Where Are All The Web Pads?

More than two years ago, various companies began demonstrating prototypes of Web pads—tablet computers dedicated to Web browsing. With a wireless link to a base station, which in turn connects to a phone line or home network, a Web pad provides Internet access with none of the complexity of a personal computer. Plus, you can use it anywhere in the house, putting the Web where the people are, instead of where the computer is.

The Web-pad concept resonated with many people. It could make Web access as convenient as reading a newspaper, consulting a phone directory, or watching TV. Yet even though prototypes have been around for years, Web pads remain essentially unavailable. Three major issues have delayed them: cost, capability, and market requirements.

The most expensive component in a Web pad is the screen. Unfortunately, liquid- crystal displays (LCDs) are in very short supply. Given their limited manufacturing capacities, LCD makers prefer to sell larger, more expensive displays for notebook computers and desktop monitors. With demand for these products outstripping supply, there's little incentive to price smaller displays aggressively.

The supply constraint is due, in part, to the Asian financial crisis, which prompted most manufacturers to delay plans for new manufacturing plants. Many new plants are now finally under construction, but it takes years—and billions of dollars—for a new plant to begin production. The supply/ demand imbalance isn't likely to get dramatically better this year, but in 2001 the situation could turn around.

Along with high costs, delivering a high-quality Web experience has proven surprisingly difficult. At first, it seemed Web standards such as HTML provided a new definition of compatibility that eliminated the need for Windows software or x86 processors in Web appliances. But the ever-changing browser plug-ins and evolving Web standards have made it tough to create browsers for Web pads that deliver a PC-quality browsing experience. The list of widely used plug-ins is surprisingly long, from streaming audio and video (RealAudio and Windows Media Player) to layout and graphics (Flash and PDF). If the browser isn't Microsoft Internet Explorer or Netscape Navigator, the operating system isn't Windows (or perhaps Macintosh), and the processor isn't x86 (or perhaps PowerPC), you need special versions of each plug-in.

Suppliers of these plug-ins have their hands full keeping up with the rapid pace of PC-centric Internet software, and because the installed base of Web appliances is tiny, there's little immediate incentive to invest in supporting them. And because the various appliance vendors are pursuing a range of technology strategies, there's no single Web-pad standard for plug-in suppliers to target. Furthermore, most plug-ins are written assuming lots of memory and a disk, which makes it hard to port them to the limited-memory world of the Web pad.

If the target market is those without computers who are looking to Web pads as an easy on-ramp to the Internet, it may be acceptable to forgo support for many plug-ins. This is the case with most Internet appliances so far, including Netpliance's i-opener and Microsoft's WebTV.

The problem is Web pads aren't cheap, and the most likely early market is not Internet neophytes but active Web users who've had PCs for years and like the idea of having an additional Web-access device at home. These users will demand a Web pad that delivers the same quality Web experience as their PCs.

Some prospective Web-pad marketers are concerned that even full-featured Web access isn't enough to get many people to buy a Web pad. These companies are working with partners to develop additional services that make Web pads more useful outside the home office. They're also seeking partners to subsidize the cost of the device by offering preset bookmarks and other forms of advertising.

Many early Web-pad developers have switched from RISC to x86 processors in order to gain easier access to plug-ins, and Transmeta has pitched its low-power x86 processor as an ideal solution for this market. Simply being x86-compatible doesn't solve the problem, though; the full set of plug-ins is available only for the Windows environment. Microsoft hopes to use this advantage to make the new Windows CE 3.0—for which a full IE 4 browser is available—the operating system of choice for Web pads.

All these factors combined mean you won't see a flood of Web pads this year. Some will ship this year, and if all goes well, many more will ship next year; then the flood could begin in 2002. Eventually, Web pads will be a major part of the computing landscape, but it's going to take a frustratingly long time.