Current Cover (3068 bytes)
Current Cover
HOME | MORE COLUMNS | Win Mobile | Palm OS | Features | Pen Computers | Tablet PCs | Rugged PCs | Subscribe | Case Studies |
Commentary by Pen Computing Magazine's editor-in-chief


By Conrad H. Blickenstorfer
Issue 48

Palm retires Graffiti
In September of 1994, in the early days of Pen Computing Magazine, I received a call from a very excited Andy Seybold. He essentially told me to stop the presses for the November '94 issue of Pen Computing Magazine because he had just seen something that had changed the way he looked at PDAs. Seybold, then and now one of the leading industry experts on handheld and wireless technology, had been on record as declaring pen and handwriting recognition as "cumbersome and not acceptable." What had changed Seybold's mind to the extent where he felt new life had been breathed into PDAs and that it made him want "to take his Newton out of the drawer and begin using it."

It was Graffiti. "Graffiti is an entirely new way of viewing the relationship of pen to computer," said Seybold. Based on his urgent call we stopped the presses, yanked a page in the front of the magazine and inserted Seybold's report on Graffiti, which he titled "Major Breakthrough for PDAs." In it, Andy explained the brilliant thought process and rationale behind Graffiti. He talked about how Jeff Hawkins, recipient of many patents regarding to pen computing and related technologies, had realized that people tolerate their own mistakes but get upset when computers make mistakes. He had also found that people liked the immediate feedback they get from a keyboard: type a character and it appears on the screen. So what Hawkins did was create a very simple alphabet that eliminated the ambiguity of the letters in the English alphabet. If people learned that simplified alphabet, the computer would recognize each letter. And each letter appeared on the screen instantaneously, providing instant feedback. Hawkins had thus solved the two major problems facing handwriting recognition: He shifted the emphasis from the computer to the user, and he provided the instant feedback people wanted.

Palm Computing made Graffiti available for the Newton, a number of other contemporary PDA and pen computing operating environments (GEOS, Magic Cap, PenRight!) and announced a version for Windows. When little Palm branched out into hardware with the Palm Pilot, Graffiti was the primary data entry method. As Andy Seybold had predicted, the Graffiti alphabet, which resembled the standard alphabet, was easy to learn and soon millions of people used it. I should mention here that no matter how brilliant Graffiti was, it alone did not make the Palm Pilot succeed. Around the same time Palm introduced the first Pilot, Hewlett Packard introduced the OmniGo, a keyboard-based PDA that also used Graffiti. According to the wisdom of the time the OmniGo with its keyboard, Graffiti, and expansion slot should have been a success whereas the keyboardless, non-expandable Pilot should have been a failure. The reason the Pilot succeeded was because it was small and handy and brilliantly conceived, not because of Graffiti.

Nonetheless, Graffiti quickly became a standard. Graffiti rejuvenated the PDA market. Graffiti, in some ways, established an entire industry. Virtually every other handwriting recognition company began offering its own Graffiti-style product. Most used slightly different alphabets that were closer to the standard alphabet (which made them less accurate than Graffiti), but almost all also had a "Graffiti mode."

Fast-forward a few years. Xerox Corporation launched a lawsuit against Palm, claiming Palm had committed patent infringement with Graffiti. High-powered law firms became involved to pursue the case, which centered on "unistroke" characters. A Xerox patent which was finally granted in April of 2002 claimed ownership of "a unistrokes symbology in which strokes of like profile (i.e., strokes that are distinguished from each other by their rotational orientation) are rotationally offset from each other by at least 90 degrees is provided. This provides a sufficient tolerance for disambiguating these strokes when they are written into hand-held pen computers and the like by users having wildly divergent hand writing styles." The patent included the Xerox alphabet, which, as you can see, is totally non-intuitive and does not resemble Graffiti nor the standard alphabet at all. The likelihood that many people would be able to memorize such a weird set of strokes is very small. Had Graffiti looked like that, it would have died a quick death. To illustrate the point, look at Gesture Mosaic, another attempt to introduce a non-intuitive alphabet. It was an interesting idea but it never really caught on.

Yet, despite the obvious gigantic difference between Hawkins' brilliant deductions and the resulting Graffiti alphabet, and the Xerox symbols, the legal battle raged on and Xerox eventually won, making very-hard-to-believe claims like "we have several years of damages with this handwriting recognition system to address." As a result, on January 13, 2003, Palm announced it was going to drop Graffiti and replace it with a new system based on CIC's Jot, a character recognizer that was originally developed specifically to compete with Graffiti. So is Palm going to be safe? According to a Reuter's report, not necessarily: "However, when asked if Xerox had any plans to target CIC for its handwriting recognition technology, he said the company is focusing for now on the Palm case." And how about Microsoft whose Pocket PC "Block Recognizer" is, let's face it, also a Graffiti clone?

The whole situation is absurd. If Xerox had gotten away with a similarly obnoxious, anti-progress legal strategy against Apple Computer there would likely be no Macintosh today, no Microsoft Windows, and no other graphical user interfaces because Xerox would have tried to keep it all to themselves based on some dubious arcane patent disputes and legal technicalities. So what does it boil down to? Our esteemed legal system handed a company with one of the worst-ever track records of converting its many ideas ideas to products a dubious, and in our opinion absurd, legal victory against a company that does convert their ideas into actual products. As a result, progress suffers, millions of people will be robbed of Graffiti thanks to Xerox's ill-conceived actions, and we're one step closer to burying the great American ingenuity under a bunch of ridiculous lawsuits.

But on to less depressing matters. The Tablet PC got off to a respectable start with many products showcased at last November's better-than-expected Comdex and then at a stellar Consumer Electronics Show in January. It's still going to be an uphill battle to convince corporations and the general public of the benefits of ink and a pen interface, but at least the machines are now on the shelves of computer stores and in the catalogs of most of the major computer mailorder companies. I am a bit concerned that prospective customers, not seeing a Tablet PC "killer app" nor an immediate need for ink, will take one look at the rather high price and relatively modest performance specs of the first generation of Tablet PCs and then order a standard notebook with a Pentium 4 for hundreds less. Our technology editor, Geoff Walker, disagrees with me on that. He believes that consumers will compare Tablet PC prices and specifications with those of premium (and similarly priced) ultralight notebooks and then consider the pen machines because they get something extra. I hope he is right.

On the PDA front, our editorial offices are labs are awash in a veritable flood of exciting new products from HP, Palm, Sony, Dell and many others. And someone seems to announce a new smartphone or convergent product almost every week. The tech market may still be in the dumps, but you sure couldn't tell that by the bustling activity and innovation out there. - -

Conrad H. Blickenstorfer, a former corporate CIO, is editor-in-chief of Pen Computing Magazine and Publication Director of Digital Camera Magazine. He can be reached via e-mail at cb@pencomputing.com.