[Coco] What would a CoCo successor have to have as a minimum?
Steve Bjork
6809er at srbsoftware.com
Tue Nov 23 11:00:01 EST 2010
I programmed on both the 68000 and 65816 back in day as home console
game designer. I will tell you that 68000 was a dream to program on,
the 65816 was a nightmare. Just as the 6502 had real problems dealing
with the 8-bit boundary (every 256 bytes) the 65816 has the same problem
with 16-bit boundary (every 64k).
The design of 6502 was in the day when CPU were expensive to design and
make. Just removing one logic gate on the CPU made a real difference in
cost. So the designers of the 6502 save logic gates by making every
register 8-bits when they could. (Even the x and y memory pointers were
just 8-bits.)
Even the great and power Woz pick the 6502 for his Apple computer not
because he liked the design of the 6502. The chip was so cheap that Woz
got one for free from the supplier and built a computer around it.
Yes, the 65816 was a step forward and loved by those that only
programmed 6502. But for those with a more diverse CPU background saw
the chip as a way to keep the worst design parts 6502 alive in the world
of 16 and 32 bit computing.
Don't get me wrong, there was a place for the 6502. It was a good
choice for small little devices that did not have much in the ram or
hardware, like the Atari 2600. I don't think that game system would
have done as well if you used a CPU that cost more in both $$$ and code
space.
Steve Bjork
On 11/22/2010 11:32 PM, Little John wrote:
> Just wanted to toss in a bit about extending the CPU...
> William Mensch extended the 6502 to 16 bits with his 65816. It could
> still run as a 6502, but also had it's 16-bit mode with a 16-Megabyte
> address space. There was also a version called the 65802 which was a
> 65816 that could be plugged directly in place of a 6502 (I don't think
> it supported the extra memory though). Might something similar not be
> possible with the 6809?
> -JohnT-
>
More information about the Coco
mailing list