February 5, 2001

Last year, the Palm-thingy industry sold 3.5 million personal information managers, handheld devices, digital organizers, or whatever they are calling them nowadays. I am not impressed. The adoption pattern is slow and the demographic base is narrow.

Handheld computer sales probably will peak this year or next. They represent a compromise technology that will be superceded as wireless Internet access improves.

The most telling indicator that handheld computers have a limited market is the fact that relatively few women own them. Perhaps it is impolite to speak in terms of gender. So let "women" be a generic term for anyone with enough common sense to evaluate technology for its practical utility, rather than buy something out of deep-seated insecurity about the size of his, er, peacock feathers.

Here is a question to test your knowledge of personal information managers:

What factor determines the number of items of information that can be stored on a handheld electronic organizer?

If you answered, "the number of megabytes of RAM," you are technically literate but otherwise a fool. If you believe that you can store thousands of addresses on a pocket gadget, then you had better hope that you do not really need any of those addresses.

If you have used one of these devices in the real world, then you would say that the factor that determines storage capacity is the owner's ability to anticipate when batteries will need changing. If you do not change the batteries in time, the number of items that they will store is zero.

Because batteries do tend to run out, the real answer is, "You cannot store information on a handheld computer, regardless of the amount of RAM." Experienced users will "synch up" their pocket organizers with a personal computer once a day.

Today's handheld computers represent a compromise between the philosophy of personal computing and server-based computing. They provide just enough local processing power to be more than thin clients, but not enough reliability to be used independently of personal computers.

The fact is that pocket organizers occupy a sour spot in between two viable architectures. At one extreme are devices that derive their intelligence from local storage, and which come with usable keyboards and readable screens. These are laptop computers.

At the other extreme are devices that derive their intelligence from a network. These include cell phones and portable email devices. Today, many of them lack adequate screens or input devices. As long as wireless Internet access suffers from poor bandwidth, there is little incentive to improve the screens on cell phones or to switch from the dialpad interface. However the ultimate form of these products probably will change as bandwidth increases in the cell network.

The real value in portable devices comes from communication, not from computing and storage. The design objective should be to maximize the communications capability of the device. The result might be something like a cell phone combined with the keyboard and screen of a Blackberry. Perhaps the form of the cell phone would be a headset, leaving your hands free to operate the keyboard.

If I had such an integrated communication device, then I would keep my address book and schedule on a server on the Internet. I would use the portable device to update the information directly on the server. I would not run applications locally or rely on local storage.

I refuse to be a slave to a device that requires me to "synch up" once a day. It offends me to be doing sneakernet maintenance in the Internet age.

My guess is that the handheld devices that reach the mass market, particularly "women," will function primarily as communication tools, not as computing tools. The ability to run applications and store information locally will not be a cost-effective use of battery power.

Laptops will stay. People want the ability to take work and/or entertainment with them on a plane, and that means having a lot of local storage and processing power. But the Palm-thingys, with neither reliable local storage nor outstanding communication functions, will go.