Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!cs.utexas.edu!sdd.hp.com!elroy.jpl.nasa.gov!ames!eos!shelby!neon!Kermit.Stanford.EDU!philip
From: philip@Kermit.Stanford.EDU (Philip Machanick)
Newsgroups: comp.sys.mac.misc
Subject: Re: System Errors, MF --> Why???
Message-ID: <1990May29.202342.3462@Neon.Stanford.EDU>
Date: 29 May 90 20:23:42 GMT
References: <58604@coherent.coherent.com> <6364.266023a1@umiami.miami.edu> <16795@phoenix.Princeton.EDU> <2545@media-lab.MEDIA.MIT.EDU>
Sender: news@Neon.Stanford.EDU (USENET News System)
Reply-To: philip@pescadero.stanford.edu
Distribution: na
Organization: Computer Science Department, Stanford University
Lines: 23

In article <58604@coherent.coherent.com>, dplatt@coherent.com (Dave
Platt) writes:
> Many applications will probably run quite happily in a preemptive-
> multitasking environment, too... but I predict that Apple will probably
> require that preemption-aware applications must set a new flag in their
> SIZE resources (much like the "MultiFinder aware" bit that's there now).
> Only applications with this bit set will be preempted; there will
> probably also be some traps that an application can use to enable and
> disable preemption.  This combination of tweaks will allow applications
> to yield time more gracefully, while ensuring that those applications
> which tweak low memory don't get preempted when low memory is in a
> "dirty" state.
Yet another hack. Apple have already shown how to implement compatibility
for old applications while introducing pre-emptive multitasking. It's called
A/UX. I don't think unix is the right route to go for a mass-market system,
but the principle still applies. My prediction: System 8 will have
protection, one address space per application, controlled sharing through
shared pages, pre-emptive multi-tasking etc. - and System 7 running in a
compatability box. Oh yes, and probably on a RISC processor, emulating the
680x0 in software.

Philip Machanick
philip@pescadero.stanford.edu