Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!ucsd!ogicse!milton!biostr.biostr.washington.edu!kraig
From: kraig@biostr.biostr.washington.edu (Kraig Eno)
Newsgroups: comp.sys.mac.system
Subject: Re: PPC, IAC, and True Multitasking (tm)
Message-ID: <6626@milton.u.washington.edu>
Date: 22 Aug 90 23:59:05 GMT
Sender: news@milton.u.washington.edu
Organization: University of Washington
Lines: 27
References:<1990Aug3.040513.14844@d.cs.okstate.edu> <2760@bridge2.ESD.3Com.COM> <13888@cbmvax.commodore.com> <923@mdavcr.UUCP>

In article <923@mdavcr.UUCP> Eric W. Mitchell writes:
> Preemptive systems are typically less reliant on good programming habits
> of the software developers.

I think you are saying that the developers should make their programs
"multifinder-aware", because if they ignore the issue entirely then their
programs won't appear to function well when run simultaneously with other
programs.

You have put your finger squarely on the issue under discussion.  All the
stuff about scheduling algorithms and performance is a side issue; the point
in question (which so far hasn't been well specified) is whether "true
multitasking" should or should not require cooperation by the individual
user programs.

I am not going to post to the net any more on this issue, because there are
just too many people talking about too many different things and bringing
it all under the heading of "multitasking".

The big problem is that everyone is right about what they are saying,
basically.  If you think a developer SHOULD have to worry about the issue
when writing his program and that's what you call multitasking, then OK.
I was taught that a multitasking system should take care of task switching
for you -- the programmer shouldn't have to make some non-portable system
call every millisecond to make sure his program will allow others to run
simultaneously.  It all takes a lot of work under MultiFinder, and I don't
call that multitasking.