Python for air traffic control?

[Bengt Richter]

On 2 Jul 2001 13:06:57 -0700, 18k11tm001 at sneakemail.com (Russ) wrote:

>I agree that the quality of the design is paramount. BTW, I would like
>to thank everyone who responded to my query. I should probably point
>out that I am an aerospace engineer with some experience in C++ and
>Tcl/Tk, but I am a Python beginner.
>
>As for function prototyping, it seems to me that argument passing by
>keyword in Python can be used to advantage. If all procedure arguments
>are passed by keyword, there is much less chance that a procedure will
>be called incorrectly, particularly when a procedure interface is
>changed. Contrast that with C++. The C++ compiler will catch only some
>of those kinds of errors. If I change the order of two ints in a
>function interface, for example, the C++ compiler won't have a clue
>that the interface has changed.
>
>What would be very desirable for Python is some kind of static checker
>that can guarantee BEFORE run time that all procedures are called
>correctly. Does anyone know of anything that does that?
>
>Also, can anyone give specific examples of Python being used in
>safety-critical applications? Thanks.
>
I am wondering if using non-deterministic[1] memory allocation is
permissible in a safety-critical system. What kind of testing
could you specify to guarantee that code would continue to
run? Or does memory allocation/garbage collection have provable
behavior under conditions that you could implement within? Also,
do you have any hard-real-time requirements?

Actually, I would be surprised if you have the freedom to choose
to use anything in your final product that doesn't formally meet
stringent requirements, if it is really safety-critical. Better read
the requirements docs with a lawyer at your elbow, I would think ;-/

[1] Obviously I don't mean that a computer proceeds randomly, any
more than it does when generating pseudo-random numbers.