Handheld computers: how it looked in September 2000

This piece first appeared in The Independent around September 2000. Given all the talk about some handheld(ish) computer released by some company or other, I thought it might be interesting to look back on…

A couple of notable phrases: “Microsoft’s failure in this market is unusual..” and at the end that “In the long term though functionality is sure to win out over form”. Debate among yourselves whether this was just history talking…

Handheld computers


Technology Editor

Handheld computers cannot do what most people want them to. This may seem surprising, given that millions of models using operating systems from Palm, Psion and Microsoft have been sold since 1984, when the British company Psion introduced its first handheld model.

But all are severely limited compared to the expectations placed upon them, which can be traced back to two sources: the 1960s TV series Star Trek, and the hit BBC radio series first broadcast in the 1970s, The Hitchhiker’s Guide To The Galaxy written by Douglas Adams. Only at the turn of the [21st] century does it look like people will soon be able to buy products with the facilities that people have been hankering after for decades.

The sight in the 1960s of William Shatner as Captain Kirk landing on alien planets and flipping out a palm-sized machine which could act as a radio, intelligent locator and general categoriser of knowledge had a subtle effect on the baby boomers’ belief about what computers of the future could and should do. It was voice-activated, and context- and location-sensitive. Similarly, in the radio series, the Hitchhiker’s Guide to the Galaxy was actually the name of a computerised guidebook. It contained as much information as the galactic hitchhiker could need. While its indexing method was hopeless by any standards – the traveller had to look up a number in the index and enter that in order to get the corresponding entry (“so bad it could have been designed by Microsoft,” Adams later quipped) – it did create the belief that someday one could build a handheld machine able to hold all the knowledge not just in the world, but in the galaxy. And if aliens had them, why shouldn’t we?

The reality of the first retail products was rather different. The Psion 1, the brainchild of David Potter, was launched in 1984. It had a mighty 10K of non-volatile memory, an alphabetic keypad and a one-line 16-character LCD screen. Entering data was tedious. It would not have passed muster with Captain Kirk. However its descendants are now widely used by people in jobs requiring simple data collection, notably including traffic wardens.

In August 1993 Apple Computer launched its $700 Newton, which seemed at the time to promise at least some Star Trek functionality. It had handwriting recognition software able to “learn” your specific cursive style; there were promises of wireless communications and word processing.

It turned out to be an example of the computer industry’s occasional hubris. The software did learn your writing style, but often failed to interpret the letters correctly. The Newton was a flop (officially abandoned in 1998, but dead some years earlier) which poisoned the well for entrepreneurs in the US handheld market for some years. Bill Gates of Microsoft reckoned it put the market for such products back by two years. (Probaly an underestimate.) Palm Computing, founded by Jeff Hawkins and Donna Dubinsky in 1992, only managed to survive by selling itself to the modem maker US Robotics early in 1996.

However, in the UK and Europe Psion was thriving, and had developed its Psion Organiser 3 series, which had a miniature keyboard and inbuilt software including limited word processing, a calendar, contacts book and spreadsheet. It looked like a miniaturised version of a laptop computer, and proved very successful in its local market.

But on the west coast of the US, Hawkins and Dubinsky were developing a palm-sized machine which would have some, at least, of the ease of use both of Captain Kirk’s communicator and The Hitchhiker’s Guide. Hawkins envisaged a machine – which later became the Palm series – that would not stand alone, but would synchronise and back up its files with a standard PC. Thus it would not have to do everything; only have enough functionality to be useful while out of touch with the PC.

For data entry, he developed a shorthand cursive system called “Graffiti” which all Palm users have to learn. He tested the ergonomics of the product by carving a block of wood into a size and shape that he could carry comfortably around in his pocket. Function and form thus developed in parallel.

