“Every time you click a mouse on an icon, or open overlapping windows on your computer screen today, you are using technology invented at [Xerox] PARC. Compose a document by word processor, and your words reach the display via software invented at PARC. Make the print larger or smaller, replace ordinary typewriter letters with a Braggadocio or Gothic typeface – that’s also technology invented at PARC, as is the means by which a keystroke speeds the finished document by cable or infrared link to a laser printer. The laser printer, too, was invented at PARC.”
That passage was penned by Michael Hiltzik in the introduction of his seminal book Dealers of Lightning, about the origins of Xerox PARC. And while it neatly illustrates the importance of PARC in the scheme of all things technological, it doesn’t cover it all, by any stretch of the imagination. For example, other technologies invented at PARC include Smalltalk (the first object-oriented programming language), Ethernet, VLSI (Very Large Scale Integration) [ a means of embedding more functionality in smaller bits of silicon – Ed], and the first 3-D computer graphics chip.
Where it all began
Xerox PARC was officially opened on 1 July 1970, but the tale is rich in irony because it was born of what has been dubbed McColough’s Folly. First, some background: Charles Peter McColough was the CEO of Xerox at the time. Early in his 25-year tenure as CEO, he was occupied with putting the wishes of his predecessor and Xerox founder Joe Wilson into action. Wilson had wanted to get into the then fledgling computer business and had been shopping around without success for several years.
It fell to McColough to continue that search and it eventually culminated in the acquisition of California-based Scientific Data Systems (SDS). As a business decision, it was pilloried in the financial press, and with good reason: he paid far too much for it. Indeed, based on SDS profits ($10 million in 1968) and the purchase price of $920 million, it would take 92 years for Xerox to see a return on its investment.
Hiltzik writes further: “... only someone with McColough’s total lack of understanding of the computer industry could delude himself into viewing SDS as Xerox’ entrée to a bright digital future. For the glittering numbers obscured some gloomy facts. SDS’s success at mining a comfortable niche in scientific computing would not easily transfer to the high-volume business data processing market, the ferociously competitive sector dominated by IBM in which McColough now expected Xerox to play a major role.
“SDS’s core market, meanwhile, was suffering from a serious economic downturn exacerbated by the drying up of government research funding. And there were nagging questions about the company’s research capabilities and depth of its management team.” If McColough had used the resources – financial and technical – that he had at his disposal, he would almost certainly have been advised against the deal. One of those who would have advised him against it was the newly appointed chief scientist Dr Jacob Goldman, who had left a similar position at motor manufacturer Ford to join Xerox. When he heard about the deal, he is said to have been shocked at not having been consulted. But he was quick to take advantage: six weeks after the AGM rubber-stamped the deal, he presented McColough with a comprehensive proposal for a dynamic new scientific facility.
According to Hiltzik, the SDS purchase had given Goldman the peg on which to hang his proposal. Recognising that Xerox yearned to be ranked alongside such paragons of industrial muscle as IBM and AT&T, he sketched out a corporate research centre engaged in basic science independent of any existing product group, exactly like IBM’s fabled Yorktown Heights research centre and AT&T’s Bell Laboratories. About half the staff would be devoted to advanced physics and materials research, and the rest to the new sciences of systems and computing. That, as it turned out, was almost exactly what Xerox PARC became.
Plans to action
The first order of business for Goldman was to find a suitable research director. He had just the person in mind: long-time associate George Pake, with whom he had worked on wartime projects at Westinghouse Research Laboratory 25 years earlier. The rumour mill had it that Pake had been offered, and turned down, Goldman’s former job at Ford so the latter knew that enticing Pake to join Xerox might not be that easy. That turned out to be the case: he had to wheel out the big guns – Wilson and McColough – to convince Pake to take the job.
Hiltzik reports on a conversation between Pake and McColough in which the former asked why Xerox wanted a second research centre – the company already had one that developed the technology in its copiers. The reply convinced him: “George, I think these people here in Rochester have had a heady success with xerography [Xerox’ name for its technology – Ed.]. But I’m not sure they’re adaptable enough to take on new and different technologies. If we are going to bring new technologies into Xerox, it would be better to do it in a whole new setting.”
Pake took the job in early 1970 and his first order of business was to find a site. A lengthy search ensued, during which many alternatives were considered. Eventually, he settled on Palo Alto. One mitigating factor in the decision was proximity to Stanford University – an institution with an active programme to attract fledgling technology firms into what would soon become Silicon Valley.
However, Pake’s second big challenge was his lack of familiarity with the science of digital computing – he was a physicist by training. He did, however, know one person who boasted a first-rate familiarity with the young Turks of the new discipline: Bob Taylor. As it happened, Taylor was marking time in Utah. The previous year, he had left the Information Processing Technologies Office at the US government’s Advanced Research Projects Agency (ARPA) after four years, three as director, and accepted an invitation to undertake a research co-ordination effort at Utah University.
But word on the street was that he wasn’t entirely happy in Salt Lake City. Pake invited Taylor to Palo Alto for a chat and, a few days later, Pake called him again and asked him to help build the computer lab. The job, he told Taylor, would involve recruiting an entire laboratory staff – including his own boss.
What Pake and Xerox had not counted on was the fact that Taylor had long held his own vision of distributed computing. Needless to say, he saw Xerox PARC as an opportunity to finally realise his ideas of computing with a handpicked team. He made this clear to all and sundry early into his tenure when, encountering Pake’s secretary Gloria Warner showing off her brand-new IBM Selectric typewriter, he remarked: “You know, we’re going to make that thing obsolete.”
