The Xerox Star: A Retrospective

1
The Xerox Star A Retrospective

• Presented by
• Yang Wang
• Michal Shmueli-Scheuer

2
This presentation is mainly based on

• The Xerox Star A Retrospective
• - Jeff Johnson, Teresa L. Roberts (US West
  Advanced Tech)
• - William Verplank (IDTwo)
• - David C. Smith (Cognition, Inc.)
• - Charles H. Irby, Marian Beard (Metaphor Com.
  Sys.)
• - Kevin Mackey (Xerox)

3
A birds-eye view of this presentation

• Part I What is Star?
• - Brief introduction
• - Distinctive features
• Part II History of Star development
• - Pre-Xerox, Star, and ViewPoint
• - Lessons from experience

4
I. What is Star?
5
Brief introduction

• Xerox introduced the 8010 Star Information
  system in April of 1981.
• Star was designed as an office automation system
  where users would produce, retrieve, distribute,
  and organize documentation, presentations, memos,
  and reports.

6
Assumptions goals

• Assumption 1
• Users are only interested in getting their
  work done and not at all interested in computers.
• Goal 1
• To make the computer as invisible to users
  as possible.

7
Assumptions goals (Cont.)

• Assumption 2
• Stars users would be casual, occasional
  users rather than people who spent most of their
  time at the machine.
• Goal 2
• Star should be easy to learn and remember.

8
Stars features

• Machine and network level
• Window and file manager level
• User interface level
• Document editor level

9
Stars features

• Machine and network level
• Window and file manager level
• User interface level
• Document editor level

10
Machine and network level

• Star was designed primarily to operate in a
  distributed computing environment.
• All personal workstations would be connected via
  Ethernet and would share access to file servers,
  printers..
• Use 2-button mouse

11
Stars features

• Machine and network level
• Window and file manager level
• User interface level
• Document editor level

12
Window and file manager level

• Windows
• Star allowed several programs to display
  information simultaneously in separate windows.
• Windows dont overlap because early testing
  showed that users spent a lot of time adjusting
  windows.
• Integrated applications
• Text, graphics, tables, and mathematical formulas
  are all edited inside documents.
• Star was not fully integrated in this sense.

13
Stars features

• Machine and network level
• Window and file manager level
• User interface level
• Document editor level

14
User interface level

• Desktop Metaphor
• Generic Commands
• Direct Manipulation and Graphical User Interface
• Icons and Iconic File Management
• Few Modes
• Objects have properties
• Progressive Disclosure
• Consistency
• Emphasis on Good Graphic and Screen Design

15
Desktop Metaphor

• Resembles the top of an office desk, together
  with surrounding furniture and equipment.
• The system associates different types of data
  files with appropriate programs, so users mainly
  deal with data files instead of programs.
• Example
• - users open a document, rather than
  invoke a text editor
• Make the system easy to learn and use

16
Generic Commands

• A small set of generic commands simplifies the
  system.
• Generic commands apply to all types of data, and
  each type of data object interprets a generic
  command in a way appropriate for it.
• Examples move, copy, open

17
Direct Manipulation Graphical User Interface

• Assumption casual users
• Design philosophy
• Recognition gt Recall
• Seeing and Pointing gt Remembering and Typing
• User control the system by manipulating graphic
  elements on the screen.
• Make the users easier to learn and use Star.

18
Icons and Iconic File Management

• Icons pictures of office objects
• Data files are represented by icons
• Each type of file has a characteristic icon
  representation
• Allow users to organize files spatially rather
  than by distinctive naming.
• Example
• Files can be kept together by putting them into a
  folder

19
Few Modes

• A system has modes if user actions differ in
  effects or availability in different situations
• Modes in interactive computer systems are
  undesirable because they restrict the functions
  available at any given point and force the user
  to keep track of the systems state to know what
  effect their actions will have ---Tesler
• Stars approaches to avoid modes
• Extensive use of generic commands
• Allow applications to operate simultaneously

20
Objects have properties

• Properties allow objects of the same type to vary
  in appearance, layout and behavior.
• In Star, properties are displayed and changed in
  graphic forms called property sheets.

21
Property Sheet
22
Progressive Disclosure

• Detail is hidden from users until they ask or
  need to see it
• To avoid overwhelming users with information
• Star provides default settings and hides settings
  that users are unlikely to change
• Example in property sheets

23
Consistency

• Star and all its applications were designed and
  developed in a unified way by a single
  development group.
• Star achieved a high degree of consistency.
• Example
• The left button of mouse always selects.
• The right button of mouse always extends the
  selection.

