Bryan Pope wrote:
> > One thing is clear: forget Windows :)
> HA! Are you sure? ;)
Positively.
> > > > #### Linux ????
> >
> I had read once about a true real-time version of Linux. But since I don't
> remember the web address that won't help...
That'd be RT-Linux. However, it won't be much of help. See below for
details.
> > > I think you can safely rule out *any* multitasking system. It introduces
> > > latencies and delays which are clearly unacceptible.
> >
> > What about VxWorks (or any other *hard* RT system)? I would not rule that
> > out "per se". (Although, I don't think it would be highly available at the
> > hobbyist's...)
> >
> What about QNX? It is free*, fast and *real-time*. You can DL the CD ISO
> from http://get.qnx.com
But it won't be fast enough.
QNX claims to be very fast (faster than VxWorks, for instance) by guaranteeing
a 6 us maximum response time unless I'm mistaken. That is a *multitude* of the
allowed deviation.
Heck, QNX claims to have a *context switch* time of only .55 us when
running on a Pentium III. Apart from the fact that the actual context switch
is just a part of the whole aspect of response time, even the bare
.55 us is *way* too much (And a Pentium III is a very fast machine..)
"true realtime" and "true multitasking" are mutually exclusive. You can go a
fair deal (See QNX or RT-Linux), as long as your service frequencies and
latencies are in the millisecond (or at least a reasonable fraction of a
millisecond) area, and as long as your real-time applications are mutally
tuned.
RT-systems a la VxWorks, RTLinux and QNX are very interesting, and very
powerful, but not quite up to this job.
Sorry to dissapoint you..
--
Martijn van Buul - Pino@dohd.org - http://www.stack.nl/~martijnb/
Geek code: G-- - Visit OuterSpace: mud.stack.nl 3333
Kees J. Bot: The sum of CPU power and user brain power is a constant.
-
This message was sent through the cbm-hackers mailing list.
To unsubscribe: echo unsubscribe | mail cbm-hackers-request@dot.tml.hut.fi.
Archive generated by hypermail 2.1.1.