Post date: Mar 11, 2019 2:2:30 PM
"In the Airbus disaster, the computer's response to an 'error' event was to abandon control, which caused a stall from which the pilots never recovered. ... Stopping on an execution error is a bad idea. It's a bad policy. The alternative is a bad policy too. ... We need to handle exceptions in a humane way."
−William Kahan, at the IFIP Working Conference on Uncertainty Quantification in Scientific Computing, at NIST in Boulder, Colorado
3 August 2011
"The Adriane 5 disaster was blamed on inadequately tested software. But what I want to know is where is the adequately tested software?"
−William Kahan, at the IFIP Working Conference on Uncertainty Quantification in Scientific Computing, at NIST in Boulder, Colorado
3 August 2011