> CPU REPORT
==========
Issue #109
----------
by Michael Arthur
CPU INSIGHTS
============
RJ Mical, and the Rise and Fall of Amiga Computer Inc.
======================================================
Gary Oberbrunner recently provided a great source of knowledge
about this, by writing and posting this essay on the Amiga newsgroup
(or message base) of Usenet. It is a transcript of a talk given by
R.J. Mical, the programmer who designed and developed the
Intuition graphical user interface for the Amiga, before the
Boston Computer Society in March, concerning the history of both
the Commodore Amiga itself, and Amiga Inc., the company who created
it. Except for modifications in its formatting, or presentation,
and various notes placed in this text to provide more information
on certain subjects, the content of Gary Oberbrunner's text is
identical....
The Early Days, Game Boxes, and the Guru Meditation
---------------------------------------------------
On Monday March 2, 1989, RJ Mical (=RJ=) spoke at the Boston
Computer Society meeting in Cambridge. Fortunately I was
momentarily possessed with an organizational passion, and I took
copious notes. I present them here filtered only through my memory
and my Ann Arbor. My comments are in [square brackets]. What
follows is a neutron-star condensed version of about three and one
half hours of completely uninterrupted discussion....
Amiga Computer Inc. had its beginnings, strangely
enough, RJ began, with the idea of three Florida doctors
who had a spare $7 million to invest.
They thought of opening a department store franchise,
but (as RJ said) they wanted to try something a bit more
exciting. So they decided to start a computer company.
"Yeah, that's it! A computer company! That's the ticket!
:-)"
They found Jay Miner, who was then at Atari, and Dave
Morse, the VP of sales (you can see their orientation
right off..) they lifted from Tonka Toys. The idea
right from the start was to make the most killer game box
they could. That was it, and nothing more. However
Jay and the techies had other ideas. Fortunately they
concealed them well, so the upper management types still
thought they were just getting a great game machine. Of
course the market for machines like that was hot
in 1982...
They got the name out of the thesaurus; they wanted to
convey the thought of friendliness, and Amiga was the first
synonym in the list. The fact that it came lexically
before Apple didn't hurt any either, said RJ.
However, before they could get a machine out the
door, they wanted to establish a "market presence" which
would give them an established name and some distribution
channels - keep thinking "game machine" - which they
did by selling peripherals and software that they bought
the rights to from other vendors. Principal among
these was the Joyboard, a sort of joystick that you stand
on, and you sway and wiggle your hips to control the
switches under the base. They had a ski game of course,
and some track & field type games that they sold with this
Joyboard. But one game the folks at Amiga Inc. thought
up themselves was the Zen Meditation game, where you sat on
the Joyboard and tried to remain perfectly motionless.
This was perfect relaxation from product development, as
well as from the ski game. And in fact, this is where
the term Guru Meditation comes from; the only way to
keep sane when your machine crashes all the time is the ol'
Joyboard. The execs tried to get them to take out the
Guru, but the early developers, bless 'em, raised such a hue
and cry they had to put it back in right away.
(Note: Recently, Commodore announced that the Term, "Guru
Meditation" would not be in AmigaDOS 1.4....)
When RJ interviewed with Amiga Computer (he had been at
Williams) in July 1983, the retail price target for the
Amiga was $400. Perfect for a killer game machine. By the
time he accepted three weeks later, the target was up to
$600 and rising fast. Partly this was due to the bottom
dropping completely out of the game market; the doctors and
the execs knew they had to have something more than just
another game box to survive. That's when the techies'
foresight in designing in everything from disk
controllers to keyboard (yes the original Amiga
had NO KEYBOARD), ports, and disk drives began to pay off.
The exciting part of the Amiga's development, in a
way its adolescence, that magical time of loss of innocence
and exposure to the beauties and cruelties of the real
world, began as plans were made to introduce it, secretly
of course, at the Winter CES on January 4th, 1984.
