+'${`%&NO CARRIER")

JohnJune 9, 2009 9:38 AM

Of course, sending professors documents in docx format is probably about as effective as this service. The few times I used that format, I got replies that they had no idea how to open the file. Good until Office 07 becomes more common, I suppose.

HJohnJune 9, 2009 9:45 AM

This is easy for a professor to resolve. Just tell students up front that backups are their responsibility because saying their file got lost or corrupted will not be an excuse.

It's in their best interest to learn early on the importance of backups anyway.

larsJune 9, 2009 9:50 AM

I think that, as many talented people are prone to doing, many posters here are overestimating the computer savy of the "average" educator. This dodge might not work well against computer science professors, but you could always try the art history department.

John FJune 9, 2009 10:07 AM

My wife is currently teaching online classes for a fairly large college, and I pointed this article out to her (maybe a week ago).

Her response: "I deduct 10% per day for late work. It's XYZ Policy, and all of the students have agreed to it." :)

Clive RobinsonJune 9, 2009 10:22 AM

Unix has two handy tools for recovering word and works documents.

If you look at a works document you will see that each char has a format byte (usually 0x00). Using the unix "tr" prog will remove these chars quite easily.

You then pass either the word file or tr'd works file through strings into a text file.

Open the text file in wordpad and check it for rouge/missing chars. Sort them out a quick cut-n-paste back into word or works and bob's your aunty...

I was doing this for students at Kingston Uni back in the early 90's and it still works today.

The new MS docx format is detailed on the web somewhere (if I remember correctly it's just compressed XML) so it likewise can be recovered.

As for the encrypted versions of .doc prior to Office97 the method of encryption and the standard opening header in word files made it fairly easy to analyse and decrypt (MS have a history of using either weak crypto/obsfication or in an incorrect way that makes it weak such as the well doccumented use of ROT13 TEA).

Rick AuricchioJune 9, 2009 10:23 AM

In the 70s, Saturday Night Live ran a parody of a FedEx commercial.

The company would backdate your shipment and apologize, claiming they caused the delay in arrival. In reality, of course, you were behind schedule shipping that important report.

Todd KnarrJune 9, 2009 10:41 AM

@Rick: "Tardis Express. When it absolutely, positively has to be there before you sent it."

JohnJune 9, 2009 10:42 AM

Rick:

Google had an April 1 service this year that did the same with e-mail, time stamping it in the past. GMail Custom Time?

KateJune 9, 2009 11:01 AM

@bill - No, they let you download as many files as you like. I only spent $11.85 on the matter. I like to triple-verify. Really, though the extra $3.95 was gratuitous.

@cryptofish -I haven't been accused of naivete in a while! Nah, not naive, just a student myself and hence painfully aware of the perfidy and lameness of some of my classmates.

@Scott - strings doesn't turn up anything, neither does just loading it into a text editor (which in a real corrupted file should produce the odd human-readable string amongst the gobbledygook.)

Pat CahalanJune 9, 2009 11:04 AM

Allow me to channel grumpy old man for a minute...

> But on the other hand, it's services like these
> that will force professors to treat corrupted
> attachments as work not yet turned in, and
> harm innocent homework submitters.

Back when I was in college (well, okay, as an undergrad), a paper that wasn't ready at due date wasn't ready. If you forgot your password, you had to show up with your ID at the sysadmin's office between 1:00-3:00pm on a Friday, and if your assignment was due on Thursday that was just too bad.

As a result, I graduated with the ability to remember passwords and make backups, two skills that are arguably as important as *anything* else I learned in college.

Davi OttenheimerJune 9, 2009 11:23 AM

"guaranteed not to open on a Mac or PC"

mac or pc? did they mean operating systems? what kind of guarantee? seems trivial to open corrupted files.

wonder how they decided to advertise corrupt files below $4. who are they targeting? the desperate students that would fall for this could probably be swindled out of far more.

Gordon DJune 9, 2009 11:41 AM

I had a stat final once that I had to do with my computer. I used AbiWord on my eee, and saved it as a doc for my professor. Little did I know Abiword's doc export function sucked, and when my professor opened it only the first page with little other than my name on it showed up, and he gave me a 5/100. Ouch!

Fortunately the text was still in the document, if you opened it with notepad. It just didn't open properly in MS Office (or OpenOffice, or even AbiWord for that matter). So I sent him a PDF of the original, which I had fortunately had the foresight to save in native format as well as .doc, and he had the campus computer people compare the corrupted .doc to the pdf I sent, to make sure I wasn't trying to pull one over on him. I don't really blame him for being cautious, given this sort of thing, but it was annoying having to wait and wonder if I was going to fail the course.

PhillipJune 9, 2009 12:13 PM

dd if=/dev/urandom of=~/my-messedup-up-word-file-5mb.doc bs=1M count=5

There, I gave you all the secret. Now you can save yourself money and get your own unique file. Thank me later by sending $2 to my paypal account. I saved you 50%!

