Feb 1996 Opinion Columns

by: Fred Langa, Editorial Director

Fred Davis eloquently presents the "pro" side of the debate raging over these cheap Internet-only boxes. The idea is seductive: For about $500--a magic number for generating volume sales--you get a World Wide Web-ready computer that's basically just a front end. Your applications and data--perhaps even your operating system--reside on the Net. Eliminating local hard drives and lots of RAM, and encouraging on-the-Net centralization of expensive computing functions, yields economies of scale and fabulously cheap computing. Companies like IBM, Sun and Oracle have enthusiastically discussed future products along these lines. It's a huge, exciting concept--check out Fred Davis' column in this issue for a fuller explanation.

But although Fred and others may be bullish on these new computers, I think the idea ranks right up there with IBM's ill-fated PCjr. Here are 10 reasons why:

1. Chickens and eggs. The $500 price point is fantasy. A minimal Windows-based diskless workstation--a rough approximation of the Network PC--costs about $750. For a Network PC to sell for substantially less, its sales volume would have to dwarf Windows' (now around 70 million copies). Does that seem likely to you?

2. Your current software won't work. To use the new Network PC, you'll have to throw away your existing operating system and applications, and buy or rent new downloadable online apps--raising your investment considerably.

3. A box without a connection is useless. The Network PC's lifeline is its data connection, but how will it connect? A cable modem? ISDN? ATM? The hardware for these is expensive. Heavy volume may bring the price down, but item 1 shoots down the "economies of scale" argument. Fast modems? Gimme a break. Even at 28.8Kb per second, downloading a Web page can try your patience. Downloading your operating system or all your applications at that speed is nuts.

4. Web browsing isn't TV (Part 1). Some of the $500 Network PCs being discussed lack keyboards and monitors. Keyboards may not be very expensive, but monitors are. Some vendors are talking about using the TV as a monitor--a throwback to the days of Pong and the Commodore 64. TVs are horrible at handling anything denser than blocky 40-characters-per-line text. Many Web pages would be illegible or just plain ugly on a TV screen.

5. Web browsing isn't TV (Part 2). Most dens or TV rooms are arranged for passive vegging out at a significant distance from the screen--the antithesis of how users interact with their computers. You think John Q. Public will burn his Barcalounger and decorate his den around an office-furniture motif?

6. "All our support technicians are busy, but please stay on the line ...." How do you feel when America Online is slow, or your local CompuServe node goes south, or your phone or TV's out? Today, these too-common occurrences are inconveniences, but if you depended on your browser box for your computing needs and you couldn't connect, you'd be toast.

7. Hacker heaven. When your data resides far away and has to travel by wire to reach your fingertips, you have to worry about human error, technical glitches and plain old greed: Your data is just a few passwords away from any interested hacker or cracker. Conversely, when you have to download all your software each time you run it, the virus risk is immense.

8. Competition from low-end PCs. I've seen Pentium 60s, including a cheap monitor, keyboard and hard drive, on sale for about $900, which is about what a Network PC would cost once it's set up, connected and equipped with applications. Even if the PC cost a little more, which would you pick: a single-purpose browser box or a do-anything general-purpose PC?

9. It's business, not revolution. The vendors pushing cheap Web boxes hope to be at the center of things. They make the databases that will house the applications and data of browser-box users, or the big-iron servers that could shovel data down the wires, or the low-cost chips that would power these cheap terminals. This isn't an evil thing--companies have to make a buck. But before you jump on the bandwagon, make sure what's good for the vendors is good for you. And that leads me to:

10. Impersonal computing? Call me a dinosaur, but I thought the whole point of personal computing was to put computing power into the hands of end users. A PC is your "anything box" that can process words, store data, communicate, crunch numbers, play games--all entirely under your local control. By taking the "P" out of PC, you shift its power and flexibility back to a distant, glass-walled mainframe room. That seems like a huge step in the wrong direction.

Maybe I'm wrong. Would you buy a $500 browser box, or recommend one to your friends? Drop me a line at flanga@cmp.com, or visit our discussion Web site at
http://techweb.cmp.com/gurus/langa/langa.htm.

by: Mike Elgan

In my December column, I mentioned that Microsoft was working on a secret telephony application. Does that ring a bell? Now I can give you the exclusive details.

It's called Microsoft Phone. This all-in-one Windows 95-only telephony application functions as a telephone, speakerphone, answering machine, PBX, interactive voice response (IVR) system and personal assistant. You can't buy it separately--only as part of third-party telephony kits and telephony-ready Win95 PCs.

Microsoft cupped a protective hand over the Phone project for a couple of years. At press time, the company was scheduled to announce the product in January and ship it shortly thereafter.

