Notes on the (relatively unknown) History of Pen-based Computing

• Jean Renard Ward
• Contact:jrward_at_alum.mit.edu
• Copyright © Jean Renard Ward, 1992, 1996, 2003, 2005, 2008.
• Permission is hereby given to use this information in publication, including confidential reports, provided that accompanying text clearly makes reference to the URL for this page, along with the statement:
  "Source: Annotated Bibliography in Pen Computing and Handwriting Recognition by Jean Renard Ward".

This is a short outline of a presentation on the history of "pen computing" I gave for the Pen Computing group of the Boston Computer Society, March 11, 1992. Unfortunately, I did not keep the full set of slides, which had pictures of the products and systems.

Topic threads for tonight ...

• Pen Computing is not new: "A lot of it has happened before"
  (... with selected examples, mostly from commercial development)
• Some specific shortcomings of the current state of the Pen-Computing art
  (... they are not what you might think)
• What this implies ... or, "What should we do?"

Why this "historical view" might be of interest to this audience:

• Historical developments are relatively unknown, and quite voluminous.
• Present a broad range of case-study examples.
  • Early examples were done by workers motivated by other interests, with a different perspective ...
    • man-machine-interfaces (sic),
    • human functionality and usability,
    • cognitive concerns
    • a study problem for pattern recognition
      ... not in modern "GUI" development.
• Current problems:
  • There are neglected / overlooked issues, falling under the rubric of human factors, that affect and may threaten the usability and viability of pen-based systems.
    Especially: "Electronic Ink" (hardware and software)
• Topics not addressed tonight:
  • No product demonstrations ...
  • No applications demonstrations ...
• Other presentations coming up:
  • Slate is presenting to BCS general meeting, May 21.
  • Society for Information Display (S.I.D.) is meeting in Boston in May.
    Videotape demo of certain applications.

Historical overview:

• There is a repeated theme in commercial developments, namely the
  "Lief Ericsson in 1992" scenario ---
• "I/we discovered/invented pen computing, pen computers, pen-based systems."

"The Archeology of Pen-based computing"

• 1914 Hyman Eli Goldberg, U.S. Patent 1,117,184, "Controller". ... on-line (not OCR) recognition of hand-written numerals to control a machine in real-time.
• 1938 George Hansel, U.S. Patent 2,143,875, machine recognition of handwriting.
• G.G. Neil Wright, "The Writing of Arabic Numerals"

"The Ancient History of Pen-based computing"

• 1956 RAND tablet ... invention of the digitizing tablet, for a handwriting recognition project.
  (Actualy the Stylator came first!)
• 1957 T.L. Dimond, "Reading Handwritten Characters", EJCC Claimed 97% recognition rate, and a better input device for humans than a keyboard
• 1962 Masterson and Hirsch, IRE Transactions on Human Factors and Electronics
  99.79% correct recognition.
• 1962 J.C.R. Licklider, "On-line Man-computer Communication", SJCC
• 1961/3 Sitari, Harmon, Frishkopf, et al cursive script recognition at Bell Labs
• 1966 Groner et al, "GRAIL System", RAND Corporation Text/character recognition, gesture commands, sketch recognition

"The Early Commercial History of Pen-based computing"

• 1973 Applicon Corporation / Ledeen recognizer
  Gesture-based command input, with hot-points, in a widely-praised GUI for a commercial CAD system
• 1974 SRI / Xebec Systems Incorporated
  Text input to computer with a pen
• 1976 S-I Hanaki, NEC handwriting kana/romanji
  billing machine product
• 1981 Several commercial vendors, small units, some portable
  MicroPad ImageData Telepad WriteAway many Japanese (Application areas: data-capture, data-entry)

"The Modern History of Pen-based computing"

• 1983 PC-based or -oriented commercial products
  PenPad CIC Handwriter NestorWriter
  Included "front-end" interfaces to word-processing systems, CAD/Paint systems, spread-sheet input
  (Note: these companies are still around)
  Special note: even perfect recognition does not solve the problem!
  Gould et al, "Composing Letters with a Simulated Listening Typewriter", CACM April 1983.
• 1985 "Experimental" systems
  AEG (Germany) Word-processing
  IBM (Tappert)
• 1986 Linus Technologies
  Application areas: walking data capture (nurses), financing from Baxter Medical
• 1986/7 MAC-based products: Personal Writer / Anatex

"Current Events of Pen-based computing"

• 1989 GridPAD Application area: remote data capture (surveys, inventory, shipping/receiving)
• 1990/91/92 Too many to mention ...
  "Pen OS" - PenPoint, Windows for Pens
  "Pocket" - Poqet , Sharpe Wizard
  "Tablets" - Wang/Momenta/Grid/Eden Group/SuperScript/ScriptWriter/ ...
  Application areas:
  • Highly-mobile data capture
  • Note taking / capture
  • Document capture and display
  • "Imaging systems"
  NOTE: All of these application areas involve "electronic ink", data that is fundamentally in written form (sketches by insurance adjusters, for example), that does not need to be recognized as text.
• There will be a short video of some representative applications and software
  (So there is some commercial content after all - Dang!)

Specific shortcomings of the current art:

• Electronic ink: State of the Art -- Special considerations:
  • Existing, alternative technology for data capture, note capture, and documents is paper and pen, or paper and pencil,
    ... not any current computer technology.
  • Paper-based systems are highly evolved --> equivalent to well designed.
  • "Ink" APIs and functions are not highly evolved:
    • editing
    • scaling
    • storage
    • embedding
  • Current hardware designs for electronic ink are NOT adequate, due partly to ignorance of physical effects of human use on part of designers.
• Displays (LCD): State of the Art
  • Low contrast
  • No backlight
  • LCD parallax / shadowing
  • Glare
  • Coarse resolution
  • Slow ("fuzzy motion"), especially when cold
  • Greyscale mapping
• Digitizers: State of the Art
  • Bad electronic ink vs. real ink
    Comparison: Teletype keyboard vs. typewriter keyboard action
  • Tablet electronic effects: (especially at low power!)
    Tablets produce very distorted images ...
    "it's not my signature"
    "it doesn't go where I point/draw"
    • Spiking
    • Fixed line distortions
    • Velocity errors
    • Angle of stylus
    • Poor tracking with height
  • Tablet mechanical effects:
    • Slickness
    • Tip force
    • Surface hardness
  • Stylus design:
    • corded / handed
    • surface texture
    • weight and handling
• There will be a short video showing digitizers

Summary:

So ... What does this history mean ... to us?

• What applications have been shown to be inappropriate?
• Patentability?
• Why should we think there is a market?

• There is a long historical perspective Examples for human-factors, application, market, U/I, and human-machine design study
• Old (obvious?) applications were not successful.
• New applications are highly mobile, in unfamiliar areas, and characterized by functionality such as electronic ink or mobile data access, not character recognition
  • Note takers
  • Daytimer (R) Pen Scheduler
  • PenBook (R)
• Hardware for electronic ink requires a design much more to ergonomic and human factors goals than has been the case.

Back / Zurück

Comments: jrward@alum.mit.edu