This is brought to you by the
Node Your Homework project, and a generous donation from the Chubb Group.
The Influence of Xerox PARC on Modern Computing
At the end of the 1960's, executives at Xerox worried that computers and other technologies were going to replace paper as the primary information source in the office of the future. Xerox created the Palo Alto Research Center (PARC) to develop plans that would lead Xerox to dominate the paperless office envisioned for the year 1990. Founded in 1970, Xerox PARC quickly became a haven for advanced research, populated with many of the best scientists and engineers of the time and a seemingly limitless budget. Innovative technologies that were developed at PARC during the 1970's have had a great impact on modern computing systems. Specifically, the laser printer, Ethernet, WYSIWYG word processing, and object oriented programming have had a far-reaching influence more than thirty years later.
In the late 1960's, Gary Starkweather, a physicist specializing in optics and lasers, had an idea that lasers could be used to "paint" an image onto a xerographic drum with greater speed and precision than ordinary white light. He began experimenting with this idea in his spare time while working in Xerox's product development lab in Webster, NY. He determined that a laser could address a photosensitive drum with enough speed to print microscopic dots as fine as 500 to the inch, each one corresponding to a bit of digital data. Starkweather's fellow workers and management thought his work was impractical and lunatic, and his section manager ordered him to stop his research or he’d be terminated. He managed to convince the research executives at Xerox headquarters to allow him to transfer to PARC.
At PARC Starkweather was assigned the task of modifying a Model 7000 copier, capable of producing 60 copies a minute into a machine that could print on paper almost any image a computer could create. Solving the problem at PARC took eleven months. The printing mechanism, named SLOT – Scanning Laser Output Terminal, was finished in November 1971.
The building of SLOT solved only half the problem of how to convert digital images to marks on paper – the back end, how to apply toner once the image was delivered to the laser beam. The front end involved translating the computer's images into something the laser could read. This issue was solved by Ron Rider, a newly hired engineer at PARC. The Research Character Generator (RCG) accepted a digital file from a computer, evaluated it scan line by scan line, and told the printer which dots to print at which point. The two pieces were successfully brought together in early 1972.
Bringing the experimental new printer to a commercial product was a long and difficult process due to many problems in the Xerox bureaucracy. The laser printer was finally launched in 1977 as the 9700 printer. It became one of Xerox's best-selling products of all time, and was the biggest success financially for Xerox of any development at PARC. The laser printer is the most common type of printer currently used in most environments where fast, reliable and good quality computer printouts are required.
When Robert Metcalfe, a networking expert from MIT, joined the ranks at Xerox PARC in June of 1972 he was charged with the task of linking PARC's time-sharing system to ARPANET (precursor to the Internet). After he finished that project he moved on to the challenge of finding a way to connect PARC's Altos (the innovative personal computers developed and used at PARC) to each other in a local network. The system to be developed had the following requirements: it had to be inexpensive, simple and reliable, easily expandable, fast, and be able to link hundreds of computers. No existing networking schemes fit the requirements adequately.
After months of studying the problem, Metcalfe devised a system which borrowed many ideas from a radio network called ALOHAnet, developed at the University of Hawaii and ARPANET. Metcalfe's system used a passive medium, simple collision detection, and had no central server to control transmission rates. He designed a scheme by which each station would listen to the line and stop transmitting the instant it heard any interference. Metcalfe named the system Ethernet and published a memo describing the system in May of 1973. In the Ethernet system, messages were sent broken up into packets containing the destination address, source address, message contents, and a verification checksum. Receiving stations would listen to the incoming packets, determine if the destination address matched their address, and handle the message accordingly. If the transmitting station detected any interference during the transmission, it would instantly stop sending, wait a random amount of time, then send again.
Robert Metcalfe was joined by David Boggs to implement the Ethernet idea as hardware devices for the Alto. When the Ethernet devices were completed months later, they were added as options to newly built Alto computers. Acceptance of the technology was slow at first, the people at the research facility preferred to transfer files manually by disk. When the laser printer arrived at PARC, the atmosphere towards Ethernet changed dramatically as researches saw the value of network printing at the push of a button. At PARC the local network was used for file sharing, e-mail, remote booting of the Altos, and network printing. The first version of Ethernet transmitted data at a rate of 2.67 megabits per second (mbps), but a later version increased the transfer rate up to 10 mbps.
Ideas used in PARC networking protocols like the PARC Universal Packet (Pup) influenced the creation of TCP/IP, the most common networking protocol used today. The Ethernet specification was released as a joint venture of Xerox, Digital Equipment Corporation and Intel in 1980. For a small licensing fee, any computer manufacturer could create devices to work with Ethernet. This successfully beat IBM's proprietary Token-Ring networking technology in terms of technical superiority and flexibility. Robert Metcalfe later left Xerox PARC to found the company 3Com, to lead the market for Ethernet devices. Ethernet is by far the most common technology for local area networks in use today.
A year after the creation of the Alto, the computer still lacked any compelling software to make it useful. Butler Lampson, a senior researcher at the Computer Science Laboratory at PARC, decided to design a text editor for the system to help make the Alto useful. Lampson's design for the text editor was implemented by Charles Simonyi, a young Stanford Ph. D from Hungary. The program, called Bravo was originally designed to be a basic text editor.