CES, THE AMIGA'S ADOLESCENCE, AND "BUSINESS IS WAR"
---------------------------------------------------
The software was done ten days before the CES, and running
fine on the simulators. Unfortunately when the hardware was
finally powered up several days later, (surprise) it didn't match
its simulations. This hardware, of course, was still not in
silicon. The custom chips were in fact large breadboards, placed
vertically around a central core and wired together round the edges
like a Cray. Each of the three custom 'chips' had one of these
towers, each one a mass of wires. According to RJ, the path
leading up to the first Amiga breadboard, with its roll-out
antistatic flooring, the antistatic walls just wide enough apart
for one person to fit through and all the signs saying Ground
Thyself, made one think of nothing so much as an altar to some
technology god.
After working feverishly right up to the opening minutes of
the CES, including most everybody working on Christmas, they had a
working Amiga, still in breadboard, at the show in the booth in a
special enclosed gray room, so they could give private demos.
Unfortunately if you rode up the exhibit-hall escalator and craned
your neck, you could see into the room from the top.
The Amiga was, RJ reminisced, the hardest he or most anyone
there had ever worked. "We worked with a great passion...my most
cherished memory is how much we cared about what we were
doing. We had something to prove...a real love for it. We created
our own sense of family out there." RJ and Dale Luck were
known as the "dancing fools" around the office because they'd play
really loud music and dance around during compiles to stay
awake.
After the first successful night of the CES, all the
marketing guys got dollar signs in their eyes because the Amiga made
SUCH a splash even though they were trying to keep it "secret."
And so, they took out all the technical staff for Italian food,
everyone got drunk and then they wandered back to the exhibit
hall to work some more on demos, quick bug fixes, features that
didn't work, and so on. At CES everyone worked about 20 hours a day,
when they weren't eating or sleeping.
Late that night, in their drunken stupor, Dale and RJ
put the finishing touches on what would become the canonical Amiga
demo, Boing.
At last! ...The true story is told.
THE COMMODORE YEARS: AMIGA FUTURES, AND BUSINESS AS USUAL
----------------------------------------------------------
After the CES, Amiga Inc. was very nearly broke and heavily in
debt. It had cost quite a bit more than the original $7 million
to bring the Amiga even that far, and lots more time and money were
needed to bring it to the market. Unfortunately the doctors wanted
out, and wouldn't invest any more. So outside funding was needed,
and quick.
The VP of Finance balanced things for a little while, and even
though they were $11 million in the hole they managed to pay off
the longest standing debts and keep one step ahead of Chapter
11. After much scrounging, they got enough money to take them to
the June CES; for that they had REAL WORKING SILICON. People kept
peeking under the skirts of the booth tables asking "Where's the
REAL computer generating these displays?"
Now money started flowing and interest was really being
generated in the media. And like most small companies, as soon as
the money came in the door it was spent. More people were added
- hardware folks to optimize and cost-reduce the design; software
people to finish the OS. Even the sudden influx of cash was only
enough to keep them out of bankruptcy, though; they were still
broke and getting broker all the time.
How much WOULD have been enough? RJ said that if he were
starting over, he'd need about $49 million to take the machine from
design idea to market. Of course Amiga Inc. had nowhere near
that much, and they were feeling the crunch. Everybody tightened
their belts and persevered somehow. They actually were at one
point so broke they couldn't meet their payroll; Dave Morse, the VP
of Sales, took out a second mortgage on his house to help cover it,
but it still wasn't enough.
They knew they were going under, and unless they could find
someone quick to buy them out they were going to be looking for jobs
very shortly. They talked to Sony, to Apple, to Phillips and HP,
Silicon Graphics (who just wanted the chips) and even Sears.
Finally...they called Atari. (Boo! Hiss! [literally - the
audience hissed at Jack Tramiel's name!] Trying to be discreet, RJ's
only personal comment on Jack Tramiel was (and it took him a
while to formulate this sentence) "an interesting product of the
capitalist system." Ahem.
