Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!uakari.primate.wisc.edu!ginosko!uunet!microsoft!brianw
From: brianw@microsoft.UUCP (Brian Willoughby)
Newsgroups: comp.sys.apple
Subject: Re: Apple's committment to the // line
Message-ID: <8192@microsoft.UUCP>
Date: 24 Oct 89 20:23:36 GMT
References: <8910161138.AA10844@trout.nosc.mil> <11322@smoke.BRL.MIL>
Reply-To: brianw@microsoft.UUCP (Brian Willoughby)
Organization: Microsoft Corp., Redmond WA
Lines: 47

In article <11322@smoke.BRL.MIL> gwyn@brl.arpa (Doug Gwyn) writes:
>In article <8910161138.AA10844@trout.nosc.mil> sysop@pro-generic.cts.com (Matthew Montano) writes:
>>386 machine is basically a superfast CP/M style machine running mostly
>>brain-dead operating systems (MS-DOS), OS/2 is something to behold but it is
>>meerly taking advantage of the bus speed and memory capacity of these souped
>>up CP/M machines. ... There have been next to none advancements in the Intel
>>world for years, they meerly stuff faster x86's in them, more memory and
>>sometimes double the bus width.
>
>You shouldn't talk about things you don't know about.  The 386
>supports 8086 (NOT 8080 or CP/M) the way the 65816 supports 6502,
>but is MUCH, MUCH better in its native mode.  In fact the first
>available 386 systems were all UNIX boxes.  Damn near any UNIX
>makes GS/OS look like the toy operating system that it is.
>Additional graphics support is pretty much orthogonal to OS and
>CPU considerations, and there are many bitmap-supporting UNIX
>implementations (even Apple offers one).  Check out the Sun 386i.

You're right about the 386, Doug, but what have you proven?  If you take
the original 4.77 MHz 8088 PC and multiply the clock speed by 6.9
(33 MHz), increase to 32 bits, and add caching, hard disk and virtual
memory, then you do have a decent machine.  The fact that it is based on
old technology, and doesn't have true general purpose registers is only a
small limiting factor - mostly the machine is so fast you don't notice
the many restrictions.  But tell me this: why do you think that a 6502
running at TEN times the clock speed (I own a 10 MHz W65C802) of a
machine comparable to the PC (the Apple I was faster than the PC in some
benchmarks), with extended 16 bit registers (eventually 32 bit), and the
rest of the support needed couldn't run UNIX?  Sure, such an Apple would
need RAM caching (what modern PC doesn't?), hard disk, possibly virtual
memory, and some controllable interrupt source (a MC6840 for task switch
interrupts) to be able to run UNIX well, but nothing prevents all this.
I don't see why the 65C816 couldn't offer an affordable alternative in an
architecture that is familiar, and a little more developer friendly than
the (still) segmented, special purpose register set 80x86.  For all of
its shortcomings, I think an 8 to 10 MHz Apple II would be so fast that
you wouldn't care (at least not any more than the 386 folks do) that it
isn't perfect.  Some of the improvements in the PC world include reducing
the number of video RAM wait states, so removing the 1 MHz video
restriction on the GS (with SE/030-style dual port video RAM) would help
greatly as well.

Brian Willoughby
UUCP:           ...!{tikal, sun, uunet, elwood}!microsoft!brianw
Bitnet          brianw@microsoft.UUCP
InterNet:       microsoft!brianw@uunet.UU.NET
  or:           microsoft!brianw@Sun.COM