NO Sh!t, Sherlock!

[Boac]

After two fatal crashes, the FAA has been under fire for certifying Boeing’s 737 Max. The agency’s head says it is examining how humans interact with ever-more automated planes.

Well, good to see someone is bothered?

[barkingmad]

THEY didn’t listen to the grumpy ould gits in the corner, if there were any present at the design stage, so now tragically for the dead and for Boeing they are finally reaping the harvest.

42 years in aviation and I met only 1 pilot who said he’d been invited to the manufacturer’s facility to play with and opine on the latest proposed widget before it progressed further towards acceptance and production.

Shome mishtake here, shurely??!

[Pontius Navigator]

I was invited to BAe once to play with a box before it was accepted. In reality it was so they could have some one that had handled it before it got the front line. It was far to late to change it.

When I finally got my hands on it in the aircraft without BAe hovering over my shoulder we went through the checklist line by line, especially the fault finding, to find the checklist bore no resemblance to the aircraft wiring.

In this case the consequences might have been serious but only cost was the reprint a new set of FRCs and 40,000 lbs of avtur for another associated but unpublished change. We had tested our area but this switch affected several other systems too which affected aircraft pressurisation.

[Fox3WheresMyBanana]

I worked with a small company near Hull on the Computer Based Training aids for the Tornado F3. I would head up there once a month to talk through the Engines and Electrics systems programs. It was quite an education for both parties, with them knowing naff all about engines, and me knowing not much about the latest programming. However, since it was a small company, the three guys I would talk with were the actual programmers and their supervisor so changes, even of the whole approach, could be agreed to on the spot. They were a very conscientious lot, and aviation-mad. It was a big thing for them to have a real pilot come talk to them. The same was true of BAe. My uncle was a senior manager in business accounts, and I once had the time, whilst waiting for a new jet to be rolled out for delivery, to amble up to his office for a cuppa. No pilot in a flying suit had ever been seen in there before, and he got a huge kudos boost in the department as a consequence. He still talks about it today. Plus, all the people who'd built that new jet really did come out, in the rain, to watch it take off. So, I think more visits are a great idea, and I think the vast majority of employees still really care. It's just the sociopathic, greedy a-holes at the top, plus the government bureaucrats, that are the problem.

On my previous squadron, I submitted a proposal to rearrange the captions on the Central Warning Panel. There were 70, and common emergencies like an engine failure would put up anything up to 13 captions in no particular pattern. Experience had shown it was very easy for crews to miss an extra caption, so my rearrangement turned all the standard emergencies into contiguous patterns of captions, so others stood out. Cost was minimal and could have been done at a Minor servicing. What killed it was the mindless bureaucracy. I remember at one point it was held up waiting for the German Navy to consider it, even though it was only proposed for the F3. Then it got binned at the time of the Gulf War as being too expensive in paperwork for amending all the documents, despite its point that it would show up battle damage much better than the existing arrangement.

[Boac]

What did we say? https://www.bloomberg.com/amp/news/arti ... f-777-jets

[Pontius Navigator]

Fox, at Kinloss we had the Nimrod software team with a mix of programmers and engineers. We soon found we could pop in and ask how something worked or could something be done differently. They were very flexible and some simply fixes had benefits far exceeding the original purpose.

Each version update was at half the interval of the preceding one. The last was after just 6 weeks. Great for operational crews, awkward for students.

On the E3 the hugely more complex software required meant a formal review and testing process lasting at least a year though we did manage some 'one-off' changes for 'single use' in the simulator. It meant users could experience proposals before the committee did.

As it happens, this was Boeing software and we got it to do things that Boeing's original either could not do or we bettered.

[Fox3WheresMyBanana]

The ability to "pop in", not to mention the sharing of bar facilities, is not as appreciated as it should be.

[Pontius Navigator]

Fox, as it happens, I popped in to find out the ballistics alogrithm. For an 8k computer minimal syntax was essential. As a result of my popping in, they discovered there were using 3 different routines when one routine and 3 constants would save a few lines.

[Fox3WheresMyBanana]

[Rwy in Sight]

The ability to "pop in", not to mention the sharing of bar facilities, is not as appreciated as it should be.

Too much popping in can be a bad thing.

[Fox3WheresMyBanana]

Agreed.
It's an art!

[Pontius Navigator]

We were also encouraged to visit the Gin Palace to follow up equipment faults (the NBS). I remember once getting there at just the moment 'my box' was being tested and the tech had just failed to find the fault. I realised he was working on the wrong side of the crate. It was an old analog computer and he was working on the N/S side whereas the fault was on the E/W side. It had also been intermittent.

