I think one of my most simple - yet evil - prank was to LD_PRELOAD a modified `malloc()` on a colleague's computer. Except where the common practice was usually to have it blow up the memory or not return enough, I had mine to work mostly correctly: in most cases it would allocate as requested, and in occasional cases it would allocate slightly less than requested, and in rare cases it would blow up the memory.
You could think I shouldn't be proud of that one, considering some classmates probably went crazy wondering why their systems behaved erratically, and that it probably didn't help with some of their assignments. Generally our pranks were rather tame at school, and if I recall I had reserved that particular one to only 2 guys that were extremely d-ckish. Can't even recall their names now.
Other pranks were your usual stuff: switch keyboard mapping to Dvorak, swap LTR to RTL, randomly modified clocktime by a few minutes (also rather mean when you need to handover assignments, I suppose...), bind some keys to the most annoying shortcuts possible, unsecured xdisplay accesses, open cdroom on ssh-accessible machines... Basic stuff.
I also grabbed all the login/passwords for an entire promotion once. To my defence, I didn't exactly use the accounts for anything else but to change the default passwords (so, technically and legally, I did access the accounts, I suppose). I was just making a point to IT that assigning default passwords with guessable sequences (if I recall, major + year of promotion) was a bad idea and that for some students it could take weeks before they'd change it and leave some accounts open for abuse (e.g. for students who dropped, were sick on the first days, or would simply not use the labs that much). They weren't pleased by the surge of people contacting them to ask why their passwords didn't work.
When scientific distributed computing came rather prominent, I also clustered an entire classroom (didn't want to hit the whole school network, only rooms that were underutilized). That got noticed by an admin though, but he didn't know what it was and I just said this was one of my projets for graphics computing (which was indeed a real project for which distributed comp was authorized).
I think one of my most simple - yet evil - prank was to LD_PRELOAD a modified `malloc()` on a colleague's computer. Except where the common practice was usually to have it blow up the memory or not return enough, I had mine to work mostly correctly: in most cases it would allocate as requested, and in occasional cases it would allocate slightly less than requested, and in rare cases it would blow up the memory.
You could think I shouldn't be proud of that one, considering some classmates probably went crazy wondering why their systems behaved erratically, and that it probably didn't help with some of their assignments. Generally our pranks were rather tame at school, and if I recall I had reserved that particular one to only 2 guys that were extremely d-ckish. Can't even recall their names now.
Other pranks were your usual stuff: switch keyboard mapping to Dvorak, swap LTR to RTL, randomly modified clocktime by a few minutes (also rather mean when you need to handover assignments, I suppose...), bind some keys to the most annoying shortcuts possible, unsecured xdisplay accesses, open cdroom on ssh-accessible machines... Basic stuff.
I also grabbed all the login/passwords for an entire promotion once. To my defence, I didn't exactly use the accounts for anything else but to change the default passwords (so, technically and legally, I did access the accounts, I suppose). I was just making a point to IT that assigning default passwords with guessable sequences (if I recall, major + year of promotion) was a bad idea and that for some students it could take weeks before they'd change it and leave some accounts open for abuse (e.g. for students who dropped, were sick on the first days, or would simply not use the labs that much). They weren't pleased by the surge of people contacting them to ask why their passwords didn't work.
When scientific distributed computing came rather prominent, I also clustered an entire classroom (didn't want to hit the whole school network, only rooms that were underutilized). That got noticed by an admin though, but he didn't know what it was and I just said this was one of my projets for graphics computing (which was indeed a real project for which distributed comp was authorized).
I miss these lab setups.