Substandard Italian workmanship renders first 787s unsafe

On Jun 25, 2:10 pm, Mxsmanic wrote:
Boeing has discovered that the Italian firm to which it subcontracted
construction of tail assemblies isn't doing the work correctly, putting the 23
aircraft it has already built at risk:

http://travel.usatoday.com/flights/p...halts-787-test...

No surprise here. When you allow Third World companies to built vital parts to
your airplane, you should take for granted that there will be potentially
dangerous defects. That's the consequence of trying to be politically correct.

Wops tell me, "never trust a wop", it's good advice, but can
it be trusted?
Ken

On Jun 28, 5:35*am, Mxsmanic wrote:

There isn't any way to test software systems for correctness.

How is MSFS tested for correctness????????????????

writes:

How is MSFS tested for correctness????????????????

It isn't. Most desktop software is only very summarily tested, and then bugs
are fixed (or not) as they are uncovered by the user community.

Unfortunately, this philosophy has influenced some software developers in
other domains, such as those who write software for Garmin and Avidyne, and to
a lesser extent the developers of more sophisticated flight management systems
and even FBW.

In article ,
says...


Everything you are describing is analog. In digital systems that use binary
numbers (the majority of electronic systems), everything is either one or
zero. There is no half-one/half-zero. There are no intermediate values. That
is the nature of digital systems.

However, all real-world, physical systems that represent information
digitally
do so using an analog infrastructure. In the analog realm, intermediate
values
are possible; in the digital realm, they are not.

For example, the written word is a digital representation of information.
There are no "half W's" or "partial A's"; either a symbol is one letter, or
it's another. But the written word is typically actually stored on analog
physical devices, such as books or disk drives. These analog representations
must be resolved unambiguously to discrete values upon entering the digital
realm. The print on a page may be blurry, but in order to read it, each
letter
on the page must be unambiguously identified--there are no partial or
intermediate letters.


You would do well to stop spewing elementary school nonsense and go back to
that textbook (or Newsweek article or whatever your course of study was) and
learn the difference between a bit and a byte. When you have grasped that, you
will understand why everything you have just said is pure nonsense.



I've been studying the safety of aviation computer systems for several
decades.

Some people are slow learners.

wrote:
In rec.aviation.piloting Mxsmanic wrote:
Tom P writes:

Factory workers make mistakes all the time. It's normal.
The number and magnitude of mistakes they make depend a great deal on
corporate and social culture.

I recall Akio Morita describing such a problem. Sony was building Trinitrons
in both Japan and the USA. In both countries, the tubes had to meet the same
tolerances. Nevertheless, the company found that the Japanese tubes were
always far closer to perfection than the USA tubes.

Finally, management figure it out. The Japanese always tried to get things
perfect, no matter what the accepted tolerances were, whereas the Americans
didn't care whether it was perfect or not, as long as it fell within the
tolerances.

To fix this, Sony made the tolerances far tighter for the USA tubes. Their
quality then improved significantly.

All of which was stupidity on Sony's part.

All things have a tolerance and tighter tolerances increase costs.

If Sony needed tighter tolerance they should have originally specified
tighter tolerances instead of ****ing and moaning that stuff made to
their specified tolerance wasn't "good enough".

It was all about culture.

Wrong, it is all about getting the specifications to be what is really
needed in the first place.

That is basic engineering.


We have a realworld problem with a part right now. It's a tube with an
internal dimension of 58mm with a specified tolerance of +0.3mm - note
not -/+0.3mm.
Parts are being rejected on arrival at the customer as being under
58mm. The problem is that the manufacturer is passing parts measured
right down to 58mm, so with the minute difference between the mfgr's and
the customer's test pieces, the only solution is for the mfgr not to
ship anything under 58.1mm.

Mxsmanic wrote:
VOR-DME writes:

I said you were out of your depth, now you are sinking fast. I gave you that
something like a slew rate is an analog performance parameter, but to claim
there are no meaningful tolerances for digital systems is screaming nonsense.
No half-written byte? Of course there are. All the time. There are incorrect
words. Checksum errors. Bytes that arrive too late. A 30cm trace on a printed
circuit board translates to a nanosecond of transit time, and at GB speeds
this is meaningful.

Everything you are describing is analog. In digital systems that use binary
numbers (the majority of electronic systems), everything is either one or
zero. There is no half-one/half-zero. There are no intermediate values. That
is the nature of digital systems.

It all went right over the top of you head, didn't it?

Your knowledge of things digital is just as superficial as everything else
you claim to know about.


--
Jim Pennino

Remove .spam.sux to reply.

wrote in message
news
Mxsmanic wrote:
VOR-DME writes:

I said you were out of your depth, now you are sinking fast. I gave you
that
something like a slew rate is an analog performance parameter, but to
claim
there are no meaningful tolerances for digital systems is screaming
nonsense.
No half-written byte? Of course there are. All the time. There are
incorrect
words. Checksum errors. Bytes that arrive too late. A 30cm trace on a
printed
circuit board translates to a nanosecond of transit time, and at GB
speeds
this is meaningful.

Everything you are describing is analog. In digital systems that use
binary
numbers (the majority of electronic systems), everything is either one or
zero. There is no half-one/half-zero. There are no intermediate values.
That
is the nature of digital systems.

It all went right over the top of you head, didn't it?

Your knowledge of things digital is just as superficial as everything else
you claim to know about.


--
Jim Pennino

Remove .spam.sux to reply.

It really is hopeless, Jim,

If you were interviewing job applicants, too many like him could drive you
to drink.

Peter

writes:

It all went right over the top of you head, didn't it?

It may be the other way around. Most people don't understand the fundamental
nature of digital vs. analog, even in industries where you'd expect them to
know.

On Jul 1, 9:59*pm, "Peter Dohm" wrote:
wrote in message

news


Mxsmanic wrote:
VOR-DME writes:

I said you were out of your depth, now you are sinking fast. I gave you
that
something like a slew rate is an analog performance parameter, but to
claim
there are no meaningful tolerances for digital systems is screaming
nonsense.
No half-written byte? Of course there are. All the time. There are
incorrect
words. Checksum errors. *Bytes that arrive too late. A 30cm trace on a
printed
circuit board translates to a nanosecond of transit time, and at GB
speeds
this is meaningful.

Everything you are describing is analog. In digital systems that use
binary
numbers (the majority of electronic systems), everything is either one or
zero. There is no half-one/half-zero. There are no intermediate values..
That
is the nature of digital systems.

It all went right over the top of you head, didn't it?

Your knowledge of things digital is just as superficial as everything else
you claim to know about.

--
Jim Pennino

Remove .spam.sux to reply.

It really is hopeless, Jim,

If you were interviewing job applicants, too many like him could drive you
to drink.

Peter

Back in the last century when I did design, much of it digital, there
were many times the electrons refused to read my schematics. But
interviewing was not that difficult, many were VERY short because
there was no reason to talk further to someone who was as simplistic
as at least one of the posters here. My comment frequently was, "You
know, I don't think there's a good fit here. Why don't we finish our
coffee and talk about" -- insert here whatever we found as an area of
mutual interest.

Mxsmanic wrote:
writes:

It all went right over the top of you head, didn't it?

It may be the other way around. Most people don't understand the fundamental
nature of digital vs. analog, even in industries where you'd expect them to
know.

Yeah, sure, it is always the rest of the world (including the people that
have spent decades in any given field) that are uninformed and never you.


--
Jim Pennino

Remove .spam.sux to reply.