PhillipJune 9, 2009 12:14 PM

If you send an extra $2, I'll tell you how to create a 10MB corrupted word doc.

pfoggJune 9, 2009 2:06 PM

"But on the other hand, it's services like these that will force professors to treat corrupted attachments as work not yet turned in, and harm innocent homework submitters."

The comments above list three strategies: test the file, require a different format (which is presumably has less potential opacity when corrupt), and announcing a policy up front of placing the burden of delivering a timely, readable document on the student (and making reasonable precautions a normal part of submitting homework).

It seems to me that all three of these solutions fail to harm 'innocent homework submitters', and if we assume that some students were already using the corrupt submission ploy, all this service does is to encourage making suitable security procedures more common, all three of which tend to reduce the case of honest submission errors being indistinguishable from fraud.

If we grant that historically at least some honest submission errors have been treated as fraud, it looks like providing this service would actually be a net benefit to the honest student...by encouraging schools to remove the likelihood of this sort of error.

elseJune 9, 2009 4:28 PM

"But on the other hand, it's services like these that will force professors to treat corrupted attachments as work not yet turned in, and harm innocent homework submitters."

When I was at uni a few years ago, that's exactly how they behaved. Uni lecturers (and tutors/markers) are time-poor. The student is responsible for turning in their work in the correct format at the correct time. Anything else is unlikely to be met with sympathy or flexibility.

stevelaudigJune 9, 2009 5:41 PM

I require both electronic and paper submission and intermittently enforce it. It seem to avoid such problems. Now I will ad a no-excuse for corrupted file rule which would require that the student check the work as it has been handed it.

paulJune 9, 2009 7:20 PM

I could easily see a certain kind of student lawyering up with the argument that it was the professor's software that corrupted the file. Want to spend the next year in court paying expert witnesses, or want to give the little brat a few extra days?

(At the moment I'm cleaning up after OpenOffice's habit of "simplifying" tables of contents and reformatting tables into invisibility when opening .doc files, so such an argument seems plausible to me.)

Michael RobertsJune 9, 2009 9:23 PM

@Rick - that wasn't a backdating service, it really was time travel. "Temporal Express: when it absolutely, positively has to be there last Tuesday." The plot was the boss on the phone, yelling, "Where's the Robinson report? Johnson, YOU'RE FIRED!" Johnson gulps, gets to work on the Robinson report, finishes it, takes it to Temporal Express, and the end of the commercial is the boss on the phone, yelling, "The Robinson report is fantastic! Johnson, I'm promoting you!" And Johnson, looking very relieved.

It was brilliant.

AnonymousJune 9, 2009 9:43 PM

To those stressing the importance of backups, it misses the point. These are the two scenarios you expect the educator to be able to distinguish between, in order to invalidate this as a method of cheating:

1:
a) Student completes assignment
b) Student emails assignment
c) Email is corrupted in transit
d) Teacher notifies student
e) Student retrieves and sends backup

2:
a) Student fails to complete assignment
b) Student sends corrupted file
c) Student continues to work
d) Teacher notifies student
d) Student sends newly complete work

In both cases, all the teacher has is a corrupted file, and the student has the uncorrupted file immediately available - the cheater having had *some* extra time to work on it. Backups are completely irrelevant.

cheaterJune 9, 2009 11:29 PM

I've been doing this for years. I once sent in 200k of /dev/random as a "pdf" file of my final paper (actually the only paper for the class), hoping to buy myself an week.

Two months later, I got an email back from the professor saying something to the effect of "Your paper was good, but would have benefitted if you came to office hours"

I got a B+.

monokleJune 10, 2009 12:25 AM

@cheater, hope you learned your lesson young man =P

This tactic has been happening for awhile albeit in different forms. The college I went to had a policy that even if the fire alarm went off, everyone would evacuate and continue the midterm outside in the assembling area, precisely because someone would pull the alarm to get extra time studying. Shifting the burden back to the students seems to be the most effective way to curb this malfeasance

JamesJune 10, 2009 12:26 AM

OMG did you all see the FAQ page? It's hilarious!!!
My Fav!

Q: Who is behind Corrupted Files?
A: Umpa Lumpas, tons of them. Pauly Shore is here too!

Who is this guy? I can't stop laughing!!!

YoYoYoYoJune 10, 2009 12:54 AM

This is the dumbest thing I've read all week. Whether in school or at work, if you don't deliver the product it's on you. Period. Corrupted file ? Who gives a sh%t. Doesn't count as being submitted. It's equivalent to the wind blew my paper away. Too bad. Next time make a copy.

