3 ms·
> This is possible the biggest software development blunder in history I hate to be that guy, but please google "therac-25". There's been absolutely no loss o
by PH95VuimJjqBqy 3y ago
> This is possible the biggest software development blunder in history
I hate to be that guy, but please google "therac-25".
There's been absolutely no loss of life, just software shittily written. Yes, it's a blunder, but the hyperbole is unwarranted here.
- zimpenfish 3y ago> I hate to be that guy, but please google "therac-25". Or "UK Post Office Fujitsu".
- DoingIsLearning 3y agoBeyond safety critical applications the Dutch automated welfare-fraud detection system (SyRI) also jumps to mind. Wrongful debt, convictions, children removed from their parents homes, and even suicides.
- taneq 3y agoAustralia's 'Robodebt' system likewise issued a whole lot of unjustified fines etc. and caused a lot of pain for a lot of people.
- multjoy 3y agoThat was, at heart, a disclosure issue. The Post Office knew that their prosecution was unsafe but went ahead with it anyway - they had a duty to disclose to the defence any material in their possession that undermined the prosecution case or advanced the defence case. This is a statutory duty and a police investigation is ongoing into the prosecution.
- taneq 3y agoThe Therac-25 issue apparently only caused six known incidents. If you want an example of software killing people, how about the 737 MAX MCAS issue, or the Toyota unintended acceleration issue?
- DoingIsLearning 3y agoIn the Therac-25 case the number of victims is kind of less relevant. I think the shock factor in Therac-25 is the contrast between the gruesomeness of radiation poisoning and radiation burns and the post-mortem investigation showing the hubris and recklessness of the Software Design team, Risk Management, lack of HW interlocks etc. MCAS deaths were precipited by greed which is bad but not shocking. Therac-25 was an entire group of Design, Quality, Regulatory teams just literally YOLOing into patient deaths.