Re:Things I Believe About Software Engineering
by jerome-jh on Monday February 03, 2020 @05:31AM
(#22222969)
It is said that the Apollo 11 lunar lander computer rebooted twice during the final approach. We could probably not say it was correct, but still it was robust. "for any meaningful definition of correct" sounds a bit to strong as an hypothesis. Maybe we could rephrase: for all piece of software S, there exist a definition of correctness for which S is incorrect. Timings in particular are very difficult to ensure.
[ Parent
| on HN ]
Re:Things I Believe About Software Engineering
by fjfaase on Monday February 03, 2020 @06:57AM
(#22223409)
I did not reboot. (I think it did not have the capability to reboot itself.) The program alarms indicated "executive overflows", meaning the guidance computer could not complete all its tasks in real time and had to postpone some of them.
https://en.wikipedia.org/wiki/Apollo_11#Lunar_descent See also http://klabs.org/history/apollo_11_alarms/eyles_2004/eyles_2... for a detailed explanation.