The Xerox “Star” a retrospective

The Xerox “Star”

• A Retrospective

Overview

• What is the “Star”

• Features – What Makes it Unique

• History of Star Development

• Xerox PARC

• Lessons Learned

• Although Star was conceived as a product in 1975 and released in 1981, the history of its development dates back three decades.

Memex(1945)

• Vannevar Bush describes his vision of a personal, desktop computer.

• This was when computers were new, room-sized and used in military applications.

• The idea languishes because of “insufficient technology and imagination”.

Memex(1945)

• Vannevar Bush describes his vision of a personal, desktop computer.

• This was when computers were new, room-sized and used in military applications.

• The idea languishes because of “insufficient technology and imagination”.

Sketchpad(1960s)

• Ivan Sutherland builds an interactive graphics system that allows a user to create graphical figures on a CRT display using a light pen.

• These figures were treated as objects and could be moved, copied, shrunk, expanded, rotated etc.

• Sketchpad heavily influenced Star’s user interface and graphic applications.

Sketchpad(1960s)

• Ivan Sutherland builds an interactive graphics system that allows a user to create graphical figures on a CRT display using a light pen.

• These figures were treated as objects and could be moved, copied, shrunk, expanded, rotated etc.

• Sketchpad heavily influenced Star’s user interface.

NLS(1960s)

• Douglas Engelbart establishes a research program at Stanford Research Institute(SRI).

• Experiments with different types of displays and input devices.

• Invents the mouse.

• Develops a system commonly called NLS.

NLS(1960s)

• NLS was different

• It used CRT displays and not teletypes.
• It was interactive(online) when almost all computing was batch.
• Full-screen oriented when other interactive systems were line-oriented.
• It had a Mouse!
• First system to organize information in trees and networks.

The Reactive Engine(1969)

• Alan Kay, a graduate student at the time, in his dissertation, developed many ideas that found their way into Star.
• Later, he brought to fruition these ideas in the Smalltalk programming language.
• Like the developers of NLS, he realized “interactive applications do not have to treat the display as a glass teletype and can share the screen with other programs”

Xerox PARC(1970)

• Palo Alto Research Center - several laboratories devoted to basic and applied research in materials science, laser physics, integrated circuitry, CAD, user interfaces etc.

• Researchers at PARC were fond of the slogan - the best way to predict the future is to invent it. So they began searching for a new approach to computing.

Xerox PARC(1970)

• Among the founding members of PARC was Alan Kay, who liked the novel approach to HCI followed by NLS.

• As a result, PARC hired several people who had worked on NLS.

• In 1971, PARC signed agreement with SRI licensing Xerox to use the mouse.

• One major outcome of this new approach was the Alto.

Alto(1972)

• Mini-computer with removable 2.5 mb hard disk pack.

• 128-256 kb memory.

• Microprogrammable instruction set.

• Full-page bitmapped graphic display.

• 50 kb of high-speed display memory.

• A mouse.

Ethernet

• Standardized layered communication protocols.

• Used to network the newly built Alto computers.

Smalltalk

• Language and programming environment.

• Refined and solidified concepts of object-oriented programming.

• Most importantly for Star, Smalltalk demonstrated power of:

• Graphical, bitmapped displays;
• Mouse-driven input;
• Windows and
• Simultaneous applications.

Pygmalion

• Doctoral thesis project of David C. Smith.

• Demonstrated

• programming is not necessarily textual; it can be done by interacting with graphical elements on screen;
• computers can be programmed in the language of the user interface;
• the idea of using icons for direct manipulation.

Bravo, Gypsy and BravoX

• Charles Simolyn and Butler Lampson write advanced document editing system called Bravo(1976-78).[WYSIWYG]

• Exemplifying modelessness, Larry Tesler writes another text-editor: Gypsy.

• Simonyl and others add style and users’ ability to control the appearance of their documents: BravoX.

Draw, Sil, Markup, Flyer and Doodle

• Draw and Sil: Graphical object editors that allowed users to construct figures out of selectable, movable, stretchable geometric forms and text.

• Markup: Bitmap graphics editor(like paint).

• Flyer: Another paint program written in Smalltalk for Alto.

• Doodle: The above inspired Doodle for a later machine, eventually evolving into Viewpoint’s Free-Hand Drawing application.

Laser Printing

• Invented at PARC.

• Press page-description language developed (uniform way to describe output to printers).

• Press -> Interpress (Xerox’s commercial page-description language) -> Postscript(Adobe’s page-description language).

Laurel and Hardy

• Though e-mail was not invented at PARC, it was made more accessible to non-engineers through Laurel(display-oriented tool for sending, receiving and organizing e-mail).

• Laurel inspires Hardy for a successor of Alto.

Officetalk

• Prototype office automation system.

• Supported standard office automation tasks.

• Tracked jobs that went from person to person in an organization.

Star(1981)

• Contrary to popular belief, Star was not developed at PARC.

• Separate organization called System Development Department(split between southern, El Segundo and northern California, Palo Alto).

• SDD used Mesa, a ‘dialect’ of Pascal as the primary product programming language.

Star-Hardware

• 8000 series network system processor
• 384 kb of real memory
• A local harddisk - 10,20 or 40 mb
• 17 inch display
• Mechanical mouse
• 8 inch floppy disk drive
• Ethernet connection.
• $ 16,000 only!

Star-Software

• Mammoth task to integrate all the software described above into one coherent design.
• About 30 person-years went into the design of the interface, functionality and hardware.
• Objects and actions: objects that users can manipulate and actions that software provided for this manipulation.

Star(1981)

• Since the SDD was split in two locations, it had to come up with an effective means of communication.

• Ethernet: 56kbps lease line.

• Design and Prototyping - Palo Alto;

• Implementation - El Segundo.

Tajo/XDE

• Since the machine was developed in parallel with the software, it was not available initially as a development platform.

• So early prototyping and development done on Altos.

• When the 8000 series workstations were available, the systems group developed XDE, known internally as Tajo.

Success?

• In spite of such exemplary vision, Star is considered a commercial failure. So why did Star fail?

• Too expensive at $16,000?
• Ahead of its time?
• Not marketed well?
• Too monolithic?

Lessons from experience

• Pay attention to industry trends

• Pay attention to what customers want

• Know your competition

• Establish firm performance goals

• Avoid geographically spilt organizations

• Don’t be dogmatic about Desktop metaphor and direct manipulation

What was right

• Iconic, direct manipulation, object-oriented user interface

• Generic commands and consistency

• Pointing device

• High resolution display

• Good graphic design

• 16-bit character set

• Distributed personal computing