Dealers of Lightning (77 page)

By 1989 the architecture that Liddle once foresaw lasting ten years was
already a relic. That year
Computer
magazine published an article by sev­eral of the machine's original designers entitled, "The Xerox Star:
A
Ret­rospective." Among its features was a roll call of lessons the team had
learned from bitter experience: "Pay attention to industry trends..
..
Pay
attention to what customers want. . . Know your competition."

Blinded by their own technology, the Star's designers had been almost
entirely unaware of the coming revolution of cheap
PCs—
the equivalent
of scaled-down Altos, as opposed to the scaled-up Star. They did not see
it coming until the moment
IBM
announced its blockbuster.
At
that
point it was too late."It was a disaster beyond words," Belleville said later, "because the
world was already different."

Bill Spencer’s introduction
to PARC
was jolly enough.
On
the day he assumed
leadership
of the
Integrated Circuits
Lab, his new colleagues
welcomed
him
with a
green
cake:
It was St. Patricks Day
1981.

But the
charm rapidly paled.
As
he got to know the
place better,
he
grew
appalled at the working atmosphere
at
3333 Coyote
Hill Road. The
lab
that a
few
short years before had been the most exciting research
venue in the country, if not the world, now appeared to be
in
the grip of
some malignant virus. Whole departments shunned each odier,
except
to
fight over resources and belittle each other’s work.
Everywhere
you
looked you found people harboring mysterious grudges
and grotesque
suspicions.

The
most routine personal transactions were like minefields.
Not
long
after
his arrival Spencer, believing that he ought to be in
touch with his
"customers"—that is, the other labs at
PARC
that would
be
using his
chips
—
sent out an e-mail message inviting several lab chiefs to a friendly
meeting to toss around ideas. Everyone accepted.

A
few days later Bob Taylor graciously offered the Computer
Sci­ence Lab
commons as a venue for the get-together.
A
gratified
Spencer sent out another round of messages informing his guests
of the new location. The next few hours brought a cascade of cancel­lations.

Bewildered, he picked up the phone and dialed Lynn Conway, the
first respondent.

"What's going on?" he asked. "Why aren't you coming?"

"It's CSL," she responded. "I'd be physically assaulted if I went in
there."

"That's ridiculous!"

"Think what you want," she replied simply. "But I won't set foot in
that lab." The meeting never took place.

The more Spencer investigated, the grimmer he found the situation.
Pake's strategy of isolating Taylor by sundering PARC in two had exacer­bated the schism, not resolved it. "At every staff meeting power grabs and
disputes over resources dominated the discussion," he observed. "Com­puter resources, office space, budget—you name it, they fought over it.
It was so acrimonious and divisive that the center had ground to a halt
while people spent all their time defending their turf."

The contention was general, but Taylor appeared to stand at its center.
His claims on PARC's resources, based on his presumption that any work
done outside his lab was scarcely worth doing at all, forced the others to
dig in against any change that threatened even tangentially to shift the
balance of power. Further roiling the atmosphere were rumors that he
was shopping the Computer Science Lab as a wholesale package to a
Xerox competitor (speculation focused on Hewlett-Packard, which was
building a research center on a Palo Alto hillside just over the ridge from
Coyote Hill). Taylor always steadfastly denied the story, but the suspicion
never entirely waned.

Yet to blame Taylor entirely for the dysfunctional atmosphere would
be unjust. To a certain extent the internal conflicts grew out of the
ineluctable life cycle of a creative institution. In one form or another
the same tension had been present almost since PARC's founding; for
most of the first decade it simply had been channeled more fruitfully.

"Up until 1977 or so the lab conflicts seemed more dynamic and
actually productive in stimulating healthy competitiveness and innova­tion," Lynn Conway observed. As projects matured, people moved on,
and outside organizations assumed the exciting work of bringing
PARC's innovations to market, there was less of substance to fight
about within the walls. And as is known to veterans of all intramural
politics—whether they take place on a university campus or inside a
corporation—the less at stake, the more vicious the battle.

At PARC the excitement had faded but the pressure cooker remained.
The internecine competition took on a reflexive quality, as though people
were going through the motions. Severo Ornstein recalled a telling inci­dent soon after CSL spun off its graphics programs into a separate lab
under Chuck Geschke and John Warnock. "These were guys we had
worked very cohesively with," Ornstein recalled. "But not two weeks
later I heard someone in the hallway refer to them as 'they.' Chuckling, I
thought, 'This is how wars start.'"

