Dealers of Lightning (2 page)

Timeline

It was April in California's Santa Clara Valley, a fine time to
be changing the world.

Very late one night in 1973 a small group assembled
inside the office of an electronics engineer named Charles P. Thacker.
The room was located on the ground floor of a low-slung building set
upon the crest of a gentle ridge in the foothills of the Santa Cruz
range. Pastureland and apricot orchards covered one side of the hill; a
spreading growth of industrial laboratories and research facilities
dotted the other, so that the ridge itself seemed to mark the divide
between the region's agricultural past and its high-technology future.
The building housing Thacker's lab, along with two others located in a
dale about a half-mile away, encompassed Xerox Corporation's Palo
Alto Research Center, known to its small but growing staff as Xerox
PARC.

The visitors had come to attend the birth of a computer. Today such
an event inevitably would be accompanied by crowds, banners, music,
speeches, multimedia shows projected on three-story-high outdoor
screens, press releases, media tours, and admiring cover pieces in all
the important magazines. Not to mention the smell of money, the
unambiguous signal of society's insatiable thirst for any technology
promising a smarter, faster, and brighter destiny.

On this occasion there was no such fanfare—a shame, given that the
machine Chuck Thacker was about to unveil to his colleagues would
help plant the seed of that modern frenzy. There was no smell of
money, only the barbed aroma of ozone and solder. None of those
present had joined PARC with the thought of becoming rich, anyway.
Xerox paid them well enough, a couple of notches over the standard
for scientists and engineers possessing their considerable skills. But
today's popular image of the computer nerd as incipient high-tech millionaire was nobody's fantasy then. Instead they had been attracted to
PARC by the thrill of pioneering. One of them compared it many years
later to the sheer joy of making the very first footprints in a field of vir­gin snow.

Thacker checked a few last electrical connections on his machine,
his cigarette smoldering nearby. He was thirty and of medium height,
with a squarish build and an unruly cowlick that seemed perpetually to
overhang his wily eyes like an awning. Among this group of youthful
Ph.D.s he was unusual in possessing merely a bachelor's degree in
physics, but their deference to him on questions of engineering was
unequivocal. Acknowledging the gifts that had already made him an
indispensable participant in the design and construction of two trail-
blazing large-scale computers, they paid him the ultimate accolade:
Chuck Thacker, they said, was an "engineer's engineer."

Thacker's designs were simple and spare, devoid of the egotism that
often spoiled the work of even the best of his fellow professionals. He
was a master of parsimony and the sworn enemy of its opposite, which
he called "biggerism." In a Thacker schematic one never found a logic
gate or a ground wire out of place, and he policed the work of his col­leagues so they would meet the same exacting standard. Any engineer
who
set forth a dubious or dishonest idea in PARC's Computer Science
Laboratory, where Thacker worked, was likely to be stopped in his
tracks by an explosive
"Bullshit!"
At PARC one found no shortage of
big egos and stern judges, but one thing on which all agreed was that
once Chuck Thacker pronounced your idea "bullshit," you had best
shut up and start shoveling.

It was therefore not surprising that when in 1972 the scientists of
PARC conceived a revolutionary kind of digital machine they relied on
Thacker to convert the concept into circuitry. The machine he and his
hand-picked team built in the course of an amazing few months con­formed to specifications never before required of a working computer.

Its most arresting element was its human scale. Where the typical
computer of this era was the size of two or three refrigerators standing
back to back and wired to many more racks of special-purpose hard­ware, the "Alto" was to be self-contained and small enough to bark a
shin on as you wheeled it under your desk.

The Alto was interactive, which meant instantly responsive to the
users demands. Contemporary computers communicated with their
users indirectly, through punch cards or teletypes so slow and awkward
that a single bleak exchange of query and response required days to
complete. It was like trying to sustain an urgent conversation by Morse
Code. But the Alto would communicate with its user via a full-sized
TV screen that could display text and images mere nanoseconds after
they were typed on a keyboard or drawn with an electronic device.

