[Coco] Good old days with many CPUs Re: Midwest VCF

[Steve Bjork]

The stacks tell the story...

The TI9900 CPU gets a Zero for 0 stack pointers
The 6502 CPU gets 1/2 because their stack is an 8-bit pointer. (Too easy 
to overflow)
The Z80 CPU get a One for a 16-bit stack pointer.
The 6809 CPU gets a Two for 2 stack pointers (system and user, both with 
index)
the 680X0 CPU get a Two (or better) for One system stack and up to 7 
User stacks pointers.

I would not want to write OS-9 for any CPU with a stack score of less 
than 2.

Steve

> You might say I had 3 actually because my next project at a small town ma&
> pa radio station was an automatic transmitter control where I designed&
> built everything but a pair of Z80 based Micro-Professor boards.  But after
> about a years coding and hardware building in my spare time, I came to the
> conclusion that the Z80 world was a heck of a good place to be _from_.  And
> never went back, that cpu was as dane bramaged as the 6502 IMO.  It turned
> out I was fighting with buggy silicon, and zilog would not replace it since
> it took me 6 months past the original 90 day warranty to determine it was
> flaky.  The $EB command byte was supposed to swap the foreground and
> background register sets, but it only worked about 95% of the time.
>
> Sorry Zilog, but none of your products was ever again considered for any
> projects I dreamed about.  Besides, the 6x09 is easily 10x smarter than
> your Z80, so there.  Zilog's loss maybe, certainly not mine.
>
> When I finally figured out what was so different as to be magic about the
> TI9900 cpu, it was no longer a contender for new designs, primarily because
> the price of a ticket to get into that arena was a 4 digit figure and the
> first digit wasn't a 1.  The TI could have been a good chip, but TI wanted
> ALL the money up front rather than a 50 cents each commission on 50 million
> of them.  Shrug.  I still have a TI99/4a though, but its not been powered
> in years.  It did work the last time though.
>
> Cheers, gene