Flying With Python (Strong versus Weak Typing)

[Jp Calderone]

On Wed, Mar 12, 2003 at 09:39:13AM -0800, Andy Salnikov wrote:
> 
> "Matt Gerrans" <mgerrans at mindspring.com> wrote in message
> news:b4jtai$nv8$1 at slb9.atl.mindspring.net...
> > Yes, as long as it wasn't pythonw.exe.
> >
> > Funny thing is, most of us have flown many times and absolutely no clue
> > what software might be involved (if any) in controling various aspects
> > of the plane and ground control.
> >
> > It is also interesting that very advanced nuclear submarines are run
> > largely on good old-fashioned mechanical technology.  They don't want
> > the whole thing dependent on computers, because it is too risky.
> >
>   Interesting I find this mechanical design closer to C++ static type
> checking than to Python dynamic type checking (I do not like weak/strong
> type, IMHO Python is strongly typed language, but it is dynamical). Every
> mechanical thing has kind of "interface" and can be joined with only those
> mechanical things which respect this "interface" (square peg vs. round
> hole). Joining typically happens during the submarine assembly, much like
> compiling C++ program. This is one of the reason why mission-critical
> systems prefer to use statically typed languages, you can "join" all parts
> together early during "asembly" (but it still does not guarantee it will
> work correctly, it's not a complete test of the system, like with the
> submarine you have to test every single part ans all parts together).
> 

  Can you "cast" a square peg to a round one, though? ;)  I think a better
analogy might be made with ML or another strongly, statically typed language
(C++ being a weakly, statically typed one).

  Jp

-- 
 up 9 days, 9:59, 9 users, load average: 0.44, 0.18, 0.06