Human error and computer error

Flight path of Air France Flight 447 on 31 May...

Image via Wikipedia

Computer errors are almost always human errors – badly written software. The arguable exception would be the rare (?) cases where cosmic rays or other natural radiation caused a bit to flip.

But computer errors are generally seen as something which arises out of some spirit of the machine – see 2001 for perhaps the most famous and much copied example.

But as this chilling account of the crash of Air France flight 447 in June 2009, shows, humans and not machines are the real dangers.

3 thoughts on “Human error and computer error

  1. As a grad student, I experienced a bit-flip once on a computer with a BASIC interpreter in ROM. It powered up one day with the values of true and false reversed, which did not do wonders for flow control in my program. I documented it by printing statements like”7<5" (true), then cycled power off and back on, after which my program ran fine.

    • In ROM? Oh dear. Presumably it loaded the interpreter into RAM for use or similar.
      I wonder are there other sorts of computer errors that are not human errors? I could not think of any other than this sort of bit garbling.

      • Back when RS232C ruled PC ports, I encountered situations where, immediately after power-up of both PC and printer, a printer simply could not talk to the PC. Powering down and back up usually cured it. It seemed that a transient voltage was screwing up something, but I couldn’t tell whether it was a bit flip or a lingering electrical charge causing a false signal on one of the pins.

Comments are closed.