From that moment through to the public introduction in April 1981
of the Xerox 8010 office system, the "Star," this basic outline would
never change. The Star would emerge as a superb feat of engineering
and perhaps the most perfectly integrated office computer system of
all time. Liddle planned on the development phase taking five years,
but he also anticipated that the system's architecture would set a standard in the industry for at least another ten.
The vision animating the Star's designers, however, was one of the
few things about SDD that would remain stable through the coming
years. Five months after launching SDD, Don Lennox was reassigned.
His replacement was Bob Sparacino, an executive with solid engineering credentials Xerox had imported from General Motors. Sparacino
took an instant dislike to Hall and a month later Hall, astonished at
how swiftly his paranoia had been fulfilled, was dumped. (He returned
to Pake's staff.) Over the next three and one-half years SDD had three
chiefs, none of them lasting more than a matter of months, until finally
in 1978, the top job devolved to Liddle, who somehow had managed to
keep his head down amid all the turmoil. As Metcalfe later joked: "For
a while the most dangerous job to have at Xerox was to be David Liddle's boss."*
SDD, meanwhile, soon burst the restraints Lennox had placed on its
size. The organization that presented the Star to the world in 1981—and
*On the day of Liddle's appointment Hall wrote him to observe that his three predecessors had lasted exactly six, twelve, and twenty-four months in the job. Accordingly,
he predicted that Liddle would serve forty-eight. Four years later almost to the day,
Liddle mailed him a copy of the note, clipped to an announcement of his resignation.
"Harold," he wrote, "you were exactly right."
developed a basic system technology for a wide range of Xerox products—was no ten-person shop but a colossus employing 180 engineers in
Palo Alto and another 100 in El Segundo. As the designated commercial
outlet for PARC technology, the division also acquired an exalted sense of
its stature as a flag-bearer for Xerox. This partially accounted for the
Stars deliberately stately design. Its target users were not secretaries and
clerks but their bosses, who were executives and professionals—an ambition that inspired skepticism among marketing experts of the time but
accurately foretold the later evolution of the personal computer as an
office device. (This ambition also drove the Stars architects to make the
user interface as mouse-oriented as possible. A Xerox promotional
brochure would state in 1981 that the Star was "designed specifically for
professional business people with little or no typing skills.")
SDD's chiefs correctly understood that Xerox headquarters expected
a product development program to possess a certain minimum heft.
"Xerox had a hard time understanding anything that wouldn't be a
$100-million business," observed one technology manager. Others
feared, however, that in waiting for the division to unveil its fully featured product, Xerox might inadvertently miss out on myriad smaller,
but still promising, opportunities.
SDD management accordingly made earnest, if infrequent, efforts to
interest headquarters in staging less ambitious market probes. As head
of the division in 1976, for instance, the former SDS executive Robert
Spinrad attempted to persuade Jim O'Neill, the head of technology, to
introduce a downscaled low-cost Alto for the office clerical market.
O'Neill responded that Xerox was uninterested in taking what he saw
as a remote beachhead. It would be safer, he argued, to let others
establish positions in the technical vanguard. Xerox would bide its time
and overrun them later with a massive, concentrated attack. "If we put
it out right now we'll tip our hand," he told Spinrad. "While we'll have
an early success, others will look at us and come in with a well-
engineered product that can be maintained better in the field. We'll
just lose the big battle."
"My notion," Spinrad recalled, "was that we could build small and
develop up. But I lost every one of those fights. There was no way the
Xerox Corporation of that era was going to do anything but full-scale
product development."
About a year and a half after SDD's founding, Dave Liddle tempted
Bob Metcalfe back to Palo Alto. He did not find it a difficult sale, for
Metcalfe had grown disaffected with his new employer, a subsidiary of
Citibank that handled the giant banks electronic fund transfers out of
a computer center in Los Angeles.