MS Phone just might mainstream computer telephony (the joining of telephones and personal computers). I'll tell you why in a minute.

It's designed to be as easy to use as a real telephone (it doesn't quite succeed, but it's still very easy) and sports some truly bleeding-edge technology.

by: Fred Davis

According to Larry Ellison, CEO of Oracle Corp., the NC will be the first mainstream information appliance, costing under $500 and providing Plug-and-Play access to the information superhighway. Oracle has put its money where its chairman's mouth is by having its R&D team design the NC. In fact, I was lucky enough to be one of the first to eye the schematics.

Oracle is inaugurating the NC effort in the same open market spirit that made the PC so successful. Rather than get into the hardware business itself, Oracle has wisely decided to stick to its knitting and stay a software company. Instead, Oracle is offering the NC's schematics and design for a nominal fee to any manufacturer interested in making them. This is the clone strategy that helped the PC and hurt proprietary systems like the Macintosh. The NC is off to a strong start because of this open-minded strategy, and several large Asian manufacturers have already expressed interest in making the devices.

But Larry and I aren't the only ones excited about the NC. IBM chairman Lou Gerstner has hailed it as "the third wave of computing." In Gerstner's world view, the first wave was the mainframe era, the second was the PC revolution and the third, created by the NC, will "change the nature of computing."

Pretty heady stuff.

And there's something else that Lou, the chairman of the world's largest computer company, and Larry, the richest guy in California, have to be excited about. Their companies, IBM and Oracle, offer networked computing a lot of technology that Microsoft doesn't. In other words, the NC market is fresh meat, and in the microcomputer arena, it's the first fresh meat in a long time that hasn't been Intel or Microsoft's primary prey.

Ellison and Gerstner aren't the only industry lions pouncing on this opportunity. Scott McNealy of Sun Microsystems said that the company will leverage Java as a new operating system and application delivery system for the new NC platform. Apple, always eager to point out that it invented everything first, said that its Newton and General Magic spinoff are both efforts at developing some form of NC.

Although the ultimate NC standard is still up for grabs, here's the basic concept behind these machines. An NC would be a network client computer with a 32-bit processor, roughly 8MB of RAM and no disk drive. Some models would plug into your TV, using it as a monitor, and others would be more like a laptop and include a flat-panel screen. You could use the NC to surf the Net, send and receive mail, and run a host of interactive applications from games to videoconferencing.

Now, getting back to that missing disk drive. One of the NC's primary features is that you'd never need to buy or upgrade an application; you'd simply run it off the Net. Of course, there's no such thing as a free lunch. What you would actually do is "rent" your applications as part of your online fees.

The implications of this are truly mind-boggling. This would create an altogether new computing paradigm.

Sun's famous line is "the network is the computer." But in this new paradigm, the network is not just the computer; it's also the operating system and application software as well as the data that the software uses.

This paradigm is based on a classic client/server architecture in which an application doesn't simply run on a PC, but instead, a portion of it runs on the server and a portion runs on the client. This structure allows the client, an NC, to behave like a powerful system with vast storage resources--it can take advantage not just of a single server but of a whole World Wide Web full of servers.

In computing's first wave--mainframes--hardware was the ultimate platform, and whoever controlled the hardware ruled. That was IBM.

In the second wave--personal computing--the operating system became the platform, and Microsoft acquired the keys to the kingdom.

In the third wave, the network is the platform. But Gertrude Stein's comment on the sprawling city of Oakland, "There's no there there," could also apply to the vast land of networking. And this means that the new wave created by NC promises an end to feudal computer kingdoms and the beginning of a true information democracy.

The PC's kingdom will certainly prosper for a long time to come. But if the democratic principles of the NC come to fruition, it will create a vast new computing market that no single standard will dominate. That's because NCs run all sorts of operating systems and use all kinds of processors. They only need to provide compatibility with the Net's information and software standards. In other words, as long as an NC has an Internet-compatible browser, it will provide access to anything on the Internet.

The Internet is an open standard that you can build on freely, and currently there are more than a dozen browsers from Netscape to Microsoft to IBM. The same goes for servers, which run everything from UNIX to Windows to NT to Macintosh.

Just because Microsoft isn't in the third wave's start-up, don't count it out. Ellison and Gerstner's enthusiasm for the NC concept was not lost on Bill Gates. After all, whether it was DOS, the Windows interface or a spreadsheet, Microsoft has proven itself time and again as the great imitator, not the great innovator. Microsoft is already hard at work on its imitation, based of course on Windows and its Internet Explorer browser.

