Summary: Sometimes people aren't content with making the same mistake twice.

Stephen H. Wildstrom's "Technology & You" editorial in the November 9th edition of Business Week sounded the death knell for the home PC. ZDNet AnchorDesk editor Jesse Berst's November 30th "Berst Alert" proclaims, "KISS YOUR SWEET PC GOOD-BYE: THE END OF THE PC ERA." It's too convoluted. It's too complex. It does too many things. People don't want the power to surf the Web, balance their checkbooks, play games, send and receive e-mail, write standard mail and faxes, do homework or business reports, create tee-shirts and business cards, track income and expenses, and another gazillion little tasks, wrapped up in a single, relatively inexpensive package. Wildstrom agrees with IBM's Paul Horn, a senior vice-president for research, whom he quotes as saying the standard desktop personal computer is "too complicated" for most users who are more likely to want individualized contraptions for the different jobs that they're presently accomplishing on their PCs. This means we'll soon see a surge in single-use devices, "appliances," to use Horn's term, as well as a sudden blossoming of that long-awaited PC killer, the Networked Computer. So, chuck out your sparkly new Pentium 400s, Cyrix or AMD-based machines and get ready for the new wave of individual, single-use appliances which are destined to take their place. The revolution is right around the corner and nothing can stop it.

Nope. Nothing.

Nothing at all, unless, of course, you consider something the pundits and Big Blue seem to have in remarkably short supply: a few pounds of common sense.

Taking It by the Numbers

There are at least five compelling reasons to look at the NC/Computing Appliance vision as simply the latest installment in IBM's popular technology series, The Next Big Idea That Failed:

1. User Convenience
2. User Control
3. User Inertia
4. Network Infrastructure
5. User Expense

Any one of these issues alone would pose a considerable challenge to this purported new wave of computer technology. The five combined will be difficult, if not impossible, to overcome.

User Convenience

To begin, let's explore this idea of the "computer as appliance" for a few minutes to see where it leads. I suspect that many of us associate the word 'appliance' with the word 'kitchen' since we seem to load our kitchens up with the damn things. And, unless I'm mistaken, most of our kitchens are areas devoted to the preparation of food, hence the gadgets which live there are most commonly designed for food preparation and service. Our silverware, food processors, blenders, mixers, microwave ovens, bowls, plates, salad shooters, refrigerators and ranges dominate this space. We could further look at these various implements, dividing them into those which store ingredients and prepared dishes (refrigerators, baggies and cupboards) from those which prepare them for cooking (food processors, paring knives, graters, etc.) to those which cook or re-heat them (ranges, waffle irons, microwaves, convection ovens and grills). Of these, the range-top oven is most like today's fat-client PC. You can heat soup or fry bacon on it, cook a casserole in it or use it to grill up steaks and burgers. You can also use it to thaw frozen foods or home process the tomatoes in your garden. While it's probably one of the most expensive appliances in the kitchen both in terms of initial cost and required storage space, it does a number of jobs well. There are other appliances on the market that can replace it, but they aren't very convenient. You can buy an electric grill for your steaks and burgers, but you have to find space on the countertop to use it and cupboard space to store it in when you're done. Convection ovens cook meat faster, but they generate a ton of heat. Microwave ovens are great for heating up soup or thawing frozen foods, but it's damn difficult to get crisp bacon or nicely-browned cookies out of one. In short, the standard range is a remarkably flexible device when it comes to cooking food. All a cook needs to set up shop is a few sauce pans, a skillet or two and a griddle.

Consider, now, the Personal Computer. It deals with information. We use it to create, modify, transfer and store information in much the same way that we use our ranges to fry, boil, bake or cook food. That is to say, we use our applications (browsers, e-mail clients, word processors, spreadsheets and databases) with our computers the same way a cook uses her pots and pans; toss the data in, stir gently, out comes a letter, inventory report or web page. Most importantly, it deals with any information. Most modern machines can crunch numbers, reproduce sounds, create text or edit images with ease. All the user needs to set up shop is a word processor, spreadsheet, graphics program and database. The NC/appliance combo that Wildstrom and Horn are talking about will only be able to accomplish this same level of flexibility by having a number of appliances which will take the place of the one we're currently using. One appliance to surf the Web. One to write a report and print it. (Gee, could this be a dedicated word processor like those ever-popular, fast-selling models that Brother and Smith-Corona are devouring the market with?) Another to do e-mail, and so on. If you want to surf the Web and pull an image from it into the spreadsheet you're working on, you'll have to have your appliances networked together. You won't be able to deal with that information with just one piece of hardware. In other words, if you want soup with your burger, you'll have to get out a hot pot and an electric grill instead of just pulling out a skillet and a saucepan. Does that sound convenient to you?