After seven months he had succeeded in weaning the operation from
its aging custom-built card readers and onto Digital PDP-11s running
interactive software. Unfortunately, he found his human colleagues less
tractable than the machines. "I'd been promised that I would be made a
vice president after six months," he recalled. But when he demanded
that the company make good on its offer, his boss reneged on tire
grounds that he was not yet thirty. "He said, 'I can't do that, you're just
too young. It would really upset the applecart. It's no big deal.' I said, 'It's
a big deal to me."
The friendship between Liddle and Metcalfe was anchored by a multitude of common traits. They were both tall, solid men with built-in swaggers, unashamed to throw their weight around. As ex-college jocks they
worked out their frustrations on the field of play, often on each other.
Metcalfe could regularly beat Liddle at tennis, his game; but Liddle, who
had briefly played varsity basketball at the University of Michigan in the
mid-1960s, could whomp him decisively under the boards.
One time his varsity background enabled Liddle to score a decisive victory off Metcalfe. The latter, living high off the hog as a newly divorced
bachelor in Los Angeles, had secured a pair of floor seats to tire L.A. Lakers games, close enough to the action for tire toes of his shoes to nuzzle
the side line. He invited Liddle, who had long bragged of having played
ball at Michigan with Lakers guard Cazzie Russell.
"We'd all say, yeah, yeah, you were at Michigan and Cazzie Russell was
there too, yeah, sure," Metcalfe recalled. On this occasion, he and Liddle
took dreir seats on tire floor of the Los Angeles Forum just as the Lakers
came out for their warm-up. As Metcalfe watched in mute astonishment,
Cazzie Russell made a beeline for Dave Liddle. "He came over and said
something to the effect of, 'How you
doin',
Liddle, my
man,
its
been
a
long time,' then he brought over another
famous
guy like
Kareem
Abdul-
Jabbar
and it's,
'Meet
my friend
David I've
been telling you about . . .'
Well,
thereinafter
I
believed
Dave Liddle could
never lie, because he had
obviously not been exaggerating
in the slightest
about his relationship
with
Cazzie
Russell."
When Metcalfe
quit Citicorp
in a huff over
his withheld vice presi
dency,
Liddle was ready to grab
him on the
rebound.
Metcalfe,
who had
inspired awe at
PARC
by becoming
the first
well-known researcher to
leave,
now became the first to return (although technically he was returning to
SDD,
rather than
PARC). The difference
between his final
PARC
salary and
his
new wage of
$37,000
amounted to the biggest raise anyone
in the building had ever received.
The strategy
of extracting
a
raise from
Xerox
by working somewhere else for a spell was known ever after
as
the
"Metcalfe
promotion," a term that further gratified its honoree.
"I
got twenty to thirty per cent from Citibank, then
Liddle
gave me
another ten to fifteen per cent
premium to get
me back,"
he
said.
"But
the problem
was I
knew people who were making much more,
and
they were people
I
didn't think
were
that much smarter than me."
Metcalfe's
new job was primarily to upgrade Ethernet's
capacity from
th
ree
to ten
megabits per second to meet the demands of
SDD's
formidable product plan. But he was also expected to ride herd,
as a sort
of
contract administrator, on the designer of the Star's mission-critical central processor, Chuck Thacker.
As
if his historically charged relationship with Thacker were not problem
enough,
on assuming his new duties in June 1976
Metcalfe
discov
ered
that the entire project was careening off the tracks. Structurally
speaking,
SDD
was a mess. The division had two headquarters, one in
Palo Alto
and the other in El Segundo, where it had taken
over a
block
of
manufacturing facilities vacated by
SDS.
This arrangement burdened
Liddle and his cavalcade of immediate superiors with two mutually
resentful semi-organizations located five hundred miles
apart—
the
northern contingent thickly seeded with
PARC
alumni and the southern
branch staffed with reassigned
SDS
employees nursing old grievances
against
the fancy
Ph.D.s
in Palo Alto. No volume of e-mail or networked
file sharing could quench the rekindled tensions, only the personal intercession of SDD managers shuttling endlessly back and forth by air.