Re: Wozniak vs Jobs

[toddpw@ugcs.caltech.edu (Todd P. Whitesel)]

supertimer@aol.com (Supertimer) writes:

>You keep wrapping it back to this, but the fact of the matter is the
>ARM was more powerful than the 68000 and Apple in fact
>designed and prototyped an ARM A2 that could run A2 software
>as fast as the real thing.

A 6mhz ARM emulating a 1mhz 6502. Right. Do we have any witnesses?

>The ARM was a hell of a lot faster than the 68000.  The 68000
>could match, what, a fast 286?  The ARM was moving at the real
>speed of a 486 before there were any real 486.  There was once
>a comparison of a game animated with shaded polygons on the
>Archimedes ARM based computer and a port to the Amiga 1000.
>Even with all of the Amiga's cool custom co-processors helping
>the huffing and puffing little 68000 along, the Amiga port could
>not keep up with the Archimedes original.  The ARM was just
>a more powerful CPU...

And the Archimedes had custom co-processors of its own. They also
had hardware floating point on the ARM, probably much earlier than
the Amiga added a 68881/2 or 68040. Hardware floating point makes
a _big_ difference for 3D polygon animation.

>Apple had a real working prototype of an ARM A2 that can be
>verified by Charlie Springer, an Apple employee of the time.

Sure, but will he verify all these great claims you keep making
about the ARM itself? Hey, I agree the ARM _should_ have one,
but the fact is it DIDN'T in the entire industry, not just at
Apple. Only when the handheld market made low power a critical
issue did the ARM start kicking American CPU's in the butt.

>It should have become the Mac because this "Mac" could run
>A2 software.

Sure, it "should" have from a purely technical perspective. What
I keep saying is that (a) its technical advantages were not as
great as you say, and (b) there were plenty of other political and
business reasons for going with the 68K family. I don't think Jobs
would have cared -- if the ARM prototype had been built early enough
to have been the Mac prototype. Instead it was viewed as an uncertain
competitor to the strongly progressing Macintosh prototype.

>You are accusing me of misusing 20/20 hindsight when you, having
>the benefit of seeing the 68000 become a standard because of its
>use in Macintosh, claim that a similar success could not have

It was not the Macintosh that made the 68000 a standard. The embedded
communications industry made the 68000 a standard. It was the Macintosh
and the explosion of 68020 based unix workstations that made the 68000
try to be a desktop processor. For years the division of Motorla that
made the 68K was the communications division, not the microprocessor
division. A few years ago they finally reorganized it to put the last
desktop chips (68040/68060) under their own management. Motorola still
makes new 68K based communications controllers, although their rate of
introduct is slower now that PowerPC communication controllers and
ColdFire (stripped-down 68K instruction set) are starting to eat into
the high and low end of the 68K family business.

There's a whole world of microprocessors out there easily 10 or 100
times larger than the desktop computer world.

>Now you are backtracking and saying ARM sucks?  Come on.
>This is blind support for Apple and denial of Apple's past mistakes!

>The ARM however DID evolve.  It WAS the superior chip to
>the 68000.  The Mac COULD have been based on it.  An ARM
>A2 WAS prototyped and could run A2 software.  Mix the last
>two and what do you get?  The realization that if not for Jobs
>stubborn insistance that the Mac and A2 lines remained
>separate, a more powerful original Mac would have appeared
>(call it ARM A2 if the name Mac offends) that would have
>been able to run A2 software.  Rather than siblicide, we
>would have had a continuous evolution of A2 into Mac like
>today's evolution of 68k Mac into PPC Mac...which is more
>logical, right?

You give Jobs way too much credit. Company politics at Apple or any
large company is a lot more complicated than that. If you could show
that the ARM A2 prototype existed before the Mac 68K motherboard
prototype had been developed much, then maybe you'd have an argument.
But especially in big companies, when you commit to a large product
and whip engineers into a frenzy, you either follow the project through
or you kill it. You don't re-route it halfway through even if that would
be "right" technically; it may be right but it's a damn sight more
expensive.

>If Jobs was in charge of the 68k Mac to PPC Mac transition,
>do you really think the PPC Mac would have been a Mac
>at all?  That does not fit with "next big thing" and "keep the
>two lines distinct and separate" mentality, you know!

Well, Jobs probably would have been ousted later instead of sooner.
I think there were enough people at Apple willing to stand firm and
boot Jobs that the survival of the Mac would have convinced them to
stand up to Jobs. He might have tried to screw the 68K->PPC transition
but I think he would have failed.

>If the 68k Mac and the PPC Mac were to remain separate
>and incompatible, would they sell better?

I am trying to describe how marketing people thought at that time. Since
then the PC clone market has proven that backwards compatibility is king,
but few people believed that when the Mac was still being developed.

>A little strange that of all Apple CEOs, ONLY Jobs could come
>up with this...

Doesn't matter. Analysts praised the move.

Todd Whitesel
toddpw @ ugcs.caltech.edu