While few of us don't have occasional gripes with our machines, we have come to appreciate the fact that we can share a wide variety of information on them with very little effort. We can import and export files or cut and paste data from wildly different sources with a few mouse clicks or keystrokes. If we find an image on the Web that would look good in a homework project for the kids, an e-mail for Aunt Sally, wallpaper for our desktop or the cover for a business project, we can easily import it into an e-mail client, word processor or spreadsheet and, bingo! we're in business. If we need to print out last year's income report for the IRS or get a copy of our latest checking account reconciliation to sort out a mess with the bank, it's right at our fingertips. If that income report needs a letter to explain it, we can create the letter while the report is printing without moving from the chair. That's convenience and I doubt that many users will gladly relinquish it so they can fill their offices or living rooms with 6 or 8 gadgets to take the place of one PC.

User Control

We can find another blind spot in the computer appliance/Network Computer vision when we consider the lack of personal control for data that's part and parcel of a network system. If NCs have only enough RAM and storage to log onto the network, there won't be any place for users to store data in their homes. They'll have to rely upon the network to provide it. While the physical problems this represents are far from overwhelming, there's a dump truck full of other problems that have somehow escaped the notice of the NC proponents: How secure will the data stored on these systems be? Who will guarantee that security? What mechanisms will be in place to compensate the home user when the network is hacked (which it will be) and sensitive personal material is vandalized, i.e., it is senselessly destroyed or gets posted on the Web for all the world to see? I can envision the day when a talented kid, sucked in by the irresistible lure of all that personal data, hacks a home NC directory and discovers some "romantic" e-mail, complete with names, dates and explicit actions. Remember when the New York Times got hacked a few months ago? Instead of generic porn, imagine the effect if the hackers had posted some lovers' e-mail? What happens if one of those lovers happens to be married to someone else? With PCs and their data scattered across the world, any hacker trying to zero in on some juicy e-mail has to deal with a highly fragmented, moving target, files that only live on mail servers for a few hours or days. What happens when those files are stored in a user-directory for weeks or months? They become much easier to track down when they live on the same networked server and the overall security of the medium takes a colossal hit.

We can identify another control issue if we consider the lack of options that end users on a network have. Want to demo a game? Fiddle with your desktop? Load a piece of shareware? You'd better check it with the Systems Administrator first. While we (sporadically and grudgingly) accept this situation in the business world, how many of us want this kind of outside control in our homes? Of course, the network will make sure that all of the most popular games are available, but what happens when you want to try out a new e-mail client, browser or graphics program? Will your network provider allow you to load up something that might take the whole network down, inconveniencing several thousand other customers? I doubt it.

For that matter, what happens to the shareware industry as a whole? As many OS/2 users know, shareware titles often represent some of the most convenient, efficient code in the world. I've been using Small Editor, Commander and SIO, to name a few, for years. They are excellent products with a proven track record but they don't have the clout that many of their less-elegant mainstream competitors have. If the world goes NC, you can bet that most, if not all, of these authors and products will disappear, leaving us with fewer and poorer choices. Even if they are successful in making the switch from coding for PCs to NCs, there's very little chance that they'll ever get their product loaded onto the network. The "try before you buy" initiative, something that has been a great opportunity and benefit for the computer consumer, will die a tragic, undeserved death. That alone should be reason enough for any thinking person to object to Wildstrom's suggestions.

Network Infrastructure

As I mentioned above, the information appliances of the future, according to Wildstrom and Horn, will require a network to be fully effective. Wildstrom remarks that a combination of hardware and software should be ready by the end of the year "to turn your existing telephone wiring into an inexpensive Ethernet network of the sort currently used to link up PCs." Better yet, Wildstrom reports that Intel is leading a consortium which is working on a wireless networking system that will automatically link up any devices in its area. While I'll be the first to admit that I'm not a networking guru, my experience with networks as an end user makes me very suspicious. Like most of you, I've seen how productivity grinds to a halt for everyone when the network is down, and while Wildstrom doesn't mention operating systems in particular, you can bet your last nickel that he isn't planning to standardize this system on Linux or OS/2. How's that for a comforting thought? A whole gaggle of wireless, single-use devices hooked up to a Windows server in your basement. Yarg!

