Finding bugs

Peter worked in QA (Quality Assurance–finding bugs in software) for a while in the mid- to late-’80s. He always said that he had a talent for finding bugs.

I think it was while he was working for Mead, testing LexisNexis, that he told us about one of his bug-finding techniques: lift both hands in the air, and drop them onto the computer keyboard, pressing a bunch of random keys at once. Apparently in a fair number of cases, whatever software he was testing would immediately crash.

If I recall correctly, his manager at the time felt that that wasn’t a severe enough bug to be worth fixing, because it was unlikely to come up in real use. I guess this was before the days when cats and babies regularly had their way with keyboards.

2 Responses to “Finding bugs”

  1. Chaos

    Heh. People use tricks like this nowadays to look for security-relevant bugs — if the software crashes when it gets random input, it might contain a buffer overflow or other such exploitable thing. The class of techniques has a name, fuzzing, and seems to have gotten popular in the past few years. Mostly, people randomly generate input based on a regexp, rather than banging on the keyboard, but it’s the same idea.

  2. Dobe

    I watched Peter carefully around computers through the years. His own, mine, and a few times, at his worksite.

    Peter as notorious for crashing computers. I personally have recollections of having my computer working fine for days and weeks (I typically never turn it off), and then Peter walks in and the PC starts to misbehave or crash altogether. This was nearly predictable.

    (I, on the other hand, have always approached computers with a screwdriver, and they start working more reliably, whether I’ve done anything or not!)

    Peter meticulously documented bugs and then ranted that most of them were never fixed. In the next version, many of the bugs would still be there. So he would test everything he’d found before and usually be rewarded with the original and new errors.

    Although he was good at finding bugs, I always had the feeling it was one of the reasons Peter “burned out” in the computer industry. It’s like doctors or nurses that see tragedy every day and then one day, just can’t go in anymore.

    It’s like he wasn’t paid attention to (no respect, dissed), in that if there were a diminishing number of bugs, over time, an end in sight, he would have been happy to look for the “next-to-last” one, but he saw ineptitude and apathy around him for so many years that he left before finally finding his calling in teaching, instead of bug-catching.

    More rewarding…


Join the Conversation