But it wasn’t all plain sailing. Taylor and Xerox PARC were very nearly headed off at the pass by the economic slowdown that started in 1970. Xerox imposed a companywide hold on new hires and went as far as debating – at board level – the possibility of shutting down PARC before it even started.
One director stopped it. In 1971, John Bardeen already had one Nobel prize under his belt, for co-inventing the transistor with William Brittain and William Shockley in 1947 [see Brainstorm February 2007]. A towering figure in the Xerox boardroom, he reminded his fellow directors that PARC’s budget ($1.7 million at the time) was barely a fly-speck on the bottom line, but that its potential was limitless.
Other research funding companies hadn’t been as far-sighted, and when fortunes began to turn and the hiring freeze ended a few months later, the market was flooded with scientists looking for tenure. Pake, Taylor and head of personnel M. Frank Squires took full advantage of the buyer’s market for research and engineering talent. It was the beginning of the legend of Xerox PARC – one far too detailed to relate here except in summary. However, one myth deserves dispelling: it has been remarked in various texts consulted for this article that Xerox, as an organisation, failed to benefit substantially from its considerable investment in PARC. Nothing could be further from the truth.
As Hiltzik says in his introduction: “The truth is that [Xerox] revenues from one invention alone, the laser printer, have come to billions of dollars – returning its investment in PARC many times over.”
Notable PARC Inventions
➚ Laser printing.
➚ Smalltalk, the first object-oriented programming language
➚ Ethernet and client/server architecture
➚ Word-processing (Bravo) and WYSIWYG editing (Gypsy)
➚ The graphical user interface with icons and pop-up menus
➚ Very Large Scale Integration (VLSI) circuit design
➚ The first suitcase-sized portable computer (Notetaker)
➚ Linguistic technology to enable spell-checkers
➚ The Interpress page description language (a fore-runner to PostScript)
➚ Fibre-optical networking
➚ Plastic semiconductor transistor array
A History of Modern Computing (second edition) by Paul Ceruzzi, MIT Press (2003);
Dealers of Lightning: Xerox PARC and the Dawn of the Computer Age by Michael Hiltzik, HarperBusiness (1999);
Inventing the Electronic Century by Alfred Chandler, Harvard University Press (2005).
Indeed, the list of innovation milestones available on the organisation’s website indicates that Xerox’ laser printing business reached $1 billion per year in 1986 – just over 15 years after PARC was established. In addition, many of the other technologies developed at PARC also made it into Xerox products over the years – whether or not those products were successful. There were also a host of products produced at PARC merely as research prototypes.
Possibly the most well known of these was the Alto – the first networked PC and also the first PC to feature a graphical user interface (GUI) and mouse. First demonstrated in 1981 – the same year the IBM PC was launched [Brainstorm February 2008], around 1 000 units were built for internal use at PARC.
Even more notably, perhaps, numerous PARC inventions resulted in spin-off companies in which Xerox was complicit and, presumably, would have retained some shareholding. This process started as early as 1980 when Xerox, Intel and Digital Equipment Corporation jointly issued a formal specification for Ethernet, making it publicly available for a nominal licensing fee.
That same year, Optimem was spun out to commercialise non-erasable magneto-optical storage device technologies developed at PARC. The company later became Cipher Data Products. Software copyright for the Smalltalk-80 object-oriented programming language was also filed in that year and the technology was later spun out as ParcPlace Systems. Spectra Diode Labs was established in 1983 as a joint venture between Xerox and Spectra Physics. The idea was to exploit PARC’s gallium arsenide-based solid-state laser research by developing, manufacturing and marketing high-power solid-state semiconductor laser diodes.
In 1985, Synoptics Communications was spun off to commercialise fibre-optic media for Ethernet. Within three years, Synoptics had its IPO and it later became Bay Networks, which was then acquired by Nortel. That same year, Microlytics was spun off to commercialise early compression technology research at PARC by bringing artificial intelligence spell-checking software, linguistic and data compression technologies to market.
Other notable PARC spinoffs include:
• Semaphore Communications to bring encryption systems for networks to the market.
• Documentum to commercialise document management solutions.
• Inxight Software to offer information visualisation and knowledge extraction software for the internet.
• Placeware to provide a live, web-based presentation solution for field and customer communication. It will become the largest internet meeting solutions provider and is eventually acquired by Microsoft to become Microsoft Live Meeting.
• dpiX to commercialise high resolution active matrix liquid crystal display (AMLCD) monitors.
• Uppercase to commercialise a lightweight, portable document reader (PDR), including a display, computer processor, battery and network connections for document access and viewing. Later, Microsoft also acquires this technology.
• Gyricon Media (later renamed Gyricon LLC) to commercialise PARC’s “electronic reusable paper”, a document display technology that is thin, flexible and portable like paper, but can be connected to a network and reused thousands of times.
• ContentGuard, a joint venture between Xerox and Microsoft, was spun off to develop and license software for digital rights management. Xerox’ interest was eventually acquired by Microsoft, Time Warner and Thomson.
• GroupFire was spun off to commercialise almost 70 PARC intellectual property claims covering information retrieval and data mining, natural language semantic analysis, and artificial intelligence. The list above should dispel any thoughts that Xerox failed to benefit from the inventions to come out of PARC. And while it’s entirely possible that the company could have made more of its investment, that is a moot point.