$220 AND FALLING (See photo.)

(May 17, 2005) Today's "personal computer" is not even a true computer, in that it is not a general-purpose device but a proprietary Wintel device. The PC is a corrupted version of the microcomputer vision that we had in the 1970s. I was there. That vision failed when Microsoft hijacked the microcomputer/small computer/home computer as we variously called it. I will show that we have the tools to take back the vision of the computer as a universally available intellectual tool -- take it back from Bill Gates. I will show that globalization is not all bad.

Famous computer visionary Nicholas Negroponte of the MIT Media Lab is developing and promoting a $100 laptop with proposed specifications including a 500-MHz processor, 1 GB of memory and an XVGA display. The Indian company Encore Software already is marking a small computer, the Mobilis, with much more modest specs, for about $230. The Mobilis may not have the features many of you want, but it is a crack in the dam. As cheap computers flood the US, upgraded versions soon will appear -- much cheaper because of no MS tax -- and much better. Both of the above computers employ the open-source Linux operating system (OS). These two machines might not change the world, and nonproprietary operating systems other than Linux might become important, but all this shows what is coming.

Bill Gates will become irrelevant going forward, so this is not about him and how he got where he is. What will kill Microsoft is truly cheap computers. A computer that sells for a C-note, that is, $100, or even for $200, cannot be the basis for a profitable Microsoft. There simply isn't enough money in it. The way to break the MS hammerlock is to starve them out. I use Mozilla and I buy only used computers. I do not benefit Microsoft at all. That is only the beginning.

"APPROPRIATE COMPUTING" MANIFESTO: FOR THE MICROSOFT KILLER, I PROPOSE CERTAIN CORE FEATURES

The Indian Mobilis has some of them. They include:

1. Nonproprietary (free) operating system (NPOS) and nonproprietary (free) applications -- word processor, browser, etc.. This would make the small computer a general-purpose device, as a computer should be -- not tied to Microsoft. Above all, the small computer must AVOID MICROSOFT INTELLECTUAL PROPERTY AND BE ABLE TO PROVE THIS IN A COURT OF LAW. Microsoft basically is a publishing company full of lawyers. (Did you ever see a publishing company get this big or a publisher get as wealthy? Not even Hearst of "Citizen Kane." They exploit the law and technological ignorance.)

2. Instant-on operation -- no waiting for the OS to load from a hard drive. Keep the OS small enough to fit in nonvolatile solid-state memory.

3. At last, an ELECTRONIC computer. What we have now, the PC, includes an electromechanical device, the motorized hard drive -- an electromagnetic device like the relays in the Harvard Mark I of 1943. With "general purposeness" and all-electronic operation (OS on chips, not a disk drive), we would finally have something that meets the definition of a real electronic computer.

4. Simplicity. Since the operating system would be on semiconductor chips, it would be much smaller than the monstrous "whatever the traffic will bear" Windows. A small operating system is a simple one. Remember DOS and the early computers? To start "computing" (limited, admittedly), one had only to know how to turn on the power switch and insert the boot disk. Computer simplicity alone could add several points to the nation's labor productivity.

5. Software drivers for common hardware such as Epson printers and HP scanners. Ability to view, print, edit, and exchange files in Microsoft formats (.doc, .xls, etc.) and to convert to and from standard file formats, including proprietary ones where legal.


ARE THESE DESIGN GOALS REALISTIC?

Of course they are, and Microsoft cannot do anything about this. India already has the Linux-based Mobilis that sells for $220US. It is simple to use because it is a simple machine. It has to be simple because you cannot build Wintel complexity into a $200 computer. And you can't build a MS Windows computer for $200, much less $100. Who needs Windows? We really need to ask.

"What this country needs is a good C-note computer." [I said.] Can we have it? Yes, and Wal-Mart could bring it to us. With that company's merchandising clout, it probably could hope to market profitably a small computer such as the Indian one for $100. Since Wal-Mart is a global operation, there is a strong incentive to do this. The MIT Media Lab is developing a $100 laptop now. Via Technologies of Taiwan offers very low cost computers. And Walmart.com (Wal-Mart's e-commerce site) already offers a $498 laptop that is actually OS-less -- you chose from Linux or Windows. It is only a matter of time before the computer becomes a commodity available at the nearest Wal-Mart store for $100 or $200. I have a $7 twelve-band short-wave radio from China via Odd Lots. A few years ago I would have had to pay over $100 for a Grundig set to get this functionality and quality (no better, and I have been at this for many years). I wouldn't have believed it economically possible until I saw it and grabbed it.

Could Microsoft stop Wal-Mart? No. Microsoft really is just a windbag (puffing 10 times more features than you need for 10 times the price) and a moneybag (lots of cash and market capitalization). If Wal-Mart doesn't do it first, someone else will.

THE INDIAN MOBILIS -- AT LAST A REAL COMPUTER (SEE ILLUSTRATION)

It comes from Encore Software of Bangalore. It has a small VGA (a color-screen standard) LCD screen and rollup keyboard, weighs about one and a half pounds (three-fourths of a kilogram) and opens up to a desktop configuration. It has flash memory (electronically erasable, programmable read-only memory), lots of it, and no hard drive.

The software is based on Linux and developments sponsored by the government of India. Already it has the most-used applications of a computer: word-processing, email, Web browsing, and a spreadsheet. Nonproprietary software and all-electronic operation provide, finally, a real computer for home and office use.

PUTTING THE OS ON A DIET

Cheap nonvolatile semiconductor memory is making it possible to put the OS on chips and eliminate the hard drive, as in the Mobilis. Abandoning the hard drive precludes the use of bloated monstrosities like Windows XP, which squanders 64 MB of storage with bundled application programs, some of which you probably haven't even noticed, and with bells and whistles that are "helpful" like too many cooks in the kitchen.