One more thing: Each Alto was to serve a single individual. This was
a revolutionary concept to users whose experience consisted exclu­sively of sharing the precious resources of university mainframes with
hundreds of other users. With the Alto there was to be no waiting in
line for a turn to run one’s own program. To use a term coined by Alan
Kay, the PARC scientist who was one of the machine s principal conceptualizes, the Alto was to be a "personal computer."

Every one of these specifications violated the accepted wisdom of
computer science. Computers were big because their hardware cir­cuits took up room. They were slow because they were serving scores
or hundreds of users at once. And they were shared because digital
technology was so expensive its cost had to be diffused among many
users per machine. It was the same rationale by which the airlines cov­ered the cost of aircraft and fuel by transporting 300 passengers at a
time in Boeing 747s. One computer per person? To contemporary
designers this seemed an act of outrageous profligacy. The computer
memory necessary to support a single user would cost nearly ten thou­sand dollars. Squandering so much money would be like giving every
passenger from Boston to San Francisco an individual plane.

But to Thacker and his colleagues such objections missed the point.
The Alto aimed to be not a machine of its time, but of the future.
Computer memory was horrifically expensive at the moment, true, but
it was getting cheaper every week. At the rate prices were falling, the
same memory that cost ten grand in 1973 would be available in 1983
for thirty dollars. The governing principle of PABC was that the place
existed to give their employer that ten-year head start on the future.
They even contrived a shorthand phrase to explain the concept. The
Alto, they said, was a time machine.

Thacker had spent much of the Alto design phase working out ways
to make things smaller while retaining just enough memory and power
to run complex software while simultaneously keeping the display
active. In quest of efficiency he lifted tricks and shortcuts from every
obscure corner of engineering science. Hardware added mass and
slowed the system down, so wherever he could he replaced hard-wired
circuits with miniature software programs called "microcode." This
allowed him to wring bulk out of the design by jettisoning circuit
boards like a balloonist dropping sandbags to gain a few more precious
feet of lift. He knew his design was spare; he was just not sure it
worked. Now the moment had come to find out.

The Alto's operating software had not yet been written, so its brains
resided temporarily in a commercial minicomputer called a Nova,
which was cabled to the Altos back panel like a resuscitator to a
comatose patient. A few members of the lab had crafted a sort of ani­mated test pattern by converting several drawings of
Sesame Street's
Cookie Monster into sequences of digital ones and zeros. Thacker
flipped a switch or two and the bitstream flowed over the cables from
the Nova into the Alto's own processor and memory. There it was
reordered into machine instructions that governed which of the dis­
play
screens half-million dots,
or
"pixels,"
were
to be turned on and
which were to be left dark.
If it
worked
properly
this
process
would
produce the series of test
images
in
black
outline against a glowing
white background.

Everyone's eyes focused on
the
screen
as
it flickered
to
life. Sud­denly the pattern appeared.
As
the
group
watched, transfixed,
Cookie
Monster
stared back at
them,
shaggy
and
bug-eyed, brandishing its
goofy grin, flashing upon the
screen while
holding the
letter "C"
in one
hand
and a cookie in the
other.

That the image itself
stood in absurd
counterpoint
to
the
sheer
power of the technology did
not matter. The
message was not in the
content, any more than the
world-altering
significance of
the
tele­phone could have been found one
century
earlier within the literal
meaning of the words,
"Mr. Watson, come here. I
want
you."

They
understood that just
as Alexander
Graham
Bell's
phrase had
once
been shot from one
point to another by
electrical
impulses
har­
nessed
in a brand new way,
so had the Cookie
Monster
been painted
onto a phosphorescent screen
by an entirely new
power:
Not drawn by
hand, but created via a
stream of electrical pulses
mapped onto mem­ory chips as digital bits and
read
out
again as
a moving image.