Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!thunder.mcrcim.mcgill.edu!snorkelwacker.mit.edu!apple!heksterb
From: heksterb@Apple.COM (Ben Hekster)
Newsgroups: comp.sys.mac.system
Subject: Re: True Multitasking
Summary: Preemptive multitasking is the answer to *all* your problems!  (?)
Message-ID: <48122@apple.Apple.COM>
Date: 15 Jan 91 22:00:41 GMT
References: <19019@shlump.nac.dec.com>
Organization: Apple Computer Inc., Cupertino, CA
Lines: 44

awessels@ccwf.cc.utexas.edu (Allen Wessels) writes:
> [...] What exactly do you need your machine to do that "true" multitasking
> would allow?

long@mcntsh.enet.dec.com (Rich Long) responds:

> Oh...let's see. How about being able to switch out of the Finder during copy
> operations or disk formatting? How about not having one's download wedge
> because a menu was held down too long? How about getting decent foreground
> performance with something running in the background that is not a "good
> citizen"? How about real print spooling?

But these things are basically unrelated to the type of multitasking
involved.  Non-preemptive multitasking implies some judgment on the part
of the application programmer as to the desirability of relinquishing the
processor at a given stage in the program.  With preemptive multitasking these
decisions are always made by a scheduler using some fixed algorithm.  In my
opinion, a well-written application has the potential of having a much finer
and well-tuned use of the processor than probably most general schedulers.  I
am certainly not disputing that this requires more effort of the application
programmer.

	As to your observation regarding the effect of menu operations (mouse
drag operations in general) on background tasks--these are typically `high-
intensity' operations that require high feedback fidelity, so suspending
background applications seems the correct thing in these cases.  Have you ever
experienced mouse operations in preemptive multitasking environments?  My
experience with X Window on high-performance workstations (various window
managers), for instance has been that feedback is frequently so *awful*
(intermittent, or no reaction to mouse movements at all for periods on the
order of many seconds) as to make complicated interactions extremely
difficult.  (No flame on X intended, just an observation on the workstation
environment)

	If you want to experiment, you might try setting up MenuHook and
DragHook to a little routine that calls WaitNextEvent, and see if performance
is more to your liking.

_______________________________________________________________________________
Ben Hekster                           | "I don't want to start any blasphemous
                                      |  rumors/ but I think that God's got a
AppleLink:   heksterb                 |  sick sense of humor/ and when I die
Internet:    heksterb@apple.com       |  I expect to find Him laughing"
BITNET:      heksterb@henut5          | --Depeche Mode [Some Great Reward]