Originally Posted by
Icarus2001
Technically, the MCAS software did not fail. But from an overall systems standpoint, it did. The software people may get off the hook for writing code that ran per the specification. But the spec missed an important failure mode. And was written for the wrong level of system criticality.
When I worked at Boeing, we had a number of instances where the hardware people finished their design (or were saddled with older systems for which a derivative needed to be produced given a limited budget and schedule). And once finished it got "thrown over the wall" to the software group. in part because hardware and software were (and still may be) viewed as two completely different kinds of design domains. It was up to the coders to develop something that 'worked', given the hardware peoples' attitude that remaining shortcomings could always be fixed in software.
It's a cultural problem (and not just at Boeing) that has always been way above my pay grade to fix. Even if I could see it.