Boing boing boing into the ground.

[Guilaumme Faury]

At one time, Boeing knew how to build airplanes.

[gfretwell]

There is nothing wrong with the plane, they have a software problem. It isn't the bean counters, it is those kids in sneakers who are writing the bugware.

BTW a number of pilots have said if you know how to turn off the MCAS (put in a little flap), you just land the plane and everyone makes their connection.

[rbowman]

I wonder if the poor bastards are using Ada?

[trader_4]

That's a good one two. And it's not really a software problem, it's a DESIGN problem. The programmers didn't make a mistake, the program does what they were told to make it do. If you bought a TV that didn't work, that you could not use because of a "software problem", would you say there was nothing wrong with the TV?

Totally wrong. There is ZERO evidence that I have seen, that this was a problem created by the programmers. The program did what it was designed to do.

Not any responsible pilots that understand the problem. The pilots in the Ethiopian crash did exactly that. The co-pilots last words were that he could not trim the plane manually. Also, "put in a little flap" was NOT the procedure that Boeing told pilots to use to recover. And that was AFTER the first crash, after they had more than a week to think about it. And it surely was not the programmers that issued that directive either.

[gfretwell]

Just trying to put the blame where it belongs. It wasn't the air frame that failed, just the software. I am sure that problem was fixed in a couple of days and now they have been picking that plane apart piece by piece for a year and when they finally say OK, it will be the safest plane in the air. In fact better pilots were not having the problem with it in the first place. It is no coincidence that the crashes were 3d world airlines and mediocre pilots with minimal flight hours compared to your average

1st world airline pilot.

[gfretwell]

So I guess you have the specs the programmers were working with. Please give us a link so we can all see.

Some programmer decided when MCAS was supposed to be operating how it was monitored. My bet, they just gave them broad parameters of what sensors were available and what actuators they had to use. They also decided to only use one AOA sensor when two were available and not to alert the pilot when they did not agree. The latter actually being the fatal flaw. My guess is the programmers don't know shit about airplanes, they just write video game type software and do not take into account dealing with hardware failures.

As for my TV. If the Netflix function takes a shit and everything else works fine, I would blame the software, not the hardware. Maybe I just have more experience in this arena tho.

The pilot told the co-pilot to stfu too. He was too arrogant to save his own life. Maybe if the co-pilot was more than a recently graduated student the pilot would have paid more attention to him.

[trader_4]

Again, the software did not fail. The software performed as it was designed to perform. A software failure would be some programming error, some coding mistake, that caused the software to do something other than it was designed to do.

I am sure that problem was fixed in a

If it was fixed in a couple days, the second one would not have crashed six months later.

and now they have been picking that plane apart piece

I would not count on that.

That's all wrong too. It was pure coincidence. This problem, the AOA sensor failing, only occurred with two planes on three flights on Lion Air and Ethiopia flights. It never happened to any other airline, nor is there any reason to believe the outcomes would have been any different. In fact, the second crash, the co-pilot with just 200 hours is the one that correctly followed the Boeing procedure that Boeing put out after the Lion Air crash to deal with the problem. His last words were that he could not move the trim wheel.

[trader_4]

If you followed this at all, it's been very clear that Boeing, FAA and all involved have made it clear that it was a DESIGN issue, not a coding problem, not a programmer's error. The programmers did not come up with MCAS, aeronautical engineers did. The programmers did not decide that it should rely on just one AOA sensor. The programmers did not decide how much trim authority the MCAS add-on should apply, in what increments, what max amount. They didn't decide what indicators should be in the cockpit and whether they should be standard equipment or a $10K extra option. And they didn't decide to increase the authority MCAS had after flight testing and to not inform the FAA that they had increased the amount of trim that MCAS could put on by a factor of three.

BS. Programmers don't make those decisions, aeronautical engineers, test pilots, system designers do. They create the system, describe how it's to operate, what the parameters are, then software writers code to it.

My bet, they just gave them broad parameters of what

That's incredibly silly. Programmers writing code don't understand how the plane flies or what is or isn't required.

They also

IDK what your experience with what programmers do has been. In my experience, in the logical world, those decisions are made by engineers with experience in the actual underlying system, not programmers. They lay out what the system is, what elements it uses, what it has to control, how much authority it has, how it operates. The software coders then implement that. The FAA also checks the system and approves it.

That's beyond silly too. Programmers don't decide what indicators or warning lights go into the cockpit. Aeronautical engineers and system designers determine that. And the planes had a AOA disagree light OPTION, if airlines wanted to pay for it. And no, the software writers didn't decide it should be a $10K option either. Boeing also knew a year before the crash, that even if an airline ordered that option, it did not work! That was the one and only error that is likely attributable to the software writers. But it's not the only error, someone failed again at system validation, making sure it actually worked. And Boeing management, not the programmers, decided to keep secret the fact that if you ordered the disagree light as an option, it did not work secret instead of alerting operators. They planned on fixing it one day with a software update. That date never came. There's also no evidence that a working disagree light would have made any difference, the pilots didn't even know MCAS was there, much less how it operated.

My guess is the programmers

Yet you think they decide what indicators go into the cockpit, what sensors are used, what authority over the aircraft systems have, etc. Go figure.

Nice try at moving the goal posts. The issue wasn't whether it was hardware or software that was the problem with the Max. The issue was you stated there was nothing wrong with the airplane, only the software. So, again, if part of your TV doesn't work, if software can cause it to catch fire, burn down the house and kill people, would you say there is "nothing wrong with the TV"?

