Re: [PATCH] scheduler patch, faster still

Albert D. Cahalan (acahalan@cs.uml.edu)
Tue, 29 Sep 1998 22:52:00 -0400 (EDT)


Justin A. Kolodziej writes:

> Ok, someone else provided an example for me (the Win-$@%#!^#$-printer).
> If Richard could show that a page from Netscape prints in something like
> 2 minute 32 seconds on a Winprinter without his proposed change and
> prints in only 1 minute 59 seconds with his change, that would be the
> sort of proof I'm talking about.

You are still thinking in terms of throughput. The Win-printer will
always print at exactly the same speed, or FAIL to print correctly.

The comparison you could get would look more like:

5 out of 1000 pages are garbage with the old code
3 out of 1000 pages are garbage with the new code

Of course I invented those numbers, but I think they should help
you understand the problems. You can't make the printer go any
faster or slower, but you can increase or decrease the chance of
producing garbage output.

-
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/