The job label had been written incorrectly.

I described the exact profile conditions for when the fault occurred, moments later he found a chaffed cable loom.

I guess the same with a car. Try and speak directly with the guy doing the work.

[Fox3WheresMyBanana]

Ah yes, NFF = Not F#cking Fixed

And the eternal "Failure occurred at M1.6, 48,000 ft": "Unable to reproduce on ground"

On our numerous detachments, a real effort was made not just to put aircrew in their "own" aeroplane, but ditto with the groundcrew working on "their" aeroplane. It worked really well. The Boss was also keen on getting the groundcrew up on pax rides with "their" pilot in "their" aeroplane. We had remarkable serviceability; it's not common for six or nine aeroplanes to head across Europe and all arrive, yet I do not recall a single case of anyone being stuck halfway.

[Pontius Navigator]

In the case I mentioned the aircraft flew a full sortie every day of the week. I flew Mon, Wed, Fri and the Tue/Thu flight by another crew. When I got the same fault on the Friday sortie I asked what they had done.

Unless a fault could be found after flight they only pulled it off if it had the same fault on two successive flights. I would have filled in the other nav until he said what they had done - quite different profiles from mine. In my case we had done 3 identical profiles.

Back to the Max - unless problems are reported and correlated . . .

[ian16th]

Ah yes, NFF = Not F#cking Fixed

And the eternal "Failure occurred at M1.6, 48,000 ft": "Unable to reproduce on ground"

On our numerous detachments, a real effort was made not just to put aircrew in their "own" aeroplane, but ditto with the groundcrew working on "their" aeroplane. It worked really well. The Boss was also keen on getting the groundcrew up on pax rides with "their" pilot in "their" aeroplane. We had remarkable serviceability; it's not common for six or nine aeroplanes to head across Europe and all arrive, yet I do not recall a single case of anyone being stuck halfway.

The antithesis of this was 'Centralised Servicing'!

Ground crew were not even on the Squadrons.

This was invoked for the Marham Valiant's, for the bomber squadrons, we being 'tankers' and specialised, kept our independence.

[TheGreenGoblin]

The ability to "pop in",,,

Too much popping in can be a bad thing.

True, all my biggest mistakes have been made after too much time in the bar but when writing real time software systems, I have found the stack and push and pop to be very efficient assembler techniques. I mean look how useful Forth is/was and how good Aida, old time language of the avionics software jockeys/gurus, was and still is (vide. Aida64) !

So very efficient and mathematically verifiable, as Fox3 probably knows!

push rax
pop ds

As they say!

[Fox3WheresMyBanana]

Brings back memories. I wrote my first long program at Uni, about 5k, direct in machine code, to control a remote control vehicle around a maze. I could only get a 6800 chip - all 8 bit, none of that fancy 16 bit stuff - and only a paper tape reader/writer to program it, hence the need for machine code. So my first cut-and-pasting really was! Intelligent use of the stack was very important, as well as remembering to clean it up every time an optional subroutine was called to ensure the correct data went to the correct argument. Critical was ensuring all optional branches were timed at the same number of cycles, so lots of NOPs.

I remember reading The Soul of a New Machine around that time - heady stuff.
http://www.ancientgeek.org.uk/ed/The_So ... achine.pdf

[TheGreenGoblin]

Brings back memories. I wrote my first long program at Uni, about 5k, direct in machine code, to control a remote control vehicle around a maze. I could only get a 6800 chip - all 8 bit, none of that fancy 16 bit stuff - and only a paper tape reader/writer to program it, hence the need for machine code. So my first cut-and-pasting really was! Intelligent use of the stack was very important, as well as remembering to clean it up every time an optional subroutine was called to ensure the correct data went to the correct argument. Critical was ensuring all optional branches were timed at the same number of cycles, so lots of NOPs.

I remember reading The Soul of a New Machine around that time - heady stuff.
http://www.ancientgeek.org.uk/ed/The_So ... achine.pdf

Stracey Kidder!

Superb!

Error made because I used to tease my sister Anastasia with that name...

She used to accuse me of being a nerd after reading same book! How dare she!

The Motorola 6800 was superb and the kids in the UK had the 6502 in the BBC to sharpen their techniques. The on-board assembler on the 6502 (as you know I am sure) was superb.

[TheGreenGoblin]

My ex boss used to work for Data General (and knew Kidder) and was instrumental in me getting my job with MCDonnell Douglas.

It has been downhill ever since!

[Rwy in Sight]

TGG, I was thinking more in the sense that pop-in create a non-formal atmosphere where mistakes can be made from oversight or "get the job" done mentality.