Xerox's indifference to PARC's work also deserves some of the blame
for nursing these unproductive conflicts. Futures Day had underscored
how divergent were the vectors between parent and offspring. More
depressing miscues followed, such as the clash of expectations over
"Interpress."

Interpress was a programming system that aimed to reconcile the dif­ferent image resolutions of computer screens and laser printers. This dif­ference often resulted in documents that looked perfect on an Alto dis­play, but emerged as gibberish on the higher-resolution printers—which
of course made a mockery of Bravo s WYSIWYG feature. After years of
painstaking work and several intermediate versions, Geschke, Warnock,
Bob Sproull, Lampson, and others had at last invented a so-called page
description language allowing printers of any type to output a document
that accurately reproduced its on-screen representation, regardless of
the incompatibilities between display and laser. But the difficulty of per­suading Xerox to integrate Interpress into its laser printers and odier
typographical products made the process of actual invention look like a
cakewalk.

"We spent months traveling around to all the divisions within Xerox
and back to corporate selling this idea," Warnock recalled. The program
was extensively rewritten to meet objections by the various divisions.

Finally, in 1982, the company agreed to make Interpress a standard com­ponent of its entire program line—but refused to announce or release it
until every product could be re-engineered to take advantage of it. The
upshot would be an unendurable delay.

"I knew that would take at least five to ten years, and really it was just
never going to happen," Warnock recalled. "Chuck Geschke and I had
a conversation in his office and said, 'You know, we need to go do
something else, because we've spent two years of our lives trying to sell
this thing and they're going to put it under a black shroud for another
five.' You were seeing PCs get announced, and Apples, and you kept
asking yourself, 'When is all this great stuff going to see the light of
day?' And you'd think about the Xerox infrastructure and the process it
would have to go through to get into products, and it became sort of
depressing."

A short time later he and Geschke resigned to start their own com­pany, Adobe Systems. After a couple of false starts they settled on a
business plan that would ultimately turn Adobe into a $l-billion-a-year
enterprise: the refining and marketing of a new "typesetting," or page
description, language, along the lines initially developed in Interpress.
This language was Postscript, a typesetting system first bundled with
Apple printers. Postscript allowed computer-generated documents to
be printed on laser printers, linotype machines, and virtually every­thing in between. In its later versions it proved able to handle graphics
and color with amazing fidelity. Within a few short years it became the
de facto
standard of computerized typesetting, and a dominating rival
of Interpress itself—yet another technology that broke free from clois­tered PARC to flourish on the outside.

Thus conflicts that might have been channeled into helping a new
product reach market got turned inward instead. But Spencer, who
arrived at a point when PARC's most impressive achievements were
already behind it, had not lived through these battles with the corpo­rate mindset. He found it difficult to view the internecine tensions as
anything other than childish squabbling. In any case, none of the quar­reling had been resolved by late 1982, when George Pake asked him to
succeed Hall as head of the half-PARC "Science Center."

To Pake's surprise, Spencer turned him down. The bifurcation of
PARC was untenable, he said, and he would not be interested in manag­ing one half of a house divided against itself. He did, however, offer to
step in as PARC director—if and when Pake saw fit to reunite the two
halves.

Pake agreed without hesitation, even with relief. He had found it
painful to watch the bickering continue even after the partition. More­over, he admired Spencer not only for his skills as a technical administra­tor but for his apparent ability to co-exist with Bob Taylor. Spencer and
Taylor played tennis together every Saturday, their wives and children
were friendly, yet Pake could see no sign that Spencer had bought into
the CSL orthodoxy. He dared hope that after more than ten years he had
finally found the one man on Earth who could keep Taylor in his place.

There was much about Taylor that Bill Spencer respected. In all his
years in research he had met few managers so adept at running small
or medium-sized teams. Taylor's method of forging personal bonds and
keeping up a jealous defense of his own role as laboratory autocrat was
one Spencer thought would fail miserably in a large lab. But there was
no denying it worked marvelously inside CSL.

"He never traveled," Spencer recalled. "He would come in every day
at ten in the morning, park his BMW in the same space every day, and
pick up his badge from the security guard. He would go in his office,
break out a Dr Pepper, and for the next eight to ten hours individually
touch every member of his lab. As a consensus manager he was extra­ordinary."