When Bad Software Can Kill

bhoman writes "A wrist computer that tracks and calculates safe diving times and limits for SCUBA divers had a dangerous software bug that may have been covered up by company executives. This SF Chronicle Article details the problem, product, company, and some of the lawsuits. According to the Chron article, company execs tried to cover up and deny the problem for years, but their official website makes it look like they did a voluntary recall."

Comments

The main flaw of modern computer science.

[Krapangor]

Crappy coding style is caused by the main flaw of modern computer science: the failure to recognize the problem of correctness as a mathematical problem.
Everybody who has an university CS degree will agree with me that much time and effort is spend to encourage students to produce nice and correct programs. However this strategy is a failure so far. Again and again bugs, errors and other problems turn up more often these days in spite of the increased educational efforts.
This is because the CS community failed to accept the core of the problem: error in programs are a mathematical problem which must be attacked by mathematical methods. All modern approaches to correct code are indeed management orientated - take programming by contract, extreme programming etc.
But what does it mean that a program fails to execute correctly ?
It means that mapping induced by the program in the trajectory space doesn't agree with mapping induced by the specification. And that is a purely mathematical problem, ladies and gentlemen. The question if two mapping coincide is a basic mathematical question (the equivalence problem) which even dates back to Euclid and Platon.
So instead of throwing more and more management rubbish at poor CS graduates, people should analyze the mathematical structure of the problem and find there the answers they seek.

I think that it's very sad that CS people still ignore this issue and stick to their old established ways. Sometimes I believe this is not motivated by scientific arguments but a rather psychological inferiority complex: as mathematician have the reputation to be smart while CS people only count as code nerds, computer scientists tend to despise most mathematical approaches as "too academic" or "imfeasible".