Re: What if we designed and built a "32bit" Apple II?

[mjmahon@aol.com (Michael J. Mahon)]

Ranando King wrote:

>This is actually a reply to both Michaels, Roy, and Wayne...

OK.

>Your comments are quite reminiscent of my feelings about the Mac. I don't
>"like the Apple II line," I LOVE the Apple II line. The simple fact is that
>the hayday for these computers was the last time I ever saw complete
>documentation on how to do absolutely anything and everything with my
>computer. It was the last time a 7 year old kid could sit down and create
>wonders without having to spend several hundred or even thousands of dollars
>in additional software. It was the last time anyone could sit down and turn
>on an otherwise "stripped down" computer and still expect to be able to use
>it. It was the last time you could write a rudimentary disk operating system
>in under a day and actually end up with a functional system for loading and
>saving work. It was the last time a great many things were possible on a
>personal computer. The Mac and the IBM PC behind it made things far too
>complicated, too bloated, and far too under-documented to use with any
>reasonable expectation of being able to do lots of amazing things on your
>own.

I don't think we could be any more in agreement if we were trying!  ;-)

>2 notes on the 65832:
>
>1. Western Digital is supposedly planning on developing this chip at some
>point. I'm not sure I buy the line I got from their sales rep, but if they
>do build it, they have stated that it will still be compatible with the 6500
>series.
>2. Even if they don't get around to doing so, I'm already working on the
>layout for a similar chip that can be put on Xilinx or Q-Logic FPGA

I encourage anyone to have fun playing with their favorite technologies.
If something you would personally find rewarding is implementing a
32-bit extension of the 6502/65816, go for it!

Don't count on enough new "design wins" for such a processor to
make it a viable market, though.  There are simply too many better
alternatives, from almost every perspective except personal fun.

>About Memory:
>
>Sure, more memory does invite the risk of more bloated software, but I'm of
>the impression that modern day bloatware is mostly caused by lazy
>programmers working with complicated, poorly documented interfaces. I'm also
>of the impression that 16 mb should be all the address space this system
>needs. To keep at least partially true to the Apple II programming style,
>anyone wanting more memory than that can play the bank switching game.

Software bloat is the well-known product of market pressure to create
ever more feature-rich software in ever less time, combined with the
lack of any good hardware reason to constrain its size, like limited
memory size and processor speed.

Stacking four levels of GUI/object libraries on top of each other to
implement a "modular" "modern" system is a sure-fire way to get
a 1.3MB "Hello, world!" program--with a re-skinnable, semi-transparent,
parametric font dismissible dialog box that can be integrated with
scripting tools.

>About Emulation:
>
>Those of you with IIgs machines end up emulating older II machines any time
>you try to boot a P8 or DOS disk. You find yourselves emulating older II's
>when your system has to switch to 1.02 Mhz to perform some timing critical
>operation like disk access or whatever. It's all hardware emulation. So who
>really cares? You love your IIgs machines anyway.

Agreed, though the "emulation" involved is much more direct than
running 6502 code on a non-6502-like processor.

>All I want to do is build another machine that is created with the same
>design spirit as the II series. I want it to be compatible with the II's as
>much as possible. If that means that some hardware components will have to
>switch and act like older components at times, then so be it. If you're
>going to complain about it being another emulator, then you should also
>complain about your IIgs emulating a IIe, your IIe emulating a ][Plus and
>your ][Plus emulating a ][. Otherwise, you should let us dreamers have our
>dream.

Well, now that you mention it, as I've said in this forum before,
I personally find the IIgs to be a perversion of the Apple II "spirit".
GSOS, wonderful though it may be, is much more Mac-like than
Apple II-like, and it raises the bar for effective application development
_far_ above where it was in the ProDOS or DOS worlds.

I believe that the //e was the last machine for which it could be
truthfully said that all the documentation needed to completely
understand and program it weighed less than the computer itself.

