Re: Randy, wake up!

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

pubpc1@library.ucla.edu writes:

>kicked themselves in the butt so that now they're trying to
>blame the cloners for their troubles.

Did I miss something? I don't see Apple blaming anybody, just pulling the
rug out from under the clones.

BTW CHRP had problems. Some of my Mac friends pronounced it "crap" on purpose.

Do you really want a machine that has both ADB and PS/2 keyboard ports on it?
We're talking a larger back panel, with at least some of those ports going
unused for each individual user. More expensive with less added value. Uh oh.
I'm not familiar with the whole list, but this was not the only questionable
thing about CHRP.

>With Steve Jobs back, they are putting Rhapsody, they're last
>hope, on the back burner.  Have you heard, Rhapsody is only
>going to be on their servers...MacOS is going to remain
>separate and their prime focus.

Actually this is good. Mac software vendors would love to not rewrite
anything until after Rhapsody is solid. Establishing a beach-head on the
servers will make it a lot easier for a later transition in the user macs.
Right now it is more important to convince everyone that Rhapsody will not
get stuck somewhere indefinitely. Servers are the safest way to make sure
that _some_ version of Rhapsody really ships ASAP.

>Killing the clones may protect the Mac hardware, but hurt
>MacOS in the long run.  If Apple wants to be like Microsoft,
>it has let others be Compaq.  Do you really think PC's would
>have advanced so much had IBM continued its monopoly?

I think your 'Microsoft' was meant to be 'IBM' up there.
IBM published practically everything about the original PC in a desperate
attempt to attract the most third parties to the PC platform. If IBM had
kept the PC architecture all to itself, the PC would have failed in the
marketplace. As it was IBM published everything but retained the copyright
on the BIOS code; once someone clean-roomed the BIOS, IBM's only advantage
on the clones was economy of scale in manufacturing, and the development
of motherboard chip sets eliminated that advantage. IBM _did_ try to regain
its monopoly with the PS/2 and MicroChannel, but that was an utter failure
and only served to prove that IBM had lost control of the PC architecture.

Read "Accidental Empires" by Robert X. Cringely. There's a lot in this book
to make Apple II users proud. But there's also a lot in it to remind us that
luck and killer apps and business savvy matter a lot more in this industry
than pure technical excellence -- Apple's traditional strength.

>near 10%.  The reason Apple perceived the cloners as eating
>into their market is because their own Macs were poor
>compared to the clones.  Had Apple's Macs been up to snuff,
>the entire MacOS market would have expanded.

Of course! But if Apple's Macs _had_ been up to snuff, they wouldn't have
been in trouble either, so they would never have had reason to cancel the
clone market to fix it. I'll say it again: Apple's taking the lesser of two
evils because they have a responsibility to their shareholders and employees
that right now is more urgent than the needs of the Mac market in general.
It sucks, but sometimes life just sucks.

>And that's another thing: they should be nice to Motorolla.
>Who makes the PPC, after all.  And Randy says the cloners
>did nothing to advance MacOS.  What about the G3 chip!

IBM also makes the PPC. Motorola's Microprocessor division does not care
who they sell processors to. If Apple brings the volume back, they will be
forgiven, by the Microprocessor division at least.

>Not according to Western Design Center.  Their story is that
>they had the faster 65C816 chips in sufficient quantities.  At
>least when the ROM 3 came out, there was NO excuse for Apple not
>to have upped the speed of the GS.  Not when AE and Zip Tech was
>coming out with accelerators in mass quantities.

I've investigated this dispute personally. Bill Mensch's idea of sufficient
quantites seemed to be a bag of 65816's waved in front of everyone at K-Fest.
Apple's idea of sufficient quantities was a contract to deliver 50,000 CPUs
at some fairly regular interval, like every quarter. Orders of magnitude
difference here. Apple sold many more IIgs's (to schools and students) than
AE+Zip sold accelerators.

Bottom line: Bill Mensch failed to convince Apple he could produce the chips.

It didn't help that WDC had a history of lame yields for fast chips; the IIgs
was supposed to run at Color Ref (3.59 Mhz for the NTSC illiterate), but WDC
could not produce enough quantity above 3 Mhz, and Apple's schedule for the
IIgs couldn't wait -- the IIgs was Apple's panic response to the PCjr before
anyone knew enough about it to realize that it was a total dud.

In a related incident, years ago a grizzled Amiga engineer told us here on
csa2 that the Amiga was supposed to have been a 65816 based follow-on to the
C-128, but _every_ sample chip that WDC squeaked out went straight to Apple.
Commodore finally got fed up and went with the 68K, and "never looked back".

>Picture this: GOOD SALES->MORE ADVERTISEMENT->BETTER SALES->
>MORE 65C816 IN USE->INCENTIVE TO PRODUCE FASTER AND DERIVED
>65C832 and 32-bit bus cousin->MORE GS SALES

>_Apple_ was the one that broke this success cycle.

True enough. Apple was reeling from Jobs' running the place, the power plays
with Sculley and his horrible choices of top executives, the triple failures
of the Apple ///, the LISA, and the early Mac. Apple management was total
chaos; there was no hope for a coordinated effort with WDC -- it would have
taken somebody as psychotic as Jobs to make it happen, but he very clearly
felt the Apple II was a cash cow which wouldn't matter once the Mac took off.

However Motorola was already a huge company with lots of production capacity.
And with the 68K they had leap frogged all the "8-bit going on 16-bit" CPUs
by designing a truly 32 bit instruction set. Even though the first two chips
(68000/68010) were in some cases really 16 bit machines quietly emulating the
32 bit instructions, you got instant software compatibility with the 32 bit
architecture when the 68020 finally appeared. Factors like this were why many
early Unix machines (Sun, Apollo, HP 9000, Tektronix, Motorola Delta 4000...)
used the 68K.

An aside: one advantage the 8088 had was its 16-bit instruction set. This
meant easy software compatibility when the 8086 based PC came out. No black &
white here, my friends, technology is too hellaciously complicated for that.

My point? The 68K was extremely sexy and pure, whereas the 65xxx's were
vulnerable to the same "it's a kludged up 8 bit CPU" argument that could be
launched at the x86 in the PC. So the less technically aware Apple employees
would all have wondered why the 65xxx shouldn't just roll over and die. When
I was an intern there, the long term effect of this attitude depressed me so
much that when I went back to school I spent a lot of spare time on the four
or so versions of "The Apple //f" and sent the second version to Apple HQ.
Later I heard it had caused quite a stir for a week or so, but ultimately
had no real effect. If anything it probably convinced Apple even more that
it was too late to recover their blunders with the IIgs, especially when
they had the option of focusing all their resources on the Mac.

Todd Whitesel
toddpw @ ugcs.caltech.edu