The Palm operating system was hugely popular, even though the basic machine only offered a calendar, address book, task (“To-do”) and notes list, plus a calculator and search system. Its success stemmed from its ability to coordinate with a PC; the openness of the operating system; and the coincidental rise of the Internet. The first point meant users could access their databases more easily than with tiny keyboards; the second that software developers could write programs to enhance the machine; and the third, that those programs could be widely and quickly distributed. Psion, with its EPOC operating system, had attracted some software developers but was held back by its European location (where Internet development lagged by a couple of years compared to the US) and lack of connectivity to PCs.

Launched early in 1996, the first Palm computer sold 1 million units in 18 months. In 1998 Hawkins and Dubinsky left with Ed Colligan, marketing head of Palm: they were dismayed by the slow working of the monolithic 3Com, which had bought US Robotics. They set up their own company Handspring, and licensed the Palm OS, which then had 80 per cent of the world market, served by 100,000 developers, while Psion and Microsoft scrabbled over the remainder.

Microsoft’s failure in this market is unusual, but seems to stem from its WindowsCE operating system (renamed and rebuilt as PocketPC in spring 2000) being too complex for the limited power of the machines. WindowsCE is used in petrol pumps and set-top boxes for decoding digital TV signals.

The future promises rapid change. Until 2000, handheld computers sat apart from mobile phones: an address list on one could not be transferred to another. As usability expert Jakob Nielsen noted, this is absurdly inconvenient. Mobile phones are no good for noting data (such as phone numbers) while you are in a call; but handhelds have been little use for making phone calls.

But Handspring especially has been forcing the pace, as its Visor machines, which use the Palm OS, include a slot called the “Springboard” where the user can plug in items such as a camera, memory module and – from autumn 2000 – a GSM modem.

That abruptly made the Handspring into the potential killer combination of handheld address list and mobile phone. Palm rapidly announced that by the end of 2000, all of its products would have wireless capability. Separately, IBM demonstrated a version of a Palm machine with an add-on board which gave it voice recognition capability, using the ViaVoice technology. Suddenly, the humble handheld was beginning to look like the machine which would be able to do everything.

But mobile phone makers and Psion are not finished. The so-called “third generation” of mobile phones, which will have high-speed data connections, were being designed in 2000, and the Symbian consortium (which uses the Epoc OS) won a contract to provide the OS for a number of phone companies.

What was still unclear at the end of 2000 was whether handheld computers would swallow mobile phones, or vice-versa. The handhelds had the functionality; the mobile phones had the usability. However the mobiles rapidly lost that edge as new WAP (Wireless Applications Protocol) phones attempting to squeeze Internet interactivity into a few lines of a monochrome LCD screen. In some respects, it was a step back to 1984. But the market’s explosive growth may mean that there is room for everyone to survive. In the long term though functionality is sure to win out over form.



  1. Windows CE wasn’t renamed and rebuilt as PocketPC: PocketPC was a variant.

    WinCE could be used in a whole range of different configurations, ranging from a headless and extremely light OS burnt into an appliance all the way up to thin clients, tablets and ‘Handheld PCs’ (as Microsoft called them – we’d call them netbooks now).

    One particular configuration was of a PDA-style device. WinCE in this configuration, along with some additional applications from Microsoft, was called PocketPC. Getting a license for WinCE didn’t get you the extra goodies PocketPC had, and licensing PocketPC didn’t get you support for all the other stuff that WinCE had. A superset of a subset, if you will.

    (Microsoft later augmented all of this by offering Windows XP Embedded, which was sort of a [much] bigger brother to WinCE, but had the advantage that it could run the same applications – dependencies allowing – as a normal copy of XP, though only on x86, and with much higher storage and RAM requirements than WinCE).

  2. I still occasionally miss the Psion 5MX – it was well ahead of its time, the keyboard / clam shell design was great. It had banking via a Lloyds Electronic Chequbook too – anyone remember that?

    http://www.theregister.co.uk/2007/06/26/psion_special/ is worth a read.

  3. It all went wrong for Palm when 3Com bought up (sorry, merged with) US Robotics in the late nineties. 3Com was the totally clueless about what it had bought; it was really after USR’s remote access products.

    As Max Boyce said, I know, ‘cos I was there…

Comments are closed.