The
thought of dealing directly
with Taylor
as one of his laboratory
directors made him blanch.
Yet inserting
someone
new as
Elkind's
replacement as
CSL
chief might be taken
as an
intolerable insult
by Taylor's
coterie.
Pake
chose to buy time
by sticking to
the status quo, informing the lab that he had promised Elkind
his
old job back
and
owed it to
the
man to be true to his word.
Meanwhile
the anti-Elkind talk left
loyalists
like Dan
Bobrow
deeply
perturbed.
Thoughtful, a bit naive, but
optimistic
that by bringing well-
meaning people together in good faith he could find a middle ground satisfying everyone, Bobrow arranged for
a group
of
CSL
engineers to
pay
a
formal visit to Elkind's task force quarters
a
short distance down the hill
from
PARC.
"I
was trying to smooth things out," he recalled.
"To
me,
Jerry
was a
superb manager who always asked good hard questions and
was
very supportive.
I
didn't see why he couldn't continue in his role, and
Bob
in his.
I
didn't realize how much of a power struggle there was behind the
scenes or how deep the division was, in the sense that some people
thought Bob was a listener and a pusher of their vision and that Jerry had
his own views."
He was about to find out.
Most
of the other participants viewed the
meeting not as an effort at reconciliation, but as the vehicle for an ouster.
The meeting opened in a strained atmosphere. Elkind was facing the
most powerful members of his laboratory, including Lampson,
Thacker, Mitchell, McCreight, Chuck Geschke, and Severo Ornstein.
Their message, as McCreight recalled, was: "How about finding something else to do?"
"I'm pretty sure he'd been warned about why we were there to see
him," Mitchell recalled. "I don't think this was completely out of the blue.
He
tried his best to be very calm. He didn't get mad and rail at us or anything. He tried to start it off with some bonhomie, trying to say, 'Gee, I'm
glad to see you,' and everything. But man, you could see him shaking in
his boots just under the surface, because he was being rejected.
It
was
very hard for him."
Elkind listened, but did not capitulate.
As
the meeting broke up, he
said: "It's up to George Pake to decide what should be done, and he
invited me to come back to CSL."
Taylor's role in this challenge to Elkind's authority is hard to establish.
He has always contended he had nothing to do with dispatching the delegation, and Elkind declines to hazard an opinion. "I don't have any idea
whether Taylor encouraged those people to come," he said. On the other
hand, Lampson, Thacker, McCreight, Geschke, and Ornstein were all
among Taylor's "Greybeards," a sort of kitchen cabinet he relied on for
technical and administrative advice. And it is far from implausible that
after seven years as associate director—especially following Elkind s yearlong absence
—
Taylor might think he deserved to be
de jure,
not
de facto,
director of CSL.
Bobrow speculates that Taylor had finally recognized the limitations of
being the power behind the throne. "When Jerry went off for a year," he
said, "Bob got a sense of what it meant to actually have a throne."
In any case, Elkind did return to CSL as its director for a brief, uneasy
period. "It was apparent that Bob had won the battle," Bobrow recalled.
All that remained was to find Elkind some face-saving retreat. Finally,
when Jack Goldman parlayed the success of Futures Day into a chance to
deploy the Alto as a "market probe" of the commercial acceptability of
small computers, the path was clear. With Jerry Elkind appointed as its
chief, the new Advanced Systems Division was authorized to put the Alto
out into the world.
Elkind's transformation from the skeptic who vetoed Alan Kay's proposal for a small computer into proselytizer-in-chief of the Alto actually
dated back to the first time he had seen Bravo running on an Alto screen,
long before ASD's creation in January 1978. "I thought to myself, seeing
the machines work and seeing the Bravo stuff, that it was just smashing."
The Taylor group's distaste for his management style notwithstanding, he had turned into one of the machine's hardiest champions. Now
his enthusiasm was fired even more by the computer's commercial
potential. Like Ellenby, Elkind was a bruised veteran of the 1976 battle between the Alto III and the Dallas-built 850 word processor. "A
lot of us were feeling very frustrated," he said. "Like everyone, I was
very anxious to get this stuff out."