Apparently Tramiel has been quoted as saying "Business is
War." Tramiel had recently left Commodore in a huff and
bought Atari "undercover" so that by the time he left C= he was
already CEO of Atari. Realizing that Commodore was coming out
with their own hot game machine, Tramiel figured he'd revenge himself
on them for dumping him by buying Amiga Inc. and driving C= down
the tubes with "his" superior product. So Atari gave them half a
million just for negotiating for a month; that money was gone in aday.
Of course Tramiel saw that Amiga Inc. wasn't in a very good
bargaining position; basically unless they were bought they were
on the street. So he offered them 98 cents a share; Dave Morse
held out for $2.00. But instead of bargaining in good faith,
every time Morse and Amiga tried to meet them halfway their bid went
down!
Amiga Inc.: "Okay, $1.50 a share."
Jack Tramiel: "No, we think we'll give you 80 cents."
Amiga Inc.: "How about $1.25?"
Jack Tramiel: "70 cents."
And so on...
Even Dave Morse, the staunchest believer in the concept that
was the Amiga, the guiding light who made everyone's hair stand on
end when he walked into the room, was getting depressed. Gloom set
in. Things looked grim.
Then, just three days before the month deadline was up,
Commodore called. Two days later they bought Amiga Inc. for $4.25
a share. They offered them $4.00, but Dave Morse TURNED THEM DOWN
saying it wasn't acceptable to his employees; he was on the verge
of walking out when they offered $4.25. He signed right then and
there.
Commodore gave them $27 million for development; they'd
never seen that much money in one place before. They went right out
and bought a Sun workstation for every software person, with Ethernet
and disk servers and everything. The excitement was back.
Commodore did many good things for the Amiga; not only
did they cost-reduce it without losing much functionality, they had
this concept of it as a business machine; this was a very
different attitude from what Amiga Inc. had been working with.
Because of that philosophy, they improved the keyboard [ha!
- garyo] and made lots of other little improvements that RJ didn't
elaborate on.
What could Commodore have given them that they didn't? The one
thing RJ wanted most from them was an extra 18 months of
development time. Unfortunately Commodore wasn't exactly rich right
then either, so they had to bring out the product ASAP [and when is
it ever any different?] Also, he said, they could have MARKETED it.
(applause!). If he'd had that extra 18 months, he could have
made Intuition a device rather than a separate kind of thing; he
could have released it much more bug-free.
The Future
----------
RJ's advice for A1000 owners: "Keep what you've got. It's not
worth it to trade up. The A1000 is really a better machine." This
may be sour grapes on RJ's part, since the Amiga 2000 was designed
in Braunschweig, West Germany, and the version of the A2000 being
worked on in Los Gatos was rejected in favor of the
Braunschweig-Commodore version. However the A1000 compares to the
A2000, though, the Los Gatos 2000 would have certainly been
better than either machine. C= management vetoed it because
Braunschweig promised a faster design turnaround (and, to their
credit, were much faster in execution than the Los Gatos group
would have been) and more cost-reduction, which was their
specialty. Los Gatos, on the other hand, wanted a dream machine
with vastly expanded capabilities in every facet of the machine.
The cruel financial facts forced C= to go with the Business Computer
Group, who did the Sidecar in Braunschweig as well, and quickly and
cheaply.
So they fired more than half the staff at the original Los
Gatos facility, one by one. That trauma was to some extent played
out on the net; no doubt many of you remember it as a very
difficult and emotional time. There are now only six people left in
Los Gatos, and their lease expired in March, so thus expires the
original Amiga group.
And..that's how RJ ended his talk; the rise and fall of
Amiga Computer Inc. The future of the Amiga is now in the hands of
Westchester and Braunschweig, and who knows what direction it will
take?
Q & A Session: Boston Computer Society and RJ Mical
----------------------------------------------------
I'll just make this part a list of technical questions
and answers, since that was the format at the talk anyway. This part
is part technical inquiries and part total rumor mill; caveat emptor.
Questions are from the audience, Answers are =RJ=.
-----------------------------
Q: Can you do double buffering with Intuition?
