6502 trivia (was Re: Non-fake Apple 1 on ebay)

[Ethan Dicks]

On 11/20/09, Brian Lanning <brianlanning at gmail.com> wrote:
> On Fri, Nov 20, 2009 at 10:16 AM, Al Kossow <aek at bitsavers.org> wrote:
>
>> Chuck Peddle told me a couple of months ago that the 6502 was never
>> intended to be a general-purpose microprocessor, it was designed to
>> be a replacement for hard-wired logic.

Well it certainly exceeded those expectations!

>> They had a die size target to hit to get to the price point they wanted
>> and pulled out things they thought were unnecessary for its use in that
>> market.

Interesting way to build it - start with a processor... discard things
until it fits in a few mm^2... profit!

>> In particular, the length of the registers. I had always wondered why they
>> built a microprocessor with an 8 bit stack pointer, when the previous 6800
>> design had 16.

The first two processors I worked with were the 1802 and the 6502.  I
do remember the wee stack size was occasionally a problem.  It would
have been nice to have been able to at least put it somewhere other
than $0100, but they probably didn't have enough room for spare
transistors to even do that.

As a little-brother to the 6800, it still does a pretty good job.  I
remember wishing for 16-bit registers, but in effect, zero-page is a
wad of slow 16-bit registers.

> lol  Who needs multiply and divide operations anyway?

When I used to write commercial games, we used to go to great lengths
to calculate screen addresses, etc., with tables and hard-coded
multiple routines (times3, times9, done with shifts and adds) since
brute-force multiple wasn't an option.

> Great info, I never knew that.

Interesting to hear the history on it.

-ethan