But Windows isn't a killer advantage in the NC market the way it is in the PC market, so I'm looking forward to seeing some genuine competition emerge. And that should be good for everybody.

by: Cheryl Currid

The notion of the workplace is changing. A Market Research Institute survey revealed that 64 percent of U.S. Fortune 1,000 companies have telecommuting programs, and 60 percent of the companies without programs plan to start one within the next three years. A survey conducted by Link Resources projects that the number of home-based workers will rise to about 50 million within five years. Even if you do only a portion of your work at home, chances are you're part of the workplace revolution.

Mushrooming sales of PCs intended for home use have made the home office the industry's hottest growth sector in the past two years. Surveys done by the Electronics Industries Association indicate that half the homes with computers have a hookup to the Internet and are set up for multimedia applications--a figure that has doubled since 1990.

Home may be where your PC is, but it isn't necessarily where you do all of your work. Many jobs require you to be productive in multiple locations, from customer sites to airline seats. So does it matter where you work as long as you get your work done? Probably not. The rules about how, when and where your job gets done have changed.

Fueling this trend--and making work feasible almost anytime, from anywhere--are two factors crucial to just about any modern-day social movement. One is consumer acceptance, and the other is the recent flood of new paradigm-busting technologies.

A scant few years ago, acceptance of technology was the province of bona fide computer geeks and forward-thinking early adopters (like you and me). Today, having a computer is mandatory in most professions. Whether you're a manager, a marketeer or an analyst, you derive data and draw conclusions with a computer. If you don't, your colleagues are likely to distrust your recommendations.

The advent of alternative work styles such as telecommuting and virtual organizations has made it critical for many businesses to make their resources available to users even when they're not in the same location or time zone. To do this, they rely on fast modems, ISDN connections, advanced paging devices and other wireless services.

That brings me to the deluge of technologies being delivered by a powerful information fire hose. Some represent small steps of progress, such as Windows 95's Briefcase icon, which lets you quickly transport files from one computer to another. Other events, such as the new desktop videoconferencing products, amount to major technology revolutions.

Intel's ProShare and Creative Technology's ShareVision let you fine-tune your ability to use your PC as a video phone. And with Compaq and Intel's recent announcement of technology partnering, look for PCs with built-in videoconferencing components. Other enhancements like VideoServer's multipoint server also let you do more online. Users in multiple locations can join conferences, extending the power and promise of desktop video technology.

Consider how this powerful technology will affect, say, banking. Rather than maintain unprofitable branch locations on expensive real estate, a bank may decide to occupy a few square feet in the local grocery store. With video conferencing, a teller at the bank can appear onscreen as needed to offer loan applications, stock market services or other investment products. Once consumers have their own videoconferencing and communications equipment, they won't even have to leave their homes to use such banking services.

Aside from breakthroughs like these, plenty of other new technologies enable you to work from almost anywhere. Server Technology, for instance, announced several new remote control power products late last fall. For as little as $100 to $200, you can set up controllers that let you call your PC to reboot it or turn it on or off. For $700 to $1,300, you can buy similar equipment to control multiple hosts. This provides a big boost for multi-PC users and workgroups who need access to their data from home, the office or the road. They can call the computer or network server that has the data, turn it on and then dial in.

Multipoint, multifunction communication is vital to emerging work styles. Organizations are coming to understand that no single computer can house all the data people need to make decisions. Rather than try to stuff all relevant information into one place, smart organizational technology managers deploy a host of communications options and strategic storage-management products.

In recent months, I've seen several IS managers hook up corporate workers to multiple information sources. One manager, for instance, issues log-ons to the Internet, CompuServe and a private online research service as part of the standard operating software. He's also investigating hierarchical storage management (HSM) products for corporate data. This technology has the potential to remove the location requirement for data by providing an orderly way to store it, either on the Net or on a server.

Users' appetite for readily available data seems insatiable. With more than 20 million LAN nodes and more being added daily, data-storage needs multiply quickly. One researcher suggests the average PC LAN server with 2.25 gigabytes of storage will require more than 8GB within two years. Even though hard-disk costs have plummeted to between 25 cents and $1 per megabyte, the cost to manage the data has held constant at about $7 per megabyte.

But every technology-induced problem seems to spawn a technology-created solution. Enter document management vendors with HSM solutions. Kodak Digital Science's Imagery HSM product, for instance, manages documents and data by allowing data to migrate across multiple servers. You don't have to know where the data is; it just arrives upon request. This type of product is a boon to any organization struggling with data management.

Sound too good to be true? This is what technology can do today. You can use your imagination---supplemented by your wildest dreams--to anticipate what's in store for tomorrow. New work styles are coming quickly. They create opportunities and competitive situations, and offer ways to improve organizations' operational efficiency. Stay tuned.

by: George Plimpton

It derives from the sportscaster Keith Jackson's signature cry "Whoa, Nellie" in moments of high excitement. I find a certain logic in giving names to inanimate objects on which one depends--automobiles, toasters, lawn mowers and so on. There's an automatic teller machine on Second Avenue to which I assigned a name after it stubbornly refused to give me money one weekend. I named her Dorothy, and ever since then she's done what I've asked of her.

