Re: Is Linux right for Embedded?

["Pat Rogers" <progers@classwide.com>]

"bgeer" <bgeer@xmission.xmission.com> wrote in message
news:9pa9of$9me$1@xmission.xmission.com...
> "Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> writes:
>  >I've used Ada to build jet and rocket engine controls with no COTS OS
(bare
>  >machine - our own "executive") where failure was not an option. Compared
to
>  >other languages we used, we got a doubling of productivity and a
four-fold
>  >reduction in errors. We considered it a major step forward in
reliability &
>  >could see no reason to go elsewhere.
>
> I participated in building a F16 radar simulator using no OS, C, &
> TMS320C30 processors running in parallel.  We had a good team willing
> to listen to those of us with extensive realtime experience.  The
> result was a system that could run essentially 24/7.  Ok, so this
> thing will never really fly nor kill anyone if it fails, but the fact
> is one can craft C code that doesn't fail.
>
> I also participated in an ADA based project that had a small team of
> "insiders" who wouldn't listen to "outsiders" & the result was worse
> code & less reliability.  This *was* a system meant to fly.
>
> Good coders can write good code regardless of language.  Language
> won't make marginal coders, marginal coding teams, or bad design any
> better.

You missed his point.  The question is not whether language is a replacement
for talent, and he did not assert that it was impossible to make high
reliability code with anything other than Ada.  The question is "for how
much money?".  His point (if I may be so bold) is that they did it much more
economically in Ada.  Other have seen these results too.

---
Patrick Rogers                       Consulting and Training in:
http://www.classwide.com          Real-Time/OO Languages
progers@classwide.com               Hard Deadline Schedulability Analysis
(281)648-3165                                 Software Fault Tolerance