The game that MS and Intel have used to amass huge capital, and power (but I repeat myself), is reminiscent of the Fifties and Sixties "horsepower race" of American car companies (foolishly reprised with SUVs, a 108-day supply on car lots now). Microsoft develops a new operating system that works less than optimally on current computers. Then Intel designs a new microprocessor with a higher level of integration (a much greater density of circuit components) and enough power to run the latest generation of Windows. A few years later, MS raises the ante and we are in for another round of computer replacements (those of you who fall for this). The rest (non-MS part) of the software industry mainly is engaged in making applications programs -- new programs for work and entertainment -- that exploit the increased power of the Intel and Windows hardware and software. Microsoft bundles as much software with Windows as the traffic, or price, will bear -- making Windows as expensive and problem-prone as they can get away with.

Of course it can be either Intel or Microsoft that initiates the next round of planned obsolescence in computers.

How ridiculous is this? Consider the amount of memory needed to accommodate various operating systems. The following shows how operating systems have grown (the amount of RAM required by operating sytems). Most of the growth since Microsoft got involved with MS-DOS version 1.0 has been the result of the Microsoft-Intel horsepower race (between themselves and the public's gullibility). Basically this is in reverse chronological order (most recent first):

Windows XP ---------------------------- 64MB
Linux Red Hat (Unix clone) -------- 32 MB
Apple Mac OS 8.6 -------------------- 24 MB
Windows 98 --------------------------- 24 MB
Windows 95 ----------------------------- 8 MB
Commodore Amiga OS 3.5 --------- 8 MB
Windows 3.11 --------------------------- 3 MB
Apple Lisa -------------------------------- 1 MB
MS-DOS 6.22 ------------------------- 512 KB
Windows CE (for Pocket PC) ----- 512 KB (in ROM)
Apple Mackintosh OS 1.0 ---------- 128 KB
CP/M --------------------------------------- 20 KB
MS-DOS 1.0 ----------------------------- 16 KB
TRS-DOS (Radio Shack) -------------- 4 KB

Sixty-four megabytes for the operating system is "the craziest darned thing I have ever heard of!" (Bill Gates rhetoric I cleaned up). The OS should fit on an inexpensive amount of flash memory. Eliminating the hard drive gets rid of a lot of cost and Windows to boot (no pun).

I wrote a book in the late 1970s (see list at end) that discussed Tandy Leather Company's offering, the Tandy-Radio Shack TRS-80. A few years before that, I remember programming microcomputers, as we called them then, bit by bit -- this was basically a microprocessor development system using machine language. I remember my relief a year or so later when I could build a Heathkit H8 and program it four bits at a time using a hexadecimal keypad -- this was basically a microprocessor development system or demonstrator. I remember Don Lancaster's "TV Typewriter" in the mid Seventies -- this was basically a dumb terminal. It could be coupled to a time-sharing computer system, but mostly it was used just for displaying 512 characters on a screen and for amateur experimentation (chatting by text via ham radio, for example). Then, in the late Seventies, with the TRS-80 and its contemporaries, which melded CRT screen and keyboard, it was possible to program a home computer in a higher level language at last, BASIC. At the time, our dreams for computing didn't go much beyond balancing your checkbook (I never heard of anyone actually doing this with those relics). This was an era of open systems 35 years ago: Many people designed operating systems for the TRS-80.

While we were playing with these things and naming ourselves "hackers" after model railroaders who hard-wired complex train controls beneath their layout tables (no connotation of malware back then), Doug Engelbart was fathering the mouse and graphical user interface (he invented windows with a small "w") at Xerox in Palo Alto. Theodor Nelson (the man who coined the word "hypertext" as in http and html) was dreaming of the useful and usable computers we still don't have -- and of something like the Web and Marshall McLuhan's "everything all at once." Nelson dreamed of putting the corpus of human knowledge on the equivalent of an e-commerce site (which Google seems headed towards finally). We all were dreaming of useful and usable computers for the masses -- and we still don't have them, thanks to Microsoft, Intel, and people of small technogical visions and large greed. At this time (the Seventies) Bill Gates was dreaming of how to make money with software while most others were counting on doing it with hardware. He never lost his lead at this, but his company never invented anything significant. The last time I was thrilled by his company was when they came out with MS-BASIC.

THE DREAM BEGINS WHERE MICROSOFT ENDS

Progress toward the dream -- cheap, simple, real computers -- will begin again when the cultural imperative becomes global and bigger than Microsoft. This will not be good for Microsoft, but it will be for nearly everyone else. This is about a kind of freedom. The time is at hand.


MICROSOFT AGAINST THE WORLD

Microsoft's leading executive, CEO Steve Ballmer, is trotting about the globe "warning nations about the potential for patent lawsuits if they use Linux." This is sovereign nations Ballmer (maybe we should add an "s" in his name) is addressing. The sovereign State of Microsoft knows no limits to its boldness, but Ballmer and Gates and company are harboring an illusory hope if they think they can duplicate in the world at large their US marketing windfall.

The Mobilis could become the MobilUS, and when it or some other cheap, simple, real computer hits our shores, the emperor of Microsoft will be exposed (financially, too) and we will be on our way toward much, much more useful computers.

RELATED BOOKS BY THE AUTHOR

"Microelectronics" (1976) -- Probably one of the earliest books with "microelectronics" in the title.
"Computerist's Handy Databook-Dictionary" (1979) -- Yes, that's what we sometimes called computer users then.
"Computerist's Handy Manual" (1979) -- This shows how much less visionary I was than Engelbart, Nelson, and Gates.

$220 AND FALLING (See photo.)