Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!samsung!xylogics!world!madd
From: madd@world.std.com (jim frost)
Newsgroups: comp.sys.apple2
Subject: Re: pipes (implementation thereof)
Message-ID: <1990Mar25.025543.9364@world.std.com>
Date: 25 Mar 90 02:55:43 GMT
References: <2284@tellab5.tellabs.com>
Organization: Saber Software
Lines: 36

toth@tellab5.tellabs.com (Joseph G. Toth Jr.) writes:
>First of all - in order to provide multi-tasking (which is refered to in
>   the articles), the OS/Shell would have to work using interrupts.

No it wouldn't.  Using a timer to reschedule processes is called
"preemptive scheduling".  When a process explicitly or implicitly
gives up control so that a task switch can take place it's called
"non-preemptive scheduling".  The Mac multifinder uses the latter
(it's not the first example of non-preemptive scheduling, but it's a
familiar one that works pretty well).

Most of the time non-preemptive OS's do task switching during system
calls, which are very frequent in most interactive systems (such as
the Mac).

>Second - in order to make any gains provided by task switching, all I/O to
>   disk, etc. would require new hardware to allow interrupting I/O control
>   and/or DMA.

No, although it certainly would help if you didn't have to sit around
waiting for I/O all the time.  Again, look at the Macintosh
multifinder.  Smarter hardware improves throughput (that's why the IBM
3090 has such fast I/O -- whole computers are used as periferals).

All of these things are well covered in most operating systems texts;
this stuff has been around for awhile.

Also note that with some compiler help you can do protected memory on
the 6502 by emulating indirect memory references and non-relative
fetch, store, and jump functions.  Slow, but possible.

Happy hacking,

jim frost
saber software
jimf@saber.com