24
Emphasis on Good Graphic and Screen Design

• Windows, icons, and property sheets are useless
  if user cannot easily distinguish them from the
  background or each other
• Xerox hired graphic designers to determine the
  appearance and placement of objects
• Some principals applied
• The illusion of manipulable objects (e.g.
  scrolling window)
• Visual order and user focus

25
Visual order and user focus

• In Star, the icon selected by the user is
  indicated by inverting the image
  
• Still popular

26
Stars features

• Machine and network level
• Window and file manager level
• User interface level
• Document editor level

27
Document editor level

• WYSIWYG (what you see is what you get)
• Star documents are displayed as they will print
• Extended character set for multilingual
  capability
• The document editor is supposedly the primary
  application, thus most applications are
  integrated into the document editor.

28
History of star development
29
Memex
Simula
Reactive Engine
NLS
Sketchpad
Alto
Bravo
SIL
Draw
SRI
Smalltalk
Flyer
Markup
Press
Gypsy
Office Talk
BravoX
Pymalion
Mess
Interpress
Doodle
Tajo
Star (8010)
Lisa
Augment
Cedar
Deluxe paint
Macintosh
ViewPoint (6085)
Postscript
interleaf
Metaphor workstation
Cognition MCAE system
ViewPoint 2.0
Mac II
30
Memex- 1945

• Vannevar Bush

• As we may think.

After 18 years.
31
Sketchpad- 1963

• Ivan Sutherland
• Interactive graphic system
• Create graphical figures on a 9 inch CRT display
  using a light pen
• Geometric shapes as Objects move, copy, rotate..

32
Sketchpad- 1963
33
NLS- 1960s

• Douglas Engelbart
• Compare different devices light pen, track ball,
  joysticks, the "Graficon and Knee brace

34
NLS- 1960s

• CRT display

• Full screen oriented
• MOUSE

• First system to organize textual and graphical
  information in trees and networks.

35
Smalltalk- 1971

• Alan Kay
• children should program in Smalltalk

• Smalltalk relied heavily on graphics, animation,
  mouse driven input, windows and simultaneous
  applications.

Fact The Star wasnt written in Smalltalk !
36
Xerox PARC- 1970

• Palo Alto Research Center was established in
  order to explore technologies

• Expansion into the office systems business.

• Personal computer in a distributed environment

• ALTO
• 2.5 MB HD, 256KB Memory, bitmapped display, 50KB
  of display memory and a mouse.
• Became operational 1972

37
Xerox PARC- 1970

• Ethernet, 1973
• Bob Metcalfe
• Standardized layered communications protocol

• Provided a way of connecting computers much more
  flexibly than previously possible.

• Soon after the first Altos were built, they were
  networked together.

38
Pygmalion- 1975

• David C. Smith
• The first large program to be written in Smalltalk

• Programming can be accomplished by interacting
  with graphical elements on a screen.

• Computers can be programmed in the language of
  the user interface
• By demonstrating what you want done and having
  the computer remember and reproduce it.

Fact The idea of using icons came mainly from
Pygmalion !
39
Pygmalion- 1975
40
Other software's for Alto

• Bravo- advanced document editing system WYSIWYG,
  underline, boldface, italic, font, size..

41
Star (Xerox 8010)- 1981

• To develop Star, Xerox created the Systems
  Development Department (SDD)

• SDD was split between two locations El Segundo
  (So. Cal) and Palo Alto (No. Cal)

• El Segundo did most of the implementation
• Palo Alto did most of the design and prototyping

• Mesa, a dialect of Pascal, was the primary
  product programming language used

• The hardware was designed in conjunction with the
  software

42
Star (Xerox 8010)- 1981

• Hardware
• 384KB- 1.5MB Memory, HD 40MB, 17 inch display, 8
  inch floppy disk a mouse and Ethernet.

• Software
• Uniformity of specifications, uniform design
• Strict format- Objects and Actions

• Objects- applications and system features
• Actions- actions that the software provide for
  manipulating the objects.

• Cost- starting from 16,500

43
Problems.

• The star didnt sale as expected.
• Performance, extensibility, maintainability, cost

• It was monolithic
• The system knew about all applications, and all
  parts of the system knew about all other parts.
  It was difficult to correct problems, add new
  features and increase performance.

• The monolithic architecture did not lend itself
  to distributed, multiparty development.

• Stars infrastructure was rewritten according to
  more flexible model-

ViewPoint
44
Lessons from experience

• Pay attention to industry trends
• PARC researchers and Stars designers didnt pay
  enough attention to the other personal computer
  revolution occurring outside of Xerox.
• Release new technologies to public- become a
  standard.

• Pay attention to what customers want
• Star should have been designed from the start to
  be open and extensible by users, as the Alto was.
  Star didnt have the functionality customers
  wanted.

45
Lessons from experience

• Know your competition
• The price of a stars workstation was too high to
  compete with other microcomputer-based PC

• Avoid geographically split development
  organizations
• The implementers didnt benefit from much of the
  prototyping done as PARC.
• Lack of shared background

46
The best way to predict the future is to invent
it..
Alan Kay