[WalterBright]

> You'd get no new broken airplane designs - which is the point of the exercise.

Nobody seems able to write software that doesn't have bugs in it, in fact it seems to be humanly impossible. So, if you're going to jail people for buggy software, no new software will get written.

The same goes for airplanes.

[jacquesm]

Buggy software will happen. Buggy software that isn't reported when it should be, that isn't fixed, reviewed and tested shouldn't happen. Software development for aviation is pretty good, that's not where the issue lies here. The issue is a systematic attempt to circumvent proper type certification for a new aircraft, and downright trickery in order to pretend it's just a minor change to an existing aircraft when in fact the aircraft was substantially different from the previous one sharing the same type number. That's not an innocent mistake.

[WalterBright]

It is a minor change. There is nothing inherently wrong with adjusting the flight controls so one airplane flies like another. In fact, engineers try to do it, as it reduces both costs and accidents. (There have been many aircraft accidents where the pilot did the right thing for airplane X, but unfortunately was flying airplane Y. These happen even when the pilot is experienced and properly trained to fly Y.)

> Software development for aviation is pretty good, that's not where the issue lies here.

MCAS is a software system, and the rules programmed into it had human factors mistakes in it.

You might want to read Aviation Week, 19-Aug, for more detailed information.