A: Pop answer: No. Thought-out: well, yes, but it's not easy. Use
MenuVerify and don't change the display while menus are up. It's
pretty hairy.
Q: How big is intuition (source code)?
A: The listings (commented) are about a foot thick, 60 lpp, 1 inch
margins.
Q: Where did MetaComCo come into the Amiga story?
A: MCC's AmigaDOS was a backup plan; the original Los Gatos-written
AmigaDOS was done with some co-developers who dropped out due to
contract and money hassles when C= bought Amiga. Then MCC had to
crank EXTREMELY hard to get their BCPL DOS into the system at the
last possible minute.
Q: Why no MMU (support in the Amiga's Operating System)?
A: Several reasons. Obviously, cost was a factor. MMUs available
at the time the Amiga was designed also consumed system time [this
is what he said- I'm just the scribe]; although newer MMUs solve
this problem they were too late for the Amiga.
Second, the original goal of the Amiga was to be a killer game
machine with easy low-level access, and an MMU didn't seem
necessary for a game machine.
Third [get this!] with an MMU, message-passing becomes MUCH
hairier and slower, since in the Amiga messages are passed by just
passing a pointer to someone else's memory. With protection,
either public memory would need to be done and system calls issued
to allocate it, etc., or the entire message would have to be
passed. Yecch. So the lack of MMU actually speeds up the basic
operation of the Amiga several fold.
Q: Why no resource tracking?
A: The original AmigaDOS/Exec had resource tracking; it's a shame it
died.
Q: How is your game coming? [??]
A: It's just now becoming a front-burner project. It's number crunch
intensive; hopefully it will even take over the PC part of the
2000 for extra crunch. It's half action, half strategy; the
'creation' part is done, only the playing part needs to be
written. Next question. :-)
Q: Will there ever be an advanced version of the chip set?
A: Well, Jay Miner isn't working on anything right now... [RUMOR
ALERT] The chip folks left in Los Gatos who are losing their lease
in March were at one time thinking about 1k square 2meg chip space
128-color graphics, although still with 4 bit color DACs
though...and even stuff like a blitter per plane (!!) They were
supposed to be done now, in the original plans; the chip designers
will be gone in March, but the design may (?) continue in West
Chester. Maybe they'll be here two years from now.
Q: What will happen to the unused Los Gatos A2000 design?
A: ??????
(Note: Reportedly, this design eventually became the Amiga
3000's Enhanced Chip Set.)
Q: Should I upgrade from my 1000 to a 2000?
A: Probably not. The 2000 isn't enough better to justify the cost.
Unless you need the PC compatibility, RJ advocated staying with
the 1000. After all the 2000 doesn't have the nifty garage for
the keyboard...:-) The A1000 keyboard is better built; you can
have Kickstart on disk; it's smaller and a LOT quieter, [maybe not
than the old internal drives!!!] and uses less power; the 2000 has
no composite video out, plus the RGB quality is a tad worse.
Composite video (PAL or NTSC) is an extra-cost option with the
2000.
Q: Have you ever seen a working Amiga-Live!?
A: Yes, I've seen it taking 32-color images at 16fps, and HAM
pictures at something like half that. [!!] It's all done and
working. I don't know why it's not out. It sure beats Digiview
at 8 seconds per image!
Q: What do you use for Amiga development tools?
A: DPaint and Infominder, Aztec C, Andy Finkel's Microemacs.
Q: What's the future of the A1000?
A: They aren't making any right now; they're just shipping from
stock. But they do claim that they intend to continue making
them.
Note: Shortly after RJ Mical's talk, news surfaced that
Commodore had decided to not make anymore Amiga 1000s, but to
make a unified front with the Amiga 2000....)
Q: Who is the competition for Amiga right now?
A: The new Macs are so expensive, they're not a threat to the 2000,
much less the 1000. Atari's new stuff "doesn't impress me."
[that's all he said.]
Q: Why are the pixels 10% higher than wide?
A: The hardware came out that way, and it would have been a pain to
do it any other way due to sync-rate-multiple timing constraints.
-EOF