Printer Friendly

Special report: the pen-based revolution.

We've become increasingly skeptical about self-proclaimed technology revolutions: Like Third World political upheavals, most seem to fizzle out after a few weeks of mass meetings and stirring manifestos; then the anciene regime shuffles a few cabinet titles and it's back to business as usual.

But this time we're convinced something like a true revolution is in the works. Pen-based computing--launched officially on January 22 with the unveiling of GO Corp.'s PenPoint operating system--is more than a gimmick. We may not see large-scale commercial implementations until the middle of the decade, but we suspect PenPoint will be accepted almost immediately by developers as a strategic platform (a status that NeXT still hasn't achieved) and that pen-based concepts soon will begin to influence software design throughout the entire PC industry.

PenPoint's impact is extraordinarily hard to predict; like the early Macintosh, PenPoint challenges basic assumptions about personal computing (perhaps that's the best evidence of how innovative GO's approach is). Even Vern Raburn, one of the first software entrepreneurs to jump on the Penpoint bandwagon, admits he "can't fundamentally articulate what pen-based computing is." But we'll still venture a few observations about PenPoint and possible markets for pen-based applications:


First, let's be clear about what pen-based computing is not: a new handwriting recognition technology. The popular perception of Penpoint and Pen Windows (Microsoft's get-it-out-the-door-fast response) is that they've been created chiefly for deciphering handwritten scrawls.

True, both environments do perform some gee-whiz tricks with neatly handwritten text and gestures. Handwriting recognition is a feature that's bound to be useful for such tasks as proofreading, filling out simple forms, and updating a spreadsheet. But nothing we've seen suggests that pen technology is ready to cope with serious note-taking, cursive writing, or even sloppy penmanship. It's equally unrealistic, we believe, to think of PenPoint or Pen Windows as solutions to the ultimately trivial problem of keyboard phobia "like designing a car for people who can't drive," says John Dvorak).

In fact, as the name suggests, pen-based computing is about pens. And it's becoming clear (because of the inevitable comparisons between PenPoint and Pen Windows) that there will be two fundamental ways to think about the pen--as an input device and as a metaphor.

As an input device, the pen (actually, an electronic stylus that interacts directly with the computer's screen) simply gives users a new tool for drawing, marking, and entering small amounts of text. The pen is a natural tool for these tasks, and it often works significantly better than the keyboard or mouse. The arrival of a new input device, of course, doesn't automatically shake up the software world; in fact, the basic assumption behind Pen Windows is that pen input can be treated as a painless extension of the Windows environment and Windows software.

GO's approach is considerably more ambitious. The pen is the point," says a company slogan--that is, the PenPoint pen is not just an input device but the primary metaphor that shapes our experience with the computer. Just as the macintosh and windows rely on the metaphor of the desktop to define the computing experience, GO now offers a "pencentric" experience that is remarkably friendly, deep, and powerful.

As a metaphor, the pen personalizes a rather abstract notion--the direct manipulation of graphical objects. Much of PenPoint consists of features that have become standard components of modern operating systems and environments: hypertext buttons, point-and-click selection, "smart" objects that guess the user's intention, compound documents, and gesture-based commands. But the PenPoint developers have figured out how to relate all of these components back to a single focal point, the pen, leaving the user with the helpful illusion that the pen controls the computer absolutely.

The Pen as Metaphor approach has important consequences for PenPoint developers, because--like the macintosh desktop--the metaphor strongly influences the look and feel of third-party applications. Rather than display applications and files as objects on a desktop, GO shows the user a series of "notebook" pages that contain documents, not individual applications. In theory, individual documents don't have to conform to consistent interface rules (eventually, PenPoint may even open a DOS file as a page). But GO puts a lot of emphasis on PenPoint's ability to create compound documents, so the idiosyncratic behavior of an underlying application is likely to be especially noticeable. Will GO succeed in imposing "fascist" control on free-spirited programmers who don't grasp the value of consistency? Stay tuned for the next thrilling episode.


At Fall Comdex, Microsoft's Bill Gates demonstrated how a notebook-sized, pen-based PC could let a traveling bakery salesman enter orders, retrieve customer information, and create instant presentations. Much of the promise of pen-based computing is its potential to bring new "mobile" users like these into the PC fold--a market expansion that could add literally tens of millions of new users to the current installed base.