As he refined and enhanced the program, Simonyi figured out how to encode such characteristics as fancy typefaces, odd margins, and page numbers. Bravo allowed the user to see on the Alto’s graphical display the document almost exactly as it would appear when printed out. The document was shown with underlining, bold face, italics, and fonts of various styles and sizes. This capability was dubbed as "What You See Is What You Get" – WYSIWYG.
Bravo had a very non-intuitive user interface. The commands were complicated, difficult to learn, and prone to being misapplied. The editing screen image appeared flat and uninviting to an ordinary user. When Simonyi and Lampson created Bravo they decided to not work on the user interface because it would be too difficult and they didn’t have the resources.
PARC researchers Larry Tesler and Tim Mott took the power of Bravo and made it easy to use with by creating of the program Gypsy. Mott studied how non-engineers would use a computer at a publishing company owned by Xerox. This form of user research was a new design concept at Xerox PARC. The users wanted the machine to mimic the creation and editing tasks they performed on paper. The terms cut and paste came into use through this study, referring to the act of moving a block of text to a different location in a document. This task was performed manually by editors rearranging manuscripts on paper.
Larry Tesler worked on the visual representation of the interface for Gypsy. They wanted Gypsy to have a modeless graphical interface that was much more intuitive than the arcane commands of Bravo. His ambition was to build it around icons and menus – thumbnail-sized illustrations that would perform discrete functions when clicked by the mouse, and lists of commands that could be executed at any given time.
When the program was finished in 1975, Mott brought a couple Altos and laser printers back to the publishing company. At first the users were skeptical, but everyone who used the system, was a convert within an hour. For the first time on a large scale, professional editors manipulated text on a screen and stored it on the computer rather than cutting, pasting, and marking a typed manuscript. The program was an outstanding success. Charles Simonyi later left Xerox PARC to join Microsoft, where he led the creation of Microsoft Word. He incorporated many of the ideas created for Bravo and Gypsy into the most popular word processor of today.
In 1969, most programming languages were based on the procedural paradigm. Languages like BASIC, C, COBOL and Lisp make a distinction between data and the procedures acting on them. Programming in procedural languages like these, often require elaborate combinations of procedures, and the complexity of getting such constructs to work increases exponentially with the length of the program. The visionary computer scientist Alan Kay was working at PARC to create an environment to teach children the art of programming. He found the procedural programming languages too complicated and burdensome on the programmer, so he decided to create a much simpler and elegant programming language. Kay took ideas from the existing languages Simula, Lisp, and Logo and incorporated them into his design. He created a new syntax of computer programming based on discrete modules of programming called objects instead of data and procedures. Objects combine data and procedures that will work with them, and are manipulated by sending them "messages". With objects, internal data and underlying calculations remain hidden within the object, never needing to be explicitly invoked by the programmer. Kay opted for a building block approach of layered design, built around the notions of encapsulation and inheritance, so that non-specialist users would not have to think of the underlying functionality when creating their programs. This system would eventually be called object oriented programming (OOP).
Alan Kay’s team of designers and programmers, including Dan Ingalls and Adele Goldberg, implemented his concepts, creating the language Smalltalk. The language was very slow at first, but amazingly flexible. The first functional version of Smalltalk was released in 1972, and was followed by revisions in 1976 and 1980. The language was used to create complex graphical displays for the Alto. In 1979 Steve Jobs and a group of engineers from Apple Computer were shown the Smalltalk system running on the Alto. This ideas presented in the demo had a large impact on the design of the Apple Lisa computer project and eventually the Apple Macintosh.
The concepts of object oriented programming spread rapidly through computing theory and industry in the late 1980's and the early 1990's. This paradigm is one of the foundations of modern computer science. Many programming languages were created around the powerful ideas of OOP, most notably C++ and Java.
In a relatively short period of time during the 1970's talented researchers at Xerox PARC created numerous innovations that helped shape the modern computing world. The scientists accurately predicted the needs of future computer users and the powerful systems used to fulfill these needs. With inventions such as the laser printer, Ethernet, WYSIWYG word processing programs, and the object oriented programming paradigm, the work of PARC researchers has influenced the world for the past thirty years. Each of these inventions spawned major multi-billion dollar industries, and brought new ways of thinking into the world of computing.
Bibliography
- Cringely, Robert X. (1996). Accidental Empires: How the Boys of Silicon Valley Make Their Millions, Battle Foreign Competition, and Still Can’t Get a Date. New York: HarperBusiness
- Hiltzik, Michael (1999). Dealers of Lightning: Xerox PARC and the Dawn of the Computer Age. New York: HarperBusiness
- Kreutzer, Wolfgang (1998, June 22). Introduction to Smalltalk. College of Computer Science, University of Canterbury. Retrieved April 28, 2001 from the World Wide Web: http://kaka.cosc.canterbury.ac.nz/~wolfgang/cosc205/smalltalk1.html
- Smith, Douglas K., Alexander, Robert C. (1988). Fumbling the Future: How Xerox Invented, Then Ignored, the First Personal Computer. New York: William Morrow and Company, Inc.