When it comes to something as quirky as a computer, it's essential not only to name it but to talk to it constantly. "Hey, Nellie," I say as I flip the on switch.

I use Nellie mostly for word processing, but that's a misnomer. She does no such thing. I don't see much processing going on, at least as I use the term. A roll of film, for example, is processed into negatives, which are turned into finished prints. A word processor doesn't process words. It merely hangs on to them.

Perhaps it's too easy to complain about newfangled devices, but it seems to me a word processor should do more. A famous John Cheever story based on the laws of probability states that if a dozen or so chimpanzees were put to work banging away on typewriters, they would eventually copy out any number of masterpieces word for word. In the Cheever story, the chimpanzees actually reached this stage and finished their work on Dickens or whomever. Fiction for sure, but it does cross my mind that the average word processor contains far more brain cells (or bytes) than an average chimpanzee.

If you cupped a chimpanzee's head between your hands and with great earnestness told it you were writing a seafaring tale about catching a white whale, and you then implored the chimp for a first line-- "Please, Oscar"--a little spit might generate on Oscar's lower lip, but not much else. On the other hand, a word processor, stuffed to the gills with all that know-how (thesauruses, dictionaries, built-in grammar checkers and so on), ought to be able to hum a bit and come up with, "Well, how about `Call me Ishmael'?"

I treasure first lines and have often thought a great first line impels one forward to good writing, almost as if you were pushed through an open door. How I envy the first words of George Orwell's 1984: "It was a bright, cold day in April and the clocks were striking thirteen." Or, "Lolita, light of my life, fire of my loins," from Nabokov's book of that name; or, "James Bond, with two double bourbons inside him, sat in the departure lounge of the Miami Airport and thought about life and death," the opening line of Ian Fleming's Goldfinger. I'm rather proud of the first words of my own Paper Lion ("I finally decided to pack the football"), but they came to me not out of a word processor, but while I was standing in a public telephone booth. How much more convenient it would be if such inspirations came out of appliances plugged into electrical outlets in one's home.

Indeed, as it hums, why shouldn't the word processor produce the original quip or aphorism from time to time? The idea would be to turn it on for the night and when one comes down for breakfast in the morning, the screen would have something to offer: "Book reviewers are little old ladies of both sexes." Or, "The scientific theory I like best is that the rings of Saturn are composed entirely of lost airline luggage." Or, "A gentleman is a man who knows how to play the accordion but doesn't." The authors of the first two are John O'Hara and Mark Russell, but the third is unknown. This is the best kind because you don't have to acknowledge that someone else said it first when you use it later that day at a cocktail party.

Writers say one of the great advantages of word processors is their built-in editing capacities. One phenomenon I've noticed as a literary magazine editor is that the individual manuscripts submitted to The Paris Review (more than 20,000 a year) are almost invariably error-free. The margins are exact. The pages are numbered. The copy looks so perfect it's presumably ready for the presses. It always comes as a shock when what seems so slick and polished turns out to have no literary value.

The manuscripts are also too long; they suffer from logorrhea. They're inventive, and written with apparent ease by authors who know what they're doing, but because every action is detailed, often with many lines of expository dialogue, the manuscripts are strangely boring and almost impossible to read through. Their creators seem unaware of G.K. Chesterton's admonition, "Murder your darlings," or Isaac Bashevis Singer's suggestion, "The wastepaper basket is the writer's best friend."

An interesting thesis could be put forth about how literary styles have changed as various inventions have come along to ease the process. The wiring between the hand and the brain is different in each case--each method producing a different syntactic reaction. Writing with a pen is analogous to drawing. With a typewriter, you form sentences by striking the keys; with a word processor, you perform a visual analysis of what's on the screen. A writer's style is bound to change as different modes become available. When at the end of his life Henry James began dictating his novels, his sentences to many became interminable and convoluted.

So, should one stick with the tools one knows best? It always fascinated me that when Ernest Hemingway started on a project, he always began by using a pencil and writing on onion-skin typewriter paper. He shifted to the typewriter only when his thoughts were flowing quickly, or when the writing was, for him at least, relatively simple dialogue.

I've decided to keep my pencils sharpened and my grandfather's Underwood on tap just in case. But I'm going to give myself over to Nellie. Perhaps when she and I grow more accustomed to each other, I'll get going on that novel. And with her help, I'll crank out a thousand pages or so--maybe more if I learn to touch-type. She might even help me with a title. By that time, who knows?