AnonymousJune 10, 2009 1:16 AM

"This is the dumbest thing I've read all week. Whether in school or at work, if you don't deliver the product it's on you. Period. Corrupted file ? Who gives a sh%t. Doesn't count as being submitted. It's equivalent to the wind blew my paper away. Too bad. Next time make a copy."

What good does a copy do if your file was corrupted in transmission? Unless you expect every student to submit some large number of copies of each of their assignments, they aren't going to know to resubmit that copy until some time after they initially submitted it (presumably near the deadline).

Anonymous DudeJune 10, 2009 1:30 AM

Here's an old fashion trick I saw someone do in college:

If the assignment is late and your professor accepts hard copies, drop it off in the box next to the one that belongs to the professor. Your professor will/might ping you for the paper later and you can say that you've dropped it on such and such date. You can even hint that maybe you've dropped it off in the wrong place by mistake. Doesn't always work but if you're desperate, worth a shot.

CodexonJune 10, 2009 2:00 AM

Interesting and sad. These students should learn to corrupt their files by themselves!

Also, everyone should just turn in reports as ascii text files. Forget fancy formats.

Tim LeggeJune 10, 2009 6:58 AM

Missing the obvious. How can I deliver exploits to unsuspecting students and professors?

Charge 3.95 for a "corrupted file" that includes an exploit to some Office bug. The student expects it not to work so seeing an error or "unexpected results" is not unexpected.

JohnJune 10, 2009 8:29 AM

Here is a pre-computer instance.

In the early 60s I had to take on more courses than comfortable because the next year my GI bill would expire and I needed that many units to graduate. Most of these courses required several papers during the semester.

One prof turned out to be a stickler for getting papers in on time. It was a large lecture class on philosophy in modern literature. A fascinating class, some 200 students in a large lecture hall. Actual interaction for a student was with the TAs. Anyhow, in the first lecture the prof warned us that we would get an automatic F if any paper was turned in late. No excuses would be accepted.

Normally, I would have dropped the course, but I was constrained by my GI Bill running out soon.

My roommate had this prof previously. He tried to get a delay from the prof, but the prof wouldn't grant it, and the result was an F.

I realized from my roommate's story that his problem was "burden of proof." Asking the prof for a delay meant an unmeetable burden of proof that the delay was merited. How, I asked myself, could I turn that around so that the burden of proof was on the prof to justify not granting the delay?

I was going to do the paper, but I could not physically make it by the due date.

Then I hit on it.

When the paper was due, I said nothing. Remember, it was a huge lecture class. There was no way to prove I didn't hand it in. So I said nothing.

In the meantime, I got caught up in my other class work and completed the late paper. I kept it handwritten on legal pad. I didn't type it up as I normally did.

I waited until the papers were returned by the TAs, again in the huge lecture class. Then I calmly asked the TA "Where is my paper?" He looked all over for it. He turned his briefcase inside out. He asked the other TA if he had my paper. The other TA searched everywhere. Finally, the first TA asked me what my paper was about. I told him. "Oh, yeah" he said, "I remember reading it."

I told him I still had the handwritten copy and didn't have the time to type it up. He could have that. He accepted, and I got a B+.

It took me a long time to figure out why he thought he read it when I hadn't turned it in. Basically, there is only so much an undergraduate can say on Hermann Hesse in a school paper, and after a couple of semesters TAing the course, you've read them all.

It took nerve on my part, but it worked.

PS. My wife will not let me tell this story in public.

sidelobeJune 10, 2009 12:06 PM

This is one case where encryption might actually harm the system.

If you were to require the students to encrypt and sign their files, then all of the detection remedies suggested here would fail. The student would simply submit an appropriately-sized file of random bits and claim an encryption or transmission error. Nobody could prove him wrong.

Pat CahalanJune 10, 2009 8:14 PM

@ Anonymous

If you really think an extra hour or day's worth of time is going to change your ability to execute a decent deliverable... well heh.

If you're a "write it the night before" kinda guy, you're going to produce a pretty mediocre paper even if you get an extra day or a few extra hours. I know, I used to be one of those guys ;)

That said, iff'n I get around to finishing this Ph.D. *and* get a tenure track position somewhere, you can bet I'm going to be pretty hardass about meeting deadlines... and by "meeting deadlines", I mean "You get your work to me in legible format by the time specified, or you don't get full credit, period."

Now, if you have a legitimate *sounding* excuse, I'll give you the opportunity to earn back said credit, but the extra credit assignment is going to be proportionate to how credible your excuse sounds :)

AntonJune 10, 2009 8:40 PM

Who needs to buy these corrupted files. I have collected oodles of them over the years.

Chris ThompsonJune 11, 2009 12:30 AM

For the CSci department at my university, our online submission tool (which most classes use for source code and project submissions, etc.) gives both a post-upload report, as well as an e-mailed report to the student submitting containing the MD5 hashes of every file included in the tar or zip file uploaded. This way, the burden is on the student to verify that everything was uploaded correctly to the submission servers, as soon as they have uploaded it (and thus have a time-stamp of submission).

