I thought I did - "bounded latency" is the heart of it. Maybe I'm the
one that's out of step with the rest of the world ;-))
This is now rather a long way off topic, so I think I'll be dropping it.
>
> > But how can you promise the customer that your system will respond to
> > events while they are current if you DON'T control the latency?
>
> You must be building your own motherboards, right? You can not make
> honest promises with hardware that might frequently retry to hide
> error conditions from the OS. The BIOS could even use a system management
> mode to steal cycles.
>
> I think this whole argument is silly. The goal should be "improve".
> There is no perfection. Think in terms of histograms with infinitely
> long tails that need to be squished as flat as possible. Every bit
> of improvement is useful -- maybe Linux goes from 5000 ruined CDs/year
> to 4352 ruined CDs/year. Good.
Lest I mislead anyone about my meaning: I personally would not attempt
"real" aka "hard" aka "realtime" realtime on a vanilla Linux box (at
least not if I cared about the outcome ;-)). I was addressing a generic
RT point, and that's why I'm going to drop it...
Neil
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/