Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!usc!apple!agate!ucbvax!LSUVM.BITNET!$CSD211
From: $CSD211@LSUVM.BITNET (Mark Orr)
Newsgroups: comp.sys.apple2
Subject: Re: New Computer?
Message-ID: <9012032009.AA06137@apple.com>
Date: 3 Dec 90 19:38:53 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 114


|There is no particular problem with either the 6502 or the 65816. Sure it

There is a problem with the 6502/65816...its old, or at least its viewed as
old. BTW, what other computer on the market uses a processor without multiply
and divide instructions.

|would be nice to have more general purpose registers (my personal favorite is
|the 88000), but so what? It is not the processor that is important, it is the
|system. The GS as a system is the best on the market for home use. It was
|years ahead in 1986, although only really practical in 1989 with system 5.0.

Yeah, I like the 88000 too...but it's too expensive for what I had in mind.
An low cost embedded RISC like the VLSI VL86C020 (the ARM), an AMD 29000/ or
29050, or the Inmos T400 are all cheap (<$100) and very very powerful.

The GS was trailing edge in 1986, and little has changed to the hardware since
then...Two basic facts: (1) The Apple II needs a major upgrade NOW, (2) If the
II gets that long overdue upgrade, it won't be Apple Inc. doing it.

It'll take more than marketing to get the II back on track. It will take an
upgraded design AND a long term commitment to the II to make software
publishers come back. Apple is going to move heaven and earth to get II users
to go to the Mac LC...its a more marketable computer. If Apple came out with
a new Apple II this late in the game, they would incur the wrath of the
computer reporting media (I can see it now, on the cover of BYTE: "Apple goes
backwards") Apple is depending on the II educational base being converted to
the Mac...'cause it can't convince business users to buy the Mac.

|What do MIPS really mean anyhow? Will Mathematica on a 20 MIPS machine really
|balance your checkbook any better than Appleworks on a .35 MIPS machine? If

MIPS mean a great deal to us. The point is we don't have Mathematica. And all
the letter writing campaigns you can devise won't change that until there is
a change in the IIs hardware, and a commitment from its manufacturer. Besides,
the IIgs couldn't handle Mathematica, even with a Zip GS built in and the
bottlenecks removed. Mathematica is also a graphic program, and the IIgs's
graphics are so out of date it's not even funny. Don't tell me about
assembly language speed...it's extremely inconvienent for software publishers
to have to work in a specific processors instruction set. You can bet that
PC, UNIX, and Mac versions of Mathematica aren't written on the metal.

|your new machine requires a hardware emulator for II compatibility and that is
|all that it is used for, then why bother with a different processor at all? IF
|you use a software emulator then how well will a 20 MIPS processor emulate a
|.35MIPS processor? DO you really gain anything for the extra cost?

I don't want to emulate any more than we have to. My idea is to use BOTH the
'816 and a RISC. The '816 should handle serial I/O, keyboard/joystick/mouse
inputs (can't do ADB), floppy disks (can't do Smartport), and MIDI (can't be
run from serial ports). It would communicate with the RISC using an area of
dual-ported RAMs (and mabye some (or all) of the '816s address space could
be mapped, not to mention possibly some mutual DMA-access). The '816 would
handle IIe emulation (built in) and IIgs emulation (public domain program
containing the GS ROM images, like the UNIX //e emulator with hardware support)
The RISC side could assist the IIe/IIgs side by doing the graphics, sound and
acting as a math booster (like the FPE or AE Fastmath), and provide access to
the extra instructions, registers, and memory. This system would be like an
Apple II with a coprocessor card built in (like the Saybrooks and PDQ IIs of
the saintly days of yore) and could boost Applesoft (or applesoft clone)
performance (just like those 68k cards did). In addition, a new dual-processor
system monitor would be built in. When you typed the equivalent of call-151
you'd get a new monitor that could 'flip' back and forth between the RISC and
the '816, along with routines that could fetch data from each others address
spaces. A new 32-bit RISC BASIC would also be included, with full access to
system resources (like GS BASIC was meant to be). You see, it's what's in the
ROMs that make this new system every bit an Apple II.

|Remeber that the new system will have zero software available initially and
|will only run II software. How many 'named' companies are going to support yet
|another system?

Add UNIX support as an option and price it reasonably and you'll get plenty.
Though a low cost RISC computer, with FutureBus+, and with those ROM extensions
should generate enough interest to sell them. I thought of many possible case
designs (i.e. Cube shaped card cage, tower, Mac IIsi style-box) but it might
pay us to put it in an old Apple II (or Franklin) style case (at least for the
low cost model)

|If you want a non-apple II that emulates an Apple II then get the Mac LC or a
|DOS Box and the Trackstar board or a UNIX Box with the IIe emulator. Why even
|bother redesigning the same wheel that others have already built, assuming you
|want this kind of wheel.

Because those systems aren't Apple IIs. This is an extension of the basic
design. Essentially a coprocessor and bus built on top of the old system.
It's open like a II, its accessable and hackable like a II, runs all IIe and
most IIgs software. Apple is designing RISC based computers to replace the
Mac. Lets beat 'em to the punch. What more glorious way of extending the
life of the II line can there be.

|I like the GS as is. It should have more motherboard speed, better graphics
|modes, and just a little support, but otherwise the design is perfectly valid
|for what I want in a computer. Dammit, there is nothing wrong with the GS
|today that will not be automatically be wrong with whatever system gets
|designed as a replacement by the time it is in production.

Step into the '90s. The GS hasn't changed in four years and is way behind
its competition. Systems selling for less (i.e. PC clones) are faster, and
have much better graphics...that combined with the lack of support is
murdering the II. Someone posted that Sierra has released a new version of
Stellar-7 and asked when a GS version would be written. It'll never be
written. The GS cant do what that program needs it to do. Stellar-7 uses
VGA (640x480 graphics) with 256 colors; the GS cant do 640x480 and cant do
256 colors in its highest (640) mode.

|UUCP: bkj386!pnet91!ericmcg
|INET: ericmcg@pnet91.cts.com

----------------------------------
|  MARK A. ORR                   |
|  $CSD211 @ LSUVM.SNCC.LSU.EDU  |
|          @ LSUVM.BITNET        |
----------------------------------