Python for embedded systems (was Re: What does Python do)

[David Bolen]

tanzer at swing.co.at (Christian Tanzer) writes:

> You're right -- the important point is `deterministic'. But ensuring
> deterministic behaviour in the presence of dynamic memory allocation
> is a royal pain.
> 
> And while the combination of `a little care', `generous upper limit',
> and `certain cases' might allow deterministic behaviour with dynamic
> memory allocation you will still hava a hard time in validating the
> system (or convincing the FAA or another third party about it).

Of course, many embedded and/or real-time systems nowadays are likely
to have some tasks that have the "hardest" deterministic requirements,
while others are provided for external monitoring or manageability of
the system.  I'd expect that in many cases, if the dynamic portion of
the system was not in the primary operational path (e.g., you don't
mind if a person with a monitoring device plugged into the system only
gets their screen updated fractions of a second late) that Python
could very well be suited to such tasks, since it could make them much
more flexible.  As long as the underlying RTOS can ensure that the hard
tasks will always preempt the softer ones deterministically, it would
probably work out well.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/