BTW, compatibility is not the same thing as emulation.  When a
new machine is designed to _incorporate_ the logic of an older
version, that is "backward compatibility".  When a new machine
is designed so that software must simulate the logic of an older,
dissimilar machine, that is "emulation".

The line occasionally becomes gray.  For example, the slow
memory banks on the IIgs can really be regarded as an emulation
of //e memory, since there is software copying between banks,
and the _real_ memory is the fast memory.

But the HGR and DHGR graphic modes of the IIgs, for example,
are _real_ Apple II graphic modes, not some simulation of them,
possessing all their features and warts.

>Maybe you can even share in the dream, lending your rather obvious
>skepticism to the more technical aspects of the design and implimentation of
>such a machine. I'd much rather hear that this dream is a pipe dream because
>it's technically impossible to have an easy to program computer with more
>modern equipment in it than the comparatively complacent quibbling you 4
>have posted thus far. Such a statement with logic to back it up would be
>very helpful indeed.

I do not hold your dream against you, except insofar as it involves
a transformation of the Apple II community to generate a rebirth of
interest in a new, non-Apple "Apple II".  This is what I regard as a
pipe dream.

There is no technical impediment to implementing virtually any
hardware dream of an Apple II, but making the new capability
actually useful is a _much_ harder problem.

You say:

>considering how amazingly some of the projects for
>new cards in the II series have worked out, it might not be so unfeasable to
>think that such a dream as mine may come to pass inside of 2-3 years...
>assuming we can all work together on this.

The cards you describe are the work of one or two people (a very
compact team) doing a very limited design job aimed at working
in a large population of already-installed machines.  The total time
to design and develop these projects has been in the one year or
less range.

You are talking about requiring a multi-year, multi-person sustained
effort for success.  Do you see why I doubt it?

It is apparent that there are (at least) two different schools of thought
in the nominal "Apple II" community.  One believes that the Apple II
line is complete and is to be celebrated and enjoyed largely as it is,
with the occasional creation of a new software or hardware capability.

In my own case, I revel in the opportunity to do something that
hasn't been done before, with _no_ special hardware extensions.  (Of
course, I also enjoy my 8MHz Zip Chip to speed up assemblies, but
I would never release software that required acceleration, since that
would decimate the pool of possibly interested users.  ;-)

The other school of thought is that the Apple II needs to be modernized
to remain interesting, with much larger memory and faster processing,
and that emulation as a means to this end is an inferior approach.
A frequent component of this train of thought is that we should "just do
what Apple failed to do" in extending the Apple II line.

It's clear that I'm in the first camp, preferring to enjoy lilies as they
are rather than gilding them with 1/4" of gold.  ;-)

More power to anyone who personally wants to make a 1GHz, 128MB
"Apple II" with XGA graphics and 3D acceleration.  Just don't expect
that there will be any significant software written to exploit the new
capabilities.  It will just run Merlin Pro faster, or sort an Appleworks
database faster.  (But, of course, emulators already do that, and
they run on my laptop.)

There are lots of hardware things you can do with a machine that has
7 or 8 slots on a 1MHz, 8-bit bus, without having a sampling scope.
Not so easy on a machine with a 33MHz, 32-bit PCI bus.

I think that people are confused by the complexity of what appear to
be tiny things, like USB devices.  Since they are so inexpensive and
small, we naturally assume that they must be relatively simple--like
a parallel port.  Nothing could be further from the truth.  They are
hideously complex kludges of functionality, no two alike, and virtually
all of this functionality is undocumented in any public form.

Talking to a USB printer is more like talking to a network than to
a parallel port, with layers and layers of protocol and complexity.
My hat's off to anyone who can reverse-engineer a Windows driver
sufficiently to be able to implement a IIgs driver that will fit in less
than a megabyte and drive a printer to do useful things.

Someone please do it and prove me wrong, but I won't be holding
my breath.  ;-)

-michael

Check out parallel computing for 8-bit Apples on my
Home page:  http://members.aol.com/MJMahon/