Re: 1571/1581 Sources: Macro WDTEST inserting NOP before reading WD177x status register / writing command register

From: Jim Brain <>
Date: Fri, 18 Mar 2022 00:28:05 -0500
Message-ID: <>
On 3/17/2022 8:08 AM, Rhialto wrote:
> On Wed 16 Mar 2022 at 11:13:10 +0000, smf wrote:
>> You implied that it was fixing hardware problems in software that killed
>> commodore, which every succesful company does (Intel and AMD software
>> mitigations for Spectre for example).
> For example the Apple ][ was the total example of fixing limitations of
> the hardware in software (or not at all). Think of the strange way it
> did composite "color", or how the disk controller was so minimal that it
> needed cycle-exact software drivers.
> -Olaf.

I have no argument for the above, though I think it should be clarified 
that the hardware limitations were in turn fixing something else, cost 
concerns. The video circuitry might have a debatable cost savings and 
might just have been a way for Woz to be cute.  As an engineer myself, 
it's a thing we sometimes do if we're not careful.  But, the disk 
controller was absolutely done to save costs, and it saved quite a bit.

Cue the "erhm, no worries, we'll patch it up in the software" notes above.

And, obviously, given the significant cost of the Apple II in spite of 
these cost savings, there's no doubt a debate on whether leaving the 
costlier circuits in place would have turned away any consumers (I mean, 
once you're paying $1500.00 in 1980 dollars, whats a couple hundred 
more), but I'll leave that up to another thread.

On the original topic, I actually think there were 2 "eras" of hardware 
design at CBM, of maybe more.  But, the PET and early 8 bit machines had 
extremely ingenious design elements.  I would place the 1581 in the 
later era, where maybe the design focus was elsewhere and cost 
containment was much more important. (It was very important in the first 
era, but somehow, the engineers managed to both cut costs *AND* do so in 
ingenious ways.)

Maybe it was that it was easier to get the MOS fab to churn out a 
slightly custom 65XX part in the early part of the decade, and that was 
not as possible in the latter half of the decade.  Not sure.


Jim Brain
Received on 2022-03-18 07:00:02

Archive generated by hypermail 2.3.0.