Hacker News Reader

Comment by lallysingh | Hacker News Reader

It's a revision of a model number. People will forget, if they remembered at all.

But that's the thing, the 737 max isnt a revision of a model number. It has a very different airframe with different flight characteristics from other 737 models.

Boeing used software to override and reinterpret pilot inputs so that the airplane behaves as if it were a 737, in order to cut costs on retraining. When that software fails, the pilots are suddenly flying a totally new airframe that would normally require additional certification with 0 experience.

The idea of there being a layer of software between the pilot's intentions and the aircraft's control input freaks me out.

Doesn't that describe fly by wire systems which describes most planes these days?

Most fly-by-wire systems implement the pilot's stated intention directly, though, is that not correct?

This layer inbetween where the computer goes "actually what I think you REALLY meant to do is..." is novel in the 737 Max.

No. The intended behavior is a nose down nudge every 10s only when a) exceeding a certain angle of attack b) above a certain altitude and c) when flaps are retracted. It's second guessing only at a particular edge of the normal flight envelope to prevent it from departing that envelope.

And all the various layers of protections in fly by wire airplanes do that - they second guess certain inputs to prevent edge cases. But in any case the input is going through software abstraction. Any apparent directness is part of that software behavior.

I'll be interested what the software update due at the end of April entails exactly. I think every airplane should get the angle of attack sensors disagree indication, standard.

I would recommend against boarding any Airbus than.

that's exactly what airbus does... there are a set of "laws" that govern flight at different stages of flight "normal, ground, flight, flare, alternate 1, alternate 2, direct, and mechanical" quite complex systems, that.

https://en.wikipedia.org/wiki/Flight_control_modes

And this is what contributed to the crash of AF 447.

That said, I prefer either the older boeings, with minimal computer interference when hand flying or the Airbus approach. Because Airbus at least appear to know what they are doing.

This Boeing-Airbus chimera isn't really working.

The stories where the computer stops the human from doing something stupid and saving the day never makes it in the news.

An intrinsically dangerous plane that needs either much more careful piloting (hard to sell...) or software assistance (bad idea) should freak you out even more.

Sure. But my point is that the name is easy to forget. It's not a top level brand like 737 or 787. It's a variant. People will write the information off as essentially trivia, because it's just one variant out of "how many?"

Sorry to go on a tangent, but if they were going the redesign the airframe anyway, why didn't just make the plane aerodynamically stable with the bigger engines (make landing gears longer?), instead of lifting the engines and using software to compensate?

Long term public opinion impact would likely be much smaller if the FAA had lead the grounding, not lagged behind: "some sub-variant of a model iteration on hold for investigation" would have been merely a footnote only those directly affected would remember a day later. But this slow-motion wave of country after country changing their minds with ever increasing drama "what will the FAA do!", this is quality news-tainment. I know that hn readership is not representative (the intersection of aviation geekery, software engineering and high stakes UX: golden), but it's quite a bit of a headline elsewhere as well.

Reply on news.ycombinator.com