6 ms·
I found something like this at my school. The administration reacted similarly. But fortunately, I was taking djb's Unix Security Holes at the time, and a har
by jrockway 14y ago
I found something like this at my school. The administration reacted similarly. But fortunately, I was taking djb's Unix Security Holes at the time, and a harshly-worded note from djb to the Computer Center folks ended up getting me a thank you.
Next semester, though, I refused to sign the new AUP (which included a clause allowing the computer center staff to seize any computer I was using, even at my off-campus home), and they kicked me out of school. (Actually what happened was they locked my course registration account, and wouldn't reinstate it until I signed the policy in their presence. I refused.)
(Sadly, I can't find the full-disclosure thread for this bug. I guess I posted it to my blog, which I deleted after being threatened by school administrators. Oh well. That was 9 years ago!)
- tptacek 14y agoDid you pass that course?
- jrockway 14y agoI got a B. The homework was to find and write an exploit for 10 security holes in deployed software, but I only found 2. (3 including the one above, which I must have found the week or so after exams. The holes I found were in nasm and in some amateur open-source smtpd.) FWIW, the exams are quite thought-provoking nearly 10 years later, here's a link to them: http://cr.yp.to/2004-494.html http://cr.yp.to/2004-494.html
- Tichy 14y agoSounds like a very cool course.
- amenonsen 14y agoWhat did you think of the course textbook ("Exploiting Software", Hoglund & McGraw)? Is there a more modern alternative that you (or anyone) can recommend?
- jrockway 14y agoThe textbook covers subject matter that won't become outdated: reverse engineering, how to craft malicious input, etc.
- tptacek 14y ago_The Art Of Software Security Assessment_ is the current canonical text. Here's a reading list; I'd add Zalewsky's _Tangled Web_ to it, but change little else: http://amzn.to/cthr46 http://amzn.to/cthr46
- tptacek 14y agoDidn't he famously fail the whole class one of the times he gave it?
- jrockway 14y agoI don't think so, as he's only taught the class once and I didn't fail it: http://cr.yp.to/courses.html http://cr.yp.to/courses.html
- tptacek 14y agoThere was a Slashdot story about it. I remember reading the course syllabus online and being jealous despite already having worked in professional vulnerability research for a few years. You're lucky to have been at the class! Was he a good lecturer?
- monsterix 14y agoReading stories/incidents like these makes me believe that education as a whole is stapled for reinvention. As they say: competition doesn't kill your business; attitude kills it.
- Ras_ 14y agoThese expulsion stories sound really weird. I mean you pay for all of your studies and still could get axed on a whim? Whereas in my country I get paid to study and have zero chance of being expelled for these kinds of events.
- monsterix 14y agoWhich country may I ask? Nordic?
- Ras_ 14y agoYes, Finland. Maybe it's because all of our schools are public? For example higher ed. providers are funded based on enrollment and rate of graduation. If someone does not graduate, significant chunk (20-30%) of money won't be paid at all. This creates some incentive for the institution to actually guide and see that people don't fall through all kinds of cracks. I guess it's necessary when there is no ordinary paying customer relationship involved.
- monsterix 14y ago> Maybe it's because all of our schools are public? There are Asian countries where this model has failed. Perhaps because of population pressure or other social factors. But I truly like the Nordic way of life.
- jeltz 14y agoCould be low corruption. The nordic countries usually score low on corruption comparisons. High corruption can make almost any system break.
- mjn 14y agoMy experience in the U.S. is that public universities aren't much different from private universities (at least the nonprofit ones) on these kinds of policies. They might be better in other respects, such as lower tuition, but they're run by similar kinds of administrators. Often literally the same administrators: there's a lot of churn as people hop between institutions. The main problem, in my view, is the professionalization of this institution-hopping class of university administrators. It used to be made up of senior faculty who got promoted to Dean, but now it's made up of an entirely separate group of people, often people who come from business management backgrounds, and who have little grounding in a particular institution's traditions or culture. They tend to think rather differently, in a more locked-down, policy-driven way, and apply broad "best practices" without much regard for how things are done in a particular place. Universities end up getting managed like a corporation, with similar kinds of policies. Things are a bit better at small colleges (Rose-Hulman, Olin, Harvey Mudd, Wesleyan, Pomona, Colgate, etc.), which typically have much lighter-weight administration and a more pro-student, pro-experimentation attitude, as well as more success in en-culturating their administrators so they "get" the local culture and work with it. But they don't scale very well (I say this despite having gone to one and being a big fan of the undergraduate-college model).
- Natsu 14y agoI'm wondering these days if you can be any sort of hacker at all without finding some kind of vulnerability in your college's network. For me, it was a way to steal the AFS space of the previous user (basically, they didn't expire the token... oops). I actually found the initial vulnerability by accident (something crashed due to network problems, reconnected and went, "WTF, those aren't my files!"), but I did find a good way to reproduce it on demand (yank Ethernet cord at proper time). Thankfully, I had read enough stories like this way back then and submitted the bug anonymously. This was ~2000 or around then, mind you. I also tried to get university management to switch people over to using SSH way back in 1998, but it was something like 4-5 years before they eventually did so. I'm guessing they had no idea what I was talking about or why it even mattered back then, even though anyone could see everyone's passwords going over the wire with all the people who had to telnet for various reasons. Maybe they assumed that log file they were writing our activity to would catch anybody doing anything weird? It was cleverly named "resugol"--read that backwards if you're confused.