[brainbrane]

About 15 years ago, my school's electrical engineering lab had a fleet of HP-UX boxen that were configured by default to dump huge core files all over the NFS shares whenever programs crashed. Two weeks before the end of the semester a junior lab assistant noticed all the core files eating a huge chunk of shared disk space and decided to slap together a script to recursively delete files named "core" in all the students' directories.

After flinging together a recursive delete command that he thought would maybe work, he fired it off with sudo at 9:00pm just before heading out for the night. The next morning everyone discovered that all their work over the semester had been summarily blown away.

No problem, we could just restore from backups, right? Oh, well, there was just one minor problem. The backup system had been broken since before the start of the semester. And nobody prioritized fixing it.

Created quite the scenario for professors who were suddenly confronted with the entire class not having any code for their final projects.

They talked about firing the kid who wrote and ran the script. I was asking why the head of I.T. wasn't on the chopping block for failing to prioritize a working backup system.