🚀 go-pugleaf

-----BEGIN PGP SIGNED MESSAGE-----

In article <41fno9$bpg@sundog.tiac.net>,
Jim Williams  <williams@tiac.net> wrote:

>Multitasking == pre-emptive multitasking.
>
>I am unwilling to accept some marketing manager's definition of
>multitasking.

would you prefer a computer scientist's definition?

"multitasking" (also known as "multiprogramming") is any scheme which
increases CPU utilization by keeping several jobs in memory at once
and switching between them so that the processor is not idling during
I/O waits.  all that is really required is for processes to give up
the CPU while waiting for I/O.

there are obviously many different scheduling policies possible;
however, any scheduling system which does not allow a task to hold
the CPU while it is waiting for I/O completion can be properly said to
be "multitasking".  it matters not if there are other ways a process
can (voluntarily or not) release the CPU.

furthermore, there is no "ideal" scheduling policy; much will depend
on the nature of the tasks run on that system, the expectation of
users, and other factors.  that's why most systems are tunable.

-----BEGIN PGP SIGNATURE-----
Version: 2.6.2

iQCVAwUBMDtzEBbgfSHT/piVAQHmiwP/STMKgSoc3njAPGiOkYRcbYnD5cO1iZuK
byZXpp76/7pVCbvw/Ts3Y+p13uz3/ENJVqinKu6ihzPb1m6tibMBqAAI8hzBR2Ra
bdj9zBH1alMhY2ENnIZGDXk60cOpp3gvbwpp9vgD5sRUVvxJoqhRelOisqwOt132
9nhfhAmz/Jg=
=EOpf
-----END PGP SIGNATURE-----