I've never had a file corrupt on me when submitting, but I damn well make sure everything went through properly before moving on to something else.

AcademicJune 11, 2009 12:40 AM

Academic here. If its corrupt you get zero. You should have checked it before upload.

MatthewJune 11, 2009 3:44 AM

I'm not sure how a modern word file would compare, but I did I my honors thesis (call it a report, if you will) using word (v4 I think) on a mac (ah, the days of 16 bits). Two days before the cut-off, my one copy (yes, I've learned) was corrupted by a crash during a save.

Back in those days, it was mostly a matter of opening it in emacs and pulling out the paragraphs - occasionally a bit of rework to put edits back in their rightful spots, then pasting the images back in. It took me a couple of hours (on a PC, running linux - eight meg and constantly swapping).

A bit of quick testing tells me that modern word files aren't very different. Any student who tried that stunt with me would get a surprise ... I'd read the thing anyway.

I only ever marked one student assignment - I was too slow, so it wasn't worth my while, but I was neurotically thorough :-)

RookJune 11, 2009 12:55 PM

A technological solution (probably school-IT-dept supplied for most teachers would be appropriate) might be to have a paper-submission-system that a student can log into a view a rendering of the submitted paper from whatever select list of formats is excepted (pdf, doc, txt, etc): if that's fine, then the student has met his burden. This is something like the MD5 hash solution mentioned earlier but it doesn't rely on non-CompSci teachers/students understanding the concept and it also tests that the file can be reasonably rendered by a reader, so if the prof is using some different version of MS Office say, and says the file is corrupt, he or she could also login and view the rendering and see that its his problem and you've met your burden. Some other checks/balances could be added (e.g. a "receipt" in case the IT dept loses the paper that its supposed to supply the prof), but I think this type of system could cover a lot of scenarios reasonably.

RookJune 11, 2009 1:14 PM

To add my own variation of the stories about how Prof handle "corruption":

For a seminar (18 student) Greek Civilization class, the first paper, as a joke, I took page 1, converted the body text to use the Greek font, and put that as the first page. Turning the page revealed the full, normal paper including the normal top title sheet and all.

Well, the prof never turned the page. He just asked me in an annoyed way at the next class meeting in front of everyone what the heck was up with my gibberish paper. I asked if he'd turned the page, it was all there. I explained my attempt at humor. He told me to submit a new copy anyway. He was clearly embarrassed that he had given it such a cursory look before bringing it up accusatorily in front of the whole class. So that made him angry at me. So I got my grade dinged and he remained a little bit standoffish from me from that point on.

MattJune 11, 2009 4:17 PM

Here's a simple solution to intentionally corrupted files that worked when I was in college from 1998-2005...

You submit a paper via the course website. You are then given a link to download the paper you submitted. There you go.

Another GooglebotJune 11, 2009 5:23 PM

As long as we are sharing, here's my story of similar shenanigans.

The programming courses in the EE program at the University of Hawaii (Manoa) were all conducted on a unix system. Someone within the department had put together a set of scripts to automate submission and evaluation of student assignments. Consequently we were required to use a submission tool that expected source code and a makefile with filenames that conformed to a set of rules.

As all of the scripts were world readable, I learned that the system determined if an assignment was late by looking at the timestamps on the submitted files only after compiling and running the submitted source code.

So when that inevitable day came where I was late turning in an assignment I spent a few extra hours to come up with a self-modifying makefile that first "touched" everything in the current directory to a timestamp before the submission deadline, built the executable as expected and then rewrote itself to remove all traces of my devious plot.

I got an A on the assignment with no comment. Although quite pleased with myself, I sometimes wonder if I got that A because I was successful in manipulating the system or if I had been caught but the prof had been impressed enough with the schemings of a freshman to let it slide.

MikePJune 14, 2009 11:01 AM

Or remove the incentive to cheat.

There's any number of ways to do that, none of which would be acceptable in academia because it makes too much money keeping things as they are. Sound like anybody else readers are likely to be familiar with?

It's not all economics, of course, but... look what happened to Denis Rancourt.

ColinRJune 15, 2009 5:34 AM

I am surprised no one has commented on the security implications of this.

If an attacker finds a weakness in a document viewer (for example unchecked array bounds), they may then use this to construct a document to exploit the weakness. While in some viewers it may be correctly handled as "a corrupt document", while in other less robust viewers may well be a launch point for a security attack.

So be very wary of where you get you corrupt documents from and the motive of the person corrupting it.

Leave a comment

Allowed HTML: <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre>

Photo of Bruce Schneier by Per Ervland.

Schneier on Security is a personal website. Opinions expressed are not necessarily those of Co3 Systems, Inc..