On Saturday, 1 September 2018 at 08:05:58 UTC, Walter Bright
wrote:
On 8/31/2018 5:47 PM, tide wrote:
I've already read them before. Why don't you explain what is
wrong with it rather than posting articles.
Because the articles explain the issues at length. Explaining
why your proposal is deeply flawed was the entire purpose I
wrote them.
I didn't write a proposal. I was explaining a flaw in your
proposal.
You are just taking one line comments without even thinking
about the context.
We can start with the observation that a fly-by-wire is not a
fundamentally different system than a fully powered hydraulic
system or even a pilot muscle cable system, when we're talking
about safety principles.
It is vastly different, do you know what fly by wire is? It means
the computer is taking input digitally and applying the commands
from the digital input into actual output. If the system
controlling that just stops working, how do you expect the pilot
to fly the plane? While all they are doing is moving a digital
sensor that is doing nothing because the system that reads it
input hit an assert.
There's nothing magic about software. It's just more
complicated (a fact that makes it even MORE important to adhere
to sound principles, not throw them out the window).
I didn't say to throw them the door, I'm saying there's a lot of
different ways to do things. And using asserts isn't the one ring
to rule all safety measures. There are different methods, and
depending on the application, as with anything, has it's pros and
cons where a different method will be more suitable.
