Re: What if... Developing a new Apple II

[Rubywand <rubywand@swbell.net>]

Sheane and Bonnie Meikle writes ...
> 
> Rubywand <rubywand@swbell.net> wrote in message
> 38BAF619.3B3DF285@swbell.net">news:38BAF619.3B3DF285@swbell.net...
> > Sheane and Bonnie Meikle writes ...
 ....
 ....
> > > I've started a thread at
> > > http://www.insidetheweb.com/mbs.cgi/mb996045 dealing with this subject,
> > > just click on "Time To Put Aside Differences!!!!"
> >  ....
> >
> > 'Putting aside differences' has a nice ring to it. A fair question is
> > "What differences?"
> 
> The bickering between platforms and within platforms
>

     Really, I haven't bickered with C-64 or Atari users for years. After both
groups got backstabbed by their respective makers in 1985, there wasn't much
to bicker about. (Well, there was a bit of flak with Amiga types in the later
80's; but, bickering with 68xxx users doesn't count.)

 
> >      Beyond the above, "agreement" may come down to building in enough
> > flexibility to permit accommodating different OS's, disk formats, etc..
> 
> I agree. As Microsoft moves towards a more closed system we need to make a
> system for people who like options. Thats why Linux interest is increasing
> so rapidly. Flexibility and some sort of Basic are probably the to most
> important thing this machine should be built on.
>

     Microsoft seems to know what it is doing. It may even have a nasty
surprise in store for the current PC world should the government be dumb
enough to significantly mess with the status quo.

 
> >      Whatever, we are not talking about scrapping C-64's, Atari 800's, or
> > Apple II's. We are talking about introducing a new, high-performance 
> > machine which has much in common with these earlier computers. This 
> > does not guarantee direct compatibility with older software; but, it
> > should be easier to do conversions and, if necessary, get good results
> > with emulation.
> 
> I agree that emulation may be the way to go, but I also think that downwords
> compatibility with at least one of these lines is important as well.
> Hardware compatibility is an important factor as well.
>

     Downward compatibility at the hardware level is an interesting issue. The
traditional approach has been to design in hardware which directly mimics the
older machine (as in the IIgs) or to just add the old machine via an optional
plug-in (as with the Mac's IIe card).

     One of the lessons of Apple II emulation is that you need surprisingly
little 'up front' hardware compatibility to run Apple II software. The same
goes for C-64, Atari, Amiga, etc. software.

     It turns out that good fast hardware is, generally, "compatible" with old
slow hardware. Suppose the IIgs been designed as a simple pure 65816 machine
running at max speed with fast cache RAM and 16MB of un-kludged memory space
for ROM and main RAM plus a 640 x 480 16-color display. With normal uP speed
upgrades-- the kind we've seen in the 80x86 uP series-- it could have been
running old-Apple software at full speed within a couple years. 

     If our new machine has an interface which can access Apple II drives it
will have very good old-Apple compatibility. (And, if it has an interface for
C-64 drives, it will have very good C-64 compatibility. ...)


 
> >      Possibly, the easiest way to get things going is for the first
> > motherboard to be very simple with lots of Slots. For instance, it could
> > consist mainly of a microprocessor, flash ROM, RAM slots, and card Slots.
> > We might borrow PCI for the latter.
 ....
> 
> Again I agree, but cards must be available at the time of the computers
> release.
>

     Depends upon what you mean by "release". The initial version would have
lots of potential but, aside from good monitor firmware, not much software.
The bus for talking to disk controllers, etc. would be in place; but, there
might not be any OS.

     The idea is to get the raw machine 'out' and into the hands of users with
minimum hassle so that users can get started developing interface software,
operating systems, languages, etc.. (I guess someone might come up with a cool
SVGA game with enough of an OS to boot from whatever drive gets attached to
whatever controller card.)

     You could not very well release the first version of such a computer into
the home market as 'The Answer' to PC. After a year or so of development by
users, that could change.  

 
> > The motherboard would have an I/O port to allow loading in software from
> > an Apple II (or other computer) and a keyboard port. The I/O port would
> > eliminate the need to start with ready to go disk I/O.
> 
> The I/O port should allow adaptation to exept different drives like the Disk
> II or the 1540/41

     I was thinking more in terms of a direct serial or parallel connection.
Including Apple II and C-64 5.25" interfaces may be a good idea. The
interfaces would be developed anyway, by users; and, including the hardware
would increase costs. On the other hand, having the interfaces in place at the
start could significantly boost interest in the product. Whether or not to
include any feature involves a similar trade-off. 


> >
> > There would also need to be a simple VGA display interface. "Simple"
> > here refers to what the video controller is, at first, programmed to do.
> > Given a flexible, fast controller and user access to programming it, the 
> > video system could handle all sorts of effects.
> 
> Maybe a card based upon the Second sight card? Maybe if we aproached the
> Sequential would make us a bunch?
 ....

     Perhaps Sequential would. Really, though, Second Sight design information
and schematics would be of more interest.

     The card itself is intended for fitting into a IIgs. This makes Second
Sight both more complex and more limited than a simpler design which is user
programmable and can access a portion of main RAM would be.

     

> > Since we would want the motherboard to work at reasonably high speeds,
> > the microprocessor might be on a Slot card so that upgrades to speedier,
> > larger processors would be easier.
> 
> I would rather see something like a socket 7 set up for the microprocessor.
 ....

     Yes; that would be necessary if the microprocessor card includes just the
uP chip. If it includes the uP, cache RAM and controller, buffers, ... , it
could go into a regular Slot.

     Whether to go with a special Slot, a regular Slot, or just put it on the
main board is something that can be decided once we have a better idea of the
chip or chip set that is best for the job.



Rubywand