Jeff Hawkins Keynote, 11/12/01

Tour of the past –

1981 – IBM PC – beginning of Intel/Microsoft duopoly.


Back then – how do you make the personal computing experience mobile?


Osborne 1 – tiny display. First portable computer. Not long term, but sold quite a few.


Grid Systems – Jeff worked there – Grid Compass – first laptop computer. Flat panel display, hinged over keyboard, integrated 1200 baud modem. Problems – very expensive - $8000, proprietary OS.


Compaq – PC compatibility was key. Very heavy, not so portable. Successful – compatible, reasonably priced.


Other types – lunchbox computers. Laptop became dominant, e.g. Compaq LTE.


Some wanted to produce something even smaller and more mobile.


1989 – Gridpad – first tablet computer. Stylus, no keyboard. PC compatible (DOS), graphical UI, focus on vertical markets (insurance, delivery people, doctors) who fill out forms.


Go – 1991 – PenPoint operating system – code name Lombard – never released. Licensed their OS. Targeted mainstream computing, had spreadsheets, WP, other software written. Also raised the attention of Microsoft – developed Windows for Pen Computing – many companies manufactured – head to head with Go. Microsoft was right – didn’t need a second OS, but also didn’t need Pen Windows. All failed.


Psion – Organizer series – no presence in North America. We didn’t pay attention. Sharp Zaurus. HP 100LX. Psion especially was successful. All keyboard based, but we were fixated on pen.


Apple Newton – great success for a short time – 100,000 in first quarter, then stopped. Was incomplete, the Edsel of the computer industry. People were not deterred.


Apple engineers left, created General Magic – MagicCap OS was desk metaphor on a mobile system. Sony, AT&T, Motorola built products. Network wasn’t ready.


MS  rolled out WinPad – didn’t bring it to market because General Magic had already failed.


MS Bob – another failure.


Windows CE – chiclet keyboard, folding screen. Lined up companies to build (same ones every time…)


Windows hard to read, slow, cumbersome UI for small screen. Productivity apps didn’t work.


1996 – Palm Pilot introduced – very small, very inexpensive, simplistic applications, easy to use, connectivity focus (synchronizing with PC). We’re just an accessory – MS left them alone.


After they were successful, MS repositioned CE – no keyboard. Ballmer at development conference targeted the Palm.


Many apps – voice recording, one handed operation. Not very good, still hard to read on black and white screen, poor synchronization. All failed.


What does Palm do? Can’t compete with MS. How to do? Build hardware, since MS doesn’t. Designed the Palm V. Brushed metal case, non-removable LiIon battery, case glued shut. Great success.


Insisted no new features on product. If no new features,  analysts can’t do a features comparison chart. Successful.


MS – auto PC – can’t build mobile devices with Windows, so put Windows on things that already move.


PocketPC – was originally called PalmPC – sued them in France, where Pilot Pen sued Palm.


MS finally got it right. Good display, fast processor, better UI.


When Apple introduced the Mac Portable – too heavy. Finally introduced Powerbook. Instant hit – pent-up demand. PocketPC had pent up demand – people waiting for a Windows handheld finally could buy one that worked.


Upshot: PC and laptop haven’t changed dramatically in 20 years. Innovation has been in handheld and mobile area.


Railroads – thought they were in the railroad business, not the transportation business. We have to avoid thinking of mobile devices as the PC industry.


The present

Most successful mobile computing device of all time? The cell phone. 1 billion users worldwide today – one person of every six.


Another successful one – the pager. Tens of millions being used.


Third – handheld (Palm, Handspring, Sharp, PPC)


What made these products successful? Four reasons:


  1. Size – the smaller the better. Pocket, purse, belt. Dumbfounded by tablet PCs – everything he knows says smaller is better.
  2. Ease of use – cell phone, pager are great examples. Handheld computers are more complex products, ease of use was harder to come by. One button access to views was added because of this.
  3. Reliability – better perform, work, not need service. Attribute of all successful mobile devices. How much time per year do people spend looking at the PC screen? 200 lifetimes every year.
  4. What are the primary applications? Talking (cell phone). Writing (pager). Organizer – maintaining contact list – how do you call them or write to them? And Calendar – when are you supposed to meet with them or talk to them? All about communications.


On Handspring road show – Woman approached her - Nokia phone, RIM pager, Handspring Visor – tired of carrying all of them. Can Jeff help? Build one device that combines all three.


Started developing the Treo.


Attributes: fairly small (actually, tiny – looks like about 2x3 inches.)


Has stylus – need one for some things, even though they tried to eliminate it.


Integrated cellular radio for voice and data.


Power switch, IR port.


Single button to mute all sounds. Can vibrate if you want it to. Should have done it a long time ago.


Headphone jack, jog switch for functions.


Four buttons:




Phone app – address book disappears into it.




Messaging – SMS by default, can be email.



Lid – contains speaker for ear and speaker phone. Also protects it. Needs no case. Answer by opening the lid. Lid is transparent so caller ID is visible on screen.


QWERTY keyboard – for creating messages. Better than Graffiti, and he invented Graffiti.


Keyboard is great way to dial the phone. Start entering name, it will dial number when recognized. Or numeric keypad.


Will sell two versions – keyboard and Graffiti.


Color and B&W – color will follow after 3 months, in March or April.


Demo Intro – Rob Itani(?), leading UI expert.


One touch speed dial dialing. If not on speed dial list, will recognize names from contact list. Also – first initial and last name will give quickest match – love to design for impatient people.


Can select which of the person’s numbers to call.


Can interact with any program on the device when on the call.


Browser application – speed dial for bookmarks. Handspring compresses and reformats any page for the screen on their proxy server.


Treo is designed for one-handed operation.


SMS – expects the full keyboard will promote the use of SMS more than the 10-key keypad.


Can place people on hold – do conference calls.


Call waiting supported – can bring a fourth people in on the conference call.


The future

Communicators will get smaller – even smaller than the Treo. Challenge of the UI – his observation is that if you make things smaller they get better.


Casio Zoomer – contemporary to Newton – tap on tabs like a Rolodex – multiple taps to get to an entry.


Palm – even smaller display – 1/3 of pixel area – made it better. First and last name, phone number on index display, no tabs because no room. Forced to let people write to find a name.


Treo – that wouldn’t work. Font too big. One line for name, one for each phone number.


One hand, more information at one time. Like remodeling a house – forced to get clever.


Web browsing – don’t need all the clutter. Want one button access to each info service – news, weather. Quicker than conventional drill-down.


Wireless applications – streaming audio/video, location based services – security is the big problem. Multiple devices, connected to cell network. Few programmable devices on these networks before – they haven’t been tested.


Worse – all carrier networks attached to the Internet. Every device is a persistently connected server, has potential to create havoc on the network.


No current hardware or software platforms can address this.


Ubiquity – everything will have a radio – access to wireless will be free. E.g. long distance land line cost has dropped dramatically. Cellular call cost has also dropped.


What is the cost of servicing the customer? Cost of building the infrastructure (poles and wires) already amortized. Cellular infrastructure will eventually be amortized too. Billing customer has a cost. Cost of servicing each call is essentially zero. This, along with higher bandwidth, is an enabling technology for freedom and democracy.