Speaking of the network, who will maintain these computer-appliance or NC LANs? If home users can't deal with one, fairly simple machine with links to the Internet and a local printer, are they really likely to learn the ins and outs of each appliance and the home server? I doubt it. This sounds like another opportunity for Microsoft (or someone) to come into your home and charge an outrageous monthly maintenance fee for a service that doesn't exist and isn't needed in the PC world.

What about home NCs? We do have the technology to make them and move them, but we don't have an infrastructure that will support them, not in the kind of numbers it would take to replace every PC in every home or office in the world. Good lord, can you imagine the nightmare that would be? Can you imagine the hue and cry when the network goes down at 11:00 p.m. on April 14, when millions of users are trying to finish up their taxes? What happens when they miss the deadline? How many nuisance lawsuits will be filed against the network by tardy taxpayers who blame the crash for penalties and interest they have to pay instead of their own lack of organization. For that matter, can you imagine the thousands of lawsuits annually that could result from a nationwide network of home users? In a nation that has the world's highest per capita ratio of lawyers to real people, what happens when the network loses your homework? your checking account records? your rough draft of The Great American Novel? The mind boggles. And for those of you who doubt the probability of a disaster like that happening, check America Online's track record. Look at how many times it's sluggish or unavailable, even though it has a small percentage of the traffic that a fully networked United States would represent. Bump that by a factor of 3 or 5 and you've got a reasonable understanding of the mess the NC 'revolution' really offers.

User Inertia

As an educator, I've seen the shift in emphasis on computer skills climb geometrically every year. My own small branch of the Pennsylvania State University overhauled our computer labs in '97 to make sure that our students had more convenient access to the machines. My school-aged kids have taken basic computer intro. courses at the local high school and have two here at home to work on. Around the country, we see more and more children growing up in wired households, surfing the 'net, playing computer games, using word processors. After introducing this generation to the PC, are they likely to dump it in favor of a pile of appliances that have little or no resemblance to the machines they started out with? Quite simply, no. These kids have become accustomed to working in a particular fashion. They're used to the convenience that the conventional PC embodies and they'll show little interest in chucking out everything they've already learned just to start over with a bunch of under-powered gadgets. Their personal inertia will keep them moving on the PC path we've defined.

For that matter, can you seriously imagine any enterprise which will cheerfully chuck its collection of PCs, peripherals and trained support staff, writing off those millions of dollars and man-hours already invested in favor of the appliances Wildstrom and Horn envision? Even though Wildstrom's concern is the home user, he's made the same mistake that IBM continues to make: He's separated home users from business users as though the former never work and the latter never go home. Can you see someone who is accustomed to having the power and convenience of a personal computer at his desk in the office come home and willingly do without it, substituting a host of dedicated, single-purpose machines in its place? I can't.

Of course, the NC doesn't suffer from this particular flaw. It will look just like a PC. The interface will be standardized across the network and, in some ways, this will make learning easier. Schools will tie into the network and their desktops will look like everyone else's desktops. Businesses could do the same. But there's a subtle flaw in this reasoning that has everything to do with the way we perceive our world. I wish I could take credit for discovering it, but I can't. I am indebted to Dilbert creator Scott Adams who analyzes the situation in his latest book, The Dilbert Future: Thriving on Stupidity in the 21st Century (HarperBusiness). With regards to the NC, Adams notes that its proponents believe, "Many people will prefer a low-cost solution, even if it means giving up some functionality and prestige." His counter-argument? "One word: Yugo," and he's right. We identify those machines in our lives that are bigger, faster and more powerful as markers for people who are larger-than-life, more powerful, and more successful. Our political leaders and captains of industry don't drive Geo Metros, even if they should. In the computer world, we brag about the amount of RAM we have installed, the speed of our processors, the size of our hard drives. Can you imagine anyone bragging up an NC? It's about as sexy as my grandmother's housecoat and, considering the current market for teal, quilted housecoats, it will sell just about as well, especially among those kids who are currently learning how to use PCs and becoming members of the online community. The very culture which gave rise to the PC judges technology by its power. Machines which aren't powerful don't cut it. After all, are you driving a Yugo?