The group he attracted to his new venture was a remarkable collection
of talented malcontents. There were Ellenby and Tim Mott, who had felt
the sting of headquarters indifference after Futures Day as well as the
taste of power that comes from pulling off a hit performance on a lavish
$l-million-plus budget. Another recruit was Charles Simonyi, who had
made himself an undesirable at SDD by openly denigrating the biggerism of the Star. At ASD he found an entirely different culture and
worked with renewed vigor, incorporating the Gypsy interface in a
rewritten Bravo specifically tailored for the commercial Altos and known
as BravoX.
Elkind took seriously ASD's mandate to develop outside markets by
selective sales and leases of "pre-products." Taking over Ellenby's El
Segundo-based Special Programs Group, he commandeered several
projects they had already started, including one to produce word processors for Sweden's government-owned telephone company. The Swedes
got Altos instead.
So, too, did the Carter White House, which awarded ASD a contract to
create a document and file system for its information office. Altos running BravoX were installed in 1978, giving the otherwise ineffectual
Carter Administration the distinction of being the first in history to use
personal computers for word processing. The Senate and House of Representatives followed suit. "It turned out that if the White House was
going to do it, then the House of Representatives and the Senate had to
do it," Elkind remembered. "That was politics: As long as they all did it,
they could all get the funding. If the White House tried to do it without
the House and Senate, then they wouldn't get the money."
Other customers included the Atlantic Richfield Company in Los
Angeles and the Seattle headquarters of Boeing Corporation. More
machines went to Xerox divisions outside PARC, where demand was so
fierce that Elkind, Pake, and a research executive from Rochester sat as a
committee of three to allocate the scarce machines to employees of their
own corporation.
For a short time Elkind's people reveled in the thrill of actually getting
PARC technology out to the marketplace. "After we sold, like, twenty
machines, we had this wonderful dinner in Palo Alto," Simonyi recalled.
"I made this wonderful speech about how today it's ten machines, and
next year it will be one hundred, and then ten thousand. Everybody was
laughing from sheer hubris."
The inevitable crash landing awaited just over the horizon. As ASD
placed more Altos out in the world, John Ellenby was receiving the all-
too-familiar impression that he was hitting a brick wall. Rather than
absorb the lesson that these were marketable products, Xerox management seemed largely to view ASD as a sort of stalking horse for the Star.
"It was a sacrificial thing," Ellenby said. "It was, 'Let's show some customers we're coming'
—
but it pissed off the customers that we didn't follow through with it." He compared himself more and more to a member
of Eisenhower's expendable vanguard, like the decimated Canadians at
Dieppe in 1944: "Market probes were good if they got follow-up. But you
don't just go and land troops on a beachhead without a means of getting
further inland or pulling them out."Adding to his frustration were the mounting signs that SDD was faltering.
The
Stars launch date
was
now off
into
the next decade, and
not a
sure thing even then.
Ellenby saw a unique
opportunity for
Xerox
to offer a computer product
with proven market
acceptance that could
be ready
in
a matter of months.
"Basically, I
said if you
really
want to
get some products out there
quickly, then we
can do it.
We
know people are willing to pay for these
things because
that has come out in the
probes.
We
do need to perform
the engineering
that makes them high-
quality, and we need to put in place
the
manufacturing. But
we
know
how to do all that."
In
late
1978
he consigned to paper
a plan to
manufacture an updated
Alto
out of electronic components
available off
the shelf, to
be
bundled
with BravoX
and a suite of SSL-developed
office
systems software known
as OfficeTalk,
as well as a black-and-white
laser
printer called the
Pen
guin.
The
result was an inch-thick
document
whose title, "Capability
Investment Proposal," aimed to stress the
point
that Xerox should implement the
Alto
program on a small scale,
but with
serious follow-through.