Re: If you could design a $300 homecomputer today, what would you choose?

["william strutts" <wrstrutts1@nospam.home.com>]

"Timothy A. Seufert" <tas@mindspring.com> wrote in message
news:tas-16D91B.21514716072001@news...
<snip>
> I believe you are 100% wrong in stating that Apple II programs never
> exceeded the "size of memory space" (I assume you mean 64K).  Unless I'm
> grossly mistaken, AppleWorks did just that.  It was and is quite
> possible to design programs which bank switch code and/or data in and
> out of the CPU's address space.  With care, it need not even be a pain
> in the butt.

I did say that they would have to be bank switched.  It would still
need an address range to move the data to like that extra 64K on
the 80 column IIe card does.  It is bank switched. What I meant
in saying they didn't exceed the size of memory was that only
64K could be accessed at one time.  To get more memory requires
the programmer to bank switch the program or to swap code
out and pull it from the Floppy/HD in his code.  No hardware
or OS automatically handled that on the Apple II.  The early
PC's like the 286 had all kinds of odd ball memory due to the
8-bit legacy of the 286 processor.  They still had 64K code
segments and any one chunk of code had to reside on a 64K
bank.  You don't want to place a loop on two banks of memory
because that would require you to bank switch constantly.

> For software like a RAMdisk, it's not a pain in the butt at all to use
> bank selection.  No different from selecting what sector of a disk drive
> to read in a driver for an ordinary electromechanical disk.

While that is true, it requires special coding on the part of the
program.  Not all Apple IIe programs supported the extra RAM
and it was the minority that did.  The program had to be aware
of the memory.  Some smarter 8-bits using the 6809, had a
MTU unit which would take care of the bank switching.  They
had a smart operating system called OS-9 that knew how to
handle the memory plus it had a hardware chip which handled
a lot of the memory management.

The other way was using the Chain command in the older
basics which would remove the app in memory and load
the referenced code.  Not all Basics had that command. That
made it a bit easier but on the Coco, it required the programmer
to write routines to access the memory.  It was not handled
automatically by the OS.
--
--
William R. Strutts - wrstrutts1@nospam.home.com - Whatever!

C'est moi!  http://www.facelink.com/wrstrutts

Just hacking away...

Remove nospam to reply.