But we're not convinced that Pen Windows is the key to unlocking these markets. The chief obstacle is hardware: Windows has been designed around desktop machines with hard disks, fairly large color VGA screens, and other components that bump up both the weight and price of the minimum windows hardware configuration. (According to Jeff Raikes, one incidental requirement of Pen Windows will be a hefty two megabytes of ROM.) And we don't foresee a leaner, entry-level version of Windows that would be appropriate for non-desktop users; if anything, Microsoft has been pushing Windows up the hardware spectrum faster than the PC marketplace at large has been moving.

Instead, we expect Pen Windows to flourish primarily among white-collar users of traditional productivity applications--spreadsheets, word processors, databases, presentation packages. Microsoft's argument that pen-based computing should be based on a "standard" (i.e., Microsoft) operating system makes sense in this context, and we expect that most mainstream software companies will adopt Pen Windows as a way to add incremental features to current product lines.

Which leaves the mobile market up for grabs. Assuming Go lines up the right hardware partners, PenPoint has a fair shot at becoming the dominant operating system for blue-collar and white-collar workers who handle large amounts of paperwork, data collection, and analysis away from their desks. GO's design goal was an operating system architecture optimized for portable, notebook-style machines rather than desktop models; thus, PenPoint presumes that users will have a monochrome display (possibly as small as a Daytimer), no disk media, and only sporadic connections with LANs or other PCs.

Not surprisingly, PenPoint has already begun to attract a following of interesting software entrepreneurs. Slate Corp., founded and staffed by an impressive group of PC industry pioneers, is by far the most visionary and ambitious, but we expect to see lots of small players with pen-based product ideas just crazy enough to succeed. Of course, most PenPoint startups will probably flame out, having failed to find customers who appreciate whacky ideas. But a few--like Aldus, Adobe, and even Microsoft in the Mac market--could turn into $100-million ventures.

Which PenPoint entrepreneurs have the best prospects for success? We see at least three markets where pen applications should do well:

* Laptop-like products: GO will get rolling with a small installed hardware owners, mostly professionals who see PenPoint machines as nifty executive toys. PenSoft's still-nameless personal information manager is likely to be the hot product for these users, along with any software that automates familiar personal tasks--expense account management, E-mail and fax communication, memo writing, etc. Many of these products will resemble the kind of applications that are popular on laptops, and will be sold through conventional retail channels. Judging from the less than spectacular sales of similar products for laptops and hand-held machines, this is a market that will support a few small companies. But we probably won't see any blockbuster hits here.

* Fleet applications: Where PenPoint applications will do best, we believe, is among so-called fleet" companies, who typically hire hundreds (or thousands) of sales reps, van drivers, meter maids, service technicians, claims adjusters, nurses, and the like. Here, the software buyer isn't a personal user, but rather a corporate developer of customized single-purpose systems. In the past, fleet customers have had to rely on proprietary automation solutions (like the bar code readers that Federal Express uses, or the hand-held machines that Hertz gives to employees who check in rental cars). PenPoint offers an open systems alternative--multi-vendor hardware and the promise of a rich choice of generic development tools. Fleet applications are Slate's primary target for its PenApps forms processor (which Raburn hopes will become the dbase of the pen world"), but there's room in the fleet market for a good many other developers of pen-based languages, object libraries (such as on-screen gauges, sliders, stop watches, and other gizmos that are often part of fleet hardware systems), and customizeable vertical market packages.

* Drawing applications: Go has been curiously quiet about the CAD market and other drawing-related applications. Nevertheless, this looks like an attractive niche. Engineers, architects, illustrators, and other design professionals are naturally pencentric; keyboards and mice have never delivered the immediacy and control of pen on paper. Our guess is that PenPoint drawing applications will gradually infiltrate the CAD and drawing market, and--if the current leaders respond too slowly--PenPoint could be the catalyst for a substantial rearrangement of market shares in these areas.

Ironically, PenPoint still has to overcome one obstacle in most of these markets--the market perception that a pen-centric architecture doesn't accommodate other sources of data input. Like all good religious zealots, the PenPoint enthusiasts tend to overstate their case by demonstrating hardware and software that operates only with a pen. That's like showing a Mac with just a mouse attached: It works, but not well enough for a full range of everyday tasks.

In reality, the pen isn't the only input device that's appropriate for mobile applications. Keyboards and numeric keypads are still far faster and more accurate for entering text and performing calculations. (Weight isn't a factor, either, considering that Poqet has produced an almost-acceptable keyboard machine that literally fits in a jacket pocket.) GO has an even bigger blind spot about system-level support for bar code readers: Bar code input is now solidly entrenched as a standard technology for data collection, especially among large fleet customers that GO badly needs to attract. By ignoring the importance of bar code, GO effectively disqualifies itself (and its developers) as serious players in major data collection markets.

