🚀 go-pugleaf

In article <41c9ut$g73@usenet.ucs.indiana.edu>, sgoehrin@copper.ucs.indiana.edu (scott goehring) writes:
:-----BEGIN PGP SIGNED MESSAGE-----
:
:In article <41astd$53u@steel.interlog.com>,
:Dave Glue <daveacg@interlog.com> wrote:
:
:>>So, what I'm trying to say is that any OS that runs two of more apps
:>>at the "same time" is basically multitasking.  However, underneath
:>>the method may not be as efficient as, say... NT!
:>
:>True.  Pre-emptive is more efficient and smoother, which is why it was
:>added into 95.
:
:preemptive multitasking is _less_ efficient than nonpreemptive
:multitasking, because of preemption overhead.  many supercomputers run
:either a nonpreemptive OS or a an OS with limited preemption (the only
:time a job can be preempted is when a job with higher priority than
:the running job enters the run queue), because this maximizes CPU
:utilization.

Depends on what you mean with efficiency. If you have a set of known
static tasks that are about to be computed will it for sure be
more effective to put them into a batch job (if one CPU) or calculate
the best schedule to balance the workload on several CPU's,
and, just as you wrote, less overhead.
Supercomputers are often used to calculate problems know before
computing starts. Say, do some simulation and when done come up with
a result. The set of jobs is static during the whole lifetime.

In a operating system is the workload a _LOT_ more dynamic and you
have no chance of planning any optimal schedules. By arranging
jobs by order of priority and interupt jobs of lower priority whenever
a more "important" job comes up can you have a static schedule to
handle a dynamic environment. By today's technology is this the best
known method to get good responsivess and performance.

With cooperative multitasking is the scheduling deligated to the jobs
themselves which have poor knowledge of the other system and jobs.
It is easy to understand the weakness in a such solution.

:however, maximizing CPU utilization is not generally the goal of most
:systems; balancing CPU utilization with user responsiveness is a more
:common goal.

Yep. There's another desired goal, minimizing the maximal allowed
responsetime. This is particulary desired in a strict realtime system.
Here's when cooperative multitasking leaves us, it can by no means
compete with PMT, even if none of them pass this goal flawless.

:preemptive multitasking was developed largely to support "time sharing
:systems".  read up on the history of computing.

Yep. The text above are wellknown operating systems concepts.

:i guess this all goes to prove that "those who do not learn from
:history are doomed to repeat it."  all of this nonsense has been done
:before, just on different computers by different companies.

Yep, that's why Apple and M$ are working their asses of themselves by
reinventing history (PMT) for their new operating systems. It took a while
for them to learn their history...

/Henrik