The PC is the world's first general information/communications device. Before its creation, we didn't have a single piece of hardware which could process, store and transport information quickly and easily. We relied upon the combination of pencils, pens, typewriters, pads, books, letters, and various means of transportation to do what the PC does today. While these methods met with varying degrees of success, they weren't an integrated solution. For centuries, the speed of communication was exactly equal to the highest speed of transportation. This changed remarkably with the invention of the telegraph. Universal telephone service further changed the landscape and the broadcast media added their own particular features, but until the advent of the PC, the Internet and the World Wide Web, there was a huge information vacuum that separated text from image from voice. That vacuum has been very adequately filled by PCs with Internet connections. It's going to take a lot of digging to open that hole back up and I suspect that the expense on both sides of the equation, for computer consumers and computer-appliance/NC vendors, will be too great to make the project economically feasible.

User Expense

As I mentioned above, the lack of expertise that most end users will have with their new computer appliances and/or their new NCs, will require more maintenance to be performed, not by the user, but by a professional. Legal liability will be another justification for this expense. After all, if you're upgrading your PC's memory and you mess up your SIMM sockets, if you flash your BIOS with the latest version and your machine promptly does a convincing imitation of a $2000 paper weight, no one is inconvenienced but you. If you were to try something similar on your NC, you could take down a whole section of the network, thus, the network provider will make sure that your machine is in working order and will replace your defective SIMMs or faulty hard drive for a fee, making sure, in theory at least, that the work is done by a trained, certified professional. For that matter, the NOS will have to be either remotely installed or have a technician on site to supervise the installation.

You won't be able to do that yourself because of the possibility that the network might go down. The added cost of said professional's salary, health care, pension plan and vacation will be passed along to you. Remember those security issues we talked about? Well, if you've got the bucks, maybe you could upgrade your security status. For an extra, say, $5 each month, your provider will back up your directory and store the media on site (You can't use it, remember? You don't have a drive to read it with. Of course, now you've got to wonder who's making sure the backups are secure.). For another $10, you can get hacker insurance. An additional $15 will get you a faster connection or more bandwidth, maybe greater storage. And so it goes. By the time the service contracts are completed, all of the initial savings that NC proponents tout as a major selling point will be eaten up but the user will still have a monthly connect fee as well as all of these other charges that I've touched on, and probably another dozen I can't imagine. For companies, it's a bonanza, a never-ending flow of revenue that's only limited by their ability to dream up unneeded maintenance or services for their helpless victims. For the customer, it's a never-ending nightmare of charges and fees and the uncertainty of not knowing for sure if the money spent is actually purchasing anything of value.

Tell me again: How is this an improvement over the Personal Computer? An improvement for the end user, not corporate America. I'm afraid I just don't see it.

It's The User, Stupid

As I'm sure you've noticed, I'm looking at this from a user-oriented perspective. That's because I'm convinced that users really do drive the computer industry. While we've seen a number of industry giants (IBM and Microsoft come to mind) who have tried to dictate to the user, we see proof every business day that their control is far from complete. Look at Linux. How many years did this OS remain in the closet, quietly gaining support and respectability while the "experts" weren't even aware of its existence? Now it has gained sufficient momentum that even Microsoft has to take it as a serious threat. Look at OS/2. If certain sections of IBM had their way, it would be as dead as all of the pundits and journalists say it is. But it remains a viable MS alternative because of users like us. Events like Warpstock, publications such as this one, groups like POSSI and SCOUG are all impediments to IBM's grand NC plan. One of the greatest problems IBM faces, however, comes from within. And, for once, their own inept marketing benefits us. After all, if they can't sell a superior product like OS/2 which works within the present, successful business paradigm, how are they ever going to sell something like a Java NC that turns it over and tosses it out?

As for Stephen Wildstrom, his dissatisfaction with the PC has a much simpler, more immediate cure. As he closed his editorial, he posed, what he must have felt to be, a question that most of his readers could closely identify with. And I believe that he was probably right. In fact, I'd bet a sack full of Oreos (one of my favorite vices) that the vast majority of his readers did nod their heads in sympathy and the solution to their dissatisfaction is the same as Wildstrom's. His question? "Wouldn't you love to type a letter or browse the Web without being told you have performed an illegal operation?" We already do that, Stephen. You can, too, without sacrificing the convenience and security of a full-blown PC. Get Warped, Stephen. Get Warped.

Excese me, ma'am, but would you like to buy an Information Appliance? See what other satisfied Network Computer users have said in our interactive forum.