The good news is that PenPoint's architecture actually doesn't exclude other forms of data input. PenPoint already permits keyboard entry (though the hardware isn't ready), and Robert Carr, Go's vice president of engineering, says that adding a bar code interface wouldn't take much effort. The tough problem will be to integrate other input devices into GO's own pen-centric vision; that's not going to be easy.


GO finds itself in a curious, perhaps awkward position: PenPoint represents the first launch of a major new computing platform that's occurred without a tangible hardware component. Initially, GO planned to make machines as well as system software, but last summer the company bowed out of the manufacturing business (except for some hand-made development machines) and switched to a pure licensing strategy. IBM has already signed on as a PenPoint licensee, along with NCR and Grid, so there shouldn't be any shortage of pen-based hardware. But what will PenPoint hardware actually look like? The fact is, nobody knows.

For traditional desktop applications, hardware has never been much of a mystery. Developers have known roughly the constraints imposed by screen size, memory, disk storage, keyboard (and mouse) availability, and other physical parameters. Even though the gap between high- and low-end machines continues to widen for both the PC and the Mac, basic hardware form factors have remained fairly stable.

PenPoint, however, is a highly scalable operating system, one that will be easily adaptable to machines as small as a pocket-sized wizard or as large as a blackboard. Particularly in the fleet market, PenPoint machines are likely to be optimized for light-weight clipboard applications; in other markets, PenPoint hardware could show up with a hodgepodge of keyboards, color screens, storage and connectivity peripherals, and other special purpose attachments.

Ultimately, we expect PenPoint will spawn the most diverse array of hardware form factors the software industry has ever encountered. Some development tools may be portable across this range of machines, but we suspect the pen-based applications market will be shaped by specialized hardware niches far more than the desktop market has ever been.

Another complication is the uncertainty about IBM's strategy as a PenPoint licensee. It's conceivable that IBM recognizes that PenPoint could be the key to control of a new mass market, a standard that will help IBM squeeze out proprietary hardware companies and (unlike the PC) that doesn't cannibalize sales of high-end machines. If that's the IBM strategy, it's possible we'll see an engineering and sales blitzkrieg that blows the pen-based software market wide open. But it's also possible that IBM will twiddle its corporate thumbs and wait for other hardware companies to soften up the beaches. what's the most likely scenario? We haven't a clue.

And then there's the question of Apple's response to pen-based computing. Apple has dabbled in handwriting recognition technologies (especially for Japanese input), but otherwise doesn't seem likely to be a major player for true pen-centric applications. one reason, perhaps, is that Apple hasn't had much luck designing portable machines; currently, the notebook and hand-held market is almost entirely Intel-based (as is PenPoint itself). Our guess is that Apple will take the same approach as Microsoft has with Pen Windows--that is, it will extend the basic Mac operating system with a pen-input layer, but otherwise leave the desktop metaphor intact.


Pen-based computing may become one of the most leisurely revolutions the PC industry has ever experienced. We won't see any PenPoint hardware or applications until the end of 1991, and industrial-strength development tools probably won't start showing up until mid-1992. Pen Windows is off to an even slower start; our guess is that Microsoft won't be able to deliver a true commercial product much earlier than mid-1993.

Moreover, the whole technology is still very rough around the edges, about on a par with the first 128K Mac. It's safe to predict, we think, that pen-based computing will have to endure as much as five years of choppy evolution and corporate shakeouts before it reaches its full potential as a serious platform. And that kind of slow progress is bound to irritate a lot of customers and investors, who like quick results. (We're waiting to see how quickly the trade press begins speculating about why PenPoint and Pen Windows "failed.")

Nevertheless, pen-based computing is a force that has to be taken seriously today. Like the graphical interface, the pen is bound to reshape the software industry in dozens of unexpected and profound ways. PenPoint just made current assumptions about operating systems and market opportunities obsolete; the only important question now is which companies will move fast enough to become our next generation of market leaders.
COPYRIGHT 1991 Soft-letter
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:pen-based computing
Author:Tarter, Jeffrey
Date:Feb 4, 1991
Previous Article:Postscript.
Next Article:Benchmark survey: sales compensation.

Related Articles
Touchless typing: electronic pens and voice-activated software are coming into their own as ways to enter data into a computer.
Microsoft. (Computers).
Logitech io pen: Logitech. (new products).
Maxell DP-101B.
Wacom to Release Pen Tablet to Allow for Different Feels for Writing.
Dai Nippon Printing, Others Develop Traceability System for People Injured in Large-scale Disasters.
Wireless Generation: mClass: see.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters