[Lazarus] *** GMX Spamverdacht *** Re: Suggestion for TRadioGroup documentation

[Giuliano Colla]

Il 08/04/2016 18:56, Jürgen Hestermann ha scritto:
> > If NASA or Airbus or Boeing engineers would use that approach, I 
> guess a lot of rockets, planes and whatnot would fall on our heads.
> > I am glad they do not seem to have this attitude.
>
> I am not sure that they do not have it.
> I saw a report on TV about a test flight of the A380 some years ago
> where technicians were wondering, why the air craft computer was
> pumping fuel from one tank to the other in a certain flight situation.
> It seemed they needed a lot of time to find it out.
> I would have expected that the complexity was not driven to a point
> where even the engineers do not fully understand what they have built.
> Could be that we just had a lot of luck.

This is more about *reading* the documentation, or maybe to 
*understanding* it, which is the subsequent step, once a decent 
documentation is available.

As the complexity increases, you cannot fit in the same page all the 
relevant information, and you must rely on reader's capability of 
understanding that what is stated at page 10 may carry implications to 
what is stated on page 900. This holds true for Airbus or Boeing, and 
for Lazarus and FPC.

I'm not an expert in avionics, but from my general knowledge even I 
could have told them the reason for pumping fuel from one tank to 
another, which is to balance the weight on the wings (fuel tanks are 
located on the wings). This function must be handled by some part of 
software related to the aircraft flight attitude most likely fully 
documented, but completely apart from the section which takes care of 
pumping fuel to motors!

A very similar case occurred with one of the first Airbus of Lufthansa. 
At landing it didn't stop at the end of the runway, and ended up in a 
cabbage field.
The subsequent investigation revealed that:

  * The Airbus had a protection preventing the reverse thrust if the
    landing gear isn't touching the ground. Reverse thrust was enabled
    only when all the wheels touch the ground.
  * Lufthansa procedure, in case of crosswind, is the sideslip landing,
    meaning the right (or left) wheels of the landing gear will touch
    ground much later.
  * As a consequence, in case of strong crosswind, reverse thrust was
    enabled too late.

It would appear that someone was unable to detect the problem generated 
by two fully documented facts, until an aircraft ended up in a cabbage 
field.

This sort of things do happen even when documentation is good, so let's 
imagine what may happen if documentation is poor or missing!

Giuliano



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lazarus-ide.org/pipermail/lazarus/attachments/20160409/c1a5126a/attachment-0003.html>