Cite for that please. Not that it has anything to do with the issue though. The simple fact is that the co-pilot of the Ethiopian crash correctly identified the problem and followed the exact procedure that Boeing put out after the Lion Air crash. It didn't work, because it was impossible to move the mechanical trim wheels. Which is a problem that Boeing has known about on all 737s since the 60s. In fact, they used to disclose that in the manuals and outline a procedure that involves putting the plane into a dive to relieve forces enough so that the wheel can be turned. Later they removed that from subsequent manuals and it disappeared. It wouldn't have mattered in the Ethiopian crahs, the plane did not have sufficient altitude.

ROFL

Another good one, you're on a roll recently. Blame the co-pilot who had it correctly identified and did exactly what Boeing said to do. And it's BS, because the pilot didn't do anything to stop the co-pilot,just more Fretwell fiction.

[Scott Lurndal]

The primary factor here was the need for a more efficient engine in order to reach the per-pax-per-flight cost reduction targets. The

737 (design started in 1964) was specifically designed to use unimproved airfields (i.e. airstairs) so it was designed very close to the ground.

This worked well with the engines of the day. The next generation (NG) models (-700, -800) used a larger fan diameter for efficiency, but needed to flatten the bottom of the cowl in order for the engine to fit under the wing.

The MAX (-7, -8) needed an even larger fan (and bypass ratio) to achieve their fuel efficiency goals (the main reason a customer would buy a MAX over an NG), unfortunately, there was no room left under the wing and a redesign of the undercarriage would push the schedule out and raise the development cost. The engineers moved the engine forward on the pylon moving the fan ahead of the wing. This changed the 'pitch moment' of the airplane (moving the engine forward made it more likely that the engine would cause the airplane to pitch up (nose up)). To ameliorate the unsafe pitch moment they implemented the MCAS subsystem, which will force the pitch back down under certain conditions (and of course, the mechanism wasn't designed with appropriate fail-safe; only one of the two angle of attack (aoa) sensors was used by the MCAS system.

Note that this design was a direct result of the refusal of management to do the right thing and either redesign the undercarriage and associated structural elements (which would mean they'd need a new production line for MAX instead of sharing with -NG) or do a clean-sheet design to replace the 737 (which is what they should have done, but they had a lot of pressure from the A321LR at the high-end of the NG product range and the MAX -8 was intended to be the short term competition for the A321LR until the NMA (New Midmarket Aircraft - a replacement for the 757) hit the market (it's been delayed yet again, as has the 777X).

Granted, there was incompetence all the way up and down the chain.

[trader_4]

I think what you're saying there is that you agree it was not a programming error, a coding error, that the software did what it was designed to do, it implemented MCAS as the aeronautical folks designed it to operate.

:)

[Ed Pawlowski]

For those that care, Boeing, not Boing.

Air frame is a factor as they had to move the engines and change the balance. Software is compensating for physical characteristics of flight.

Seems that some shortcuts were taken in pilot training also. Overall it has become a shitshow with some push back from flight attendants that may refuse to fly on it.

[gfretwell]

Maybe you should read a little more about this

OUT

[Clare Snyder]

It's the bean counters that made the second detector optional and put kids in sneakers in charge of critical software. It's bean counters that pushed the decision to not require ransitional training from the 727 to the MAX. It was, to a large extent, beancounters that drove the design to the point MCAS was required in the first place.

[Clare Snyder]

If the beanies said "we need to cut the base price, so only put in one AOA sensor - make a second one an extra cost option" the code kiddies design around that parameter. Can't blame the code kiddies for that. The engineers have to have the balls to stand up to the engineers for a change and say "not on my watch - cut corners somewhere else - redundancy is not on he table".

[gfretwell]

Both sensors are there. I doubt some suit in Seattle was involved at that level. It was just a few more lines of code to compare the sensors and turn on a light if they disagreed. In fact that was an option. Those 3d world suits decided not to buy it. I agree the training was lacking but it seems the 3d world airlines had all the trouble, in spite of flying the max a fraction of what airlines like Southwest were flying.

That may be true but we don't know what really happened and Boeing isn't talking. It is clear from the leaks that there were engineers who were monday morning quarter backing tho.

[rbowman]

Boing Boing never crashed a plane, afaik.

But then they never tried to build a plane.

[trader_4]

Good one. When you have a link to anything that shows anything I have stated here is incorrect, let me know.

[trader_4]

Correct, but there is no reason to believe having the disagree light would have made any difference. Remember, there was no explanation of what MCAS was, how it works in the aircraft manuals. The only mention of it that could be found was in the glossary of terms, where MCAS was listed.

Which again is totally unfair. The AOA sensors only failed on two aircraft, LionAir and Ethiopian Air. We don't know that the outcome would have been any different if it had happened to SW or American. We do know that the Ethiopian airline crew followed the exact procedure that Boeing recommended after the Lion Air crash and that it did not work. The co-pilots last words were that he was unable to move the trim wheel by hand. Those are the facts, not speculation. Boeing even screwed up the procedure to recover that they put in place after the first crash. There is obviously a whole lot of incompetence at Boeing.

But Boeing has gotten something right, at least so far, the new 777-X made it's first flight a few hours ago.

[Ed Pawlowski]

Looks nice, I hope they do better than the Max. I've only ever piloted single engine stuff, would love to just take off, circle and land one of those. There is a place not too far from me where I could get simulator time.

[gfretwell]

That is certainly what lead to two crashes but I haven't really heard anyone say those were the only two AOA sensor failures.