On Tue, 21 Aug 2001, Curt Sampson wrote:

> I can't find these functions you reference in a any library on NetBSD.
>
> Also, perhaps because I am not a very skilled C programmer, I don't
> understand the coding tricks you're using here. I notice, for example,
> that you compare cpuload with 0.9 and 0.95. Yet, cpuload is an int,
> and so the closest it would ever get would be 0 or 1. How does this work?

I assume that it works thanks to clever magic in the pseudocode
compiler. :)

The principle looks reasonable, but it would seem like you could get
better results keeping this kind of stuff in the kernel, where it can
react more quickly to a transient increase in CPU load, as is typical of
many systems.

-- 
Gillette - the best a man can forget