Some stories
Taken from Peter G. Neumann, Computer-Related Risks
Also see the "risks" archive and forums
http://www.csl.sri.com/users/neumann/neumann.html
Usenet: comp.risks
ACM SEN (Software Engineering Notes)
Therac-25 (C-RR, pg. 68)
Well-known case of software failure (1985-1987)
Radiation therapy machine
Two modes of operation (direct radiation, indirect)
Software replaced hardware interlocks
User-interface defects permitted unsafe combinations
Hardware interlocks had been replaced by software
More than software failure
Operators ignored/minimized patient complaints
Excessive faith in computer-based systems
Result
At least three deaths
Numerous injuries and complications
AT&T long distance switching system (pg. 14)
Daylight saving time (pg. 89)
Patriot clock drift (pg. 34)