Thursday, October 7, 2010

Computers are wonderful and horrible - the view from 1982

Here's another gem from the Atlantic's project of digging up old technology articles from its archives: a 1982 article by James Fallows (who still writes for the magazine) called "Living with a Computer."

The Atlantic's blog post about the article points out how obsolete the references to companies are now, as they've since been the victim of creative destruction:

Nearly all the companies building system components that Fallows mentioned are gone or bit players now: Optek, Ball Corporation . . . , Lanier, Wang, Digital Research, Heath-Zenith, Victor. As Fallows noted, it was like the automobile industry of 1910, "a thousand little hustlers trying to claim a piece of the action." Neither Apple nor Microsoft made it onto his radar.
I find it fascinating to read Fallows describing his encounters with a machine we use all the time without thinking about it, back when this was such a novel experience that it merited a lengthy essay in the Atlantic. I'm tempted to copy and paste the whole article into this blog post, but that would be inappropriate. Here's the gist of it:
I'd sell my computer before I'd sell my children. But the kids better watch their step. When have the children helped me meet a deadline? When has the computer dragged in a dead cat it found in the back yard?

The Processor Technology SOL-20 came into my life when Darlene went out. It was a bleak, frigid day in January of 1979, and I was finishing a long article for this magazine. The final draft ran for 100 pages, double-spaced. Interminable as it may have seemed to those who read it, it seemed far longer to me, for through the various stages of composition I had typed the whole thing nine or ten times. My system of writing was to type my way through successive drafts until their ungainliness quotient declined. This consumed much paper and time. In the case of that article, it consumed so much time that, as the deadline day drew near, I knew I had no chance of retyping a legible copy to send to the home office.

I turned hopefully to the services sector of our economy. I picked a temporary-secretary agency out of the phone book and was greeted the next morning by a gum-chewing young woman named Darlene. I escorted her to my basement office and explained the challenge. The manuscript had to leave my house by 6:30 the following evening. No sweat, I thought, now that a professional is on hand.

But five hours after Darlene's arrival, I glanced at the product of her efforts. Stacked in a neat pile next to the typewriter were eight completed pages. This worked out to a typing rate of about six and a half words per minute. In fairness to Darlene, she had come to a near-total halt on first encountering the word "Brzezinski" and never fully regained her stride. Still, at this pace Darlene and I would both be dead—first I'd kill her, then I'd kill myself—before she came close to finishing the piece. Hustling her out the door at the end of the day, with $49 in wages in her pocket and eleven pages of finished manuscript left behind, I trudged downstairs to face the typewriter myself. Twenty-four hours later, I handed the bulky parcel to the Federal Express man and said, "Never again."

Over the next few weeks, my thoughts often drifted to the advertisements I had seen in airline magazines, in which trim and cheerful secretaries effortlessly produced documents by typing in front of computer screens. Were these devices real? I checked with a salesman for a company called Lanier and discovered that while their word-processing system, called "No Problem," was quite real, it cost some $15,000. If I had drawn a pie chart representing my annual income, No Problem would have been a very large piece of pie. I called Wang, Digital Equipment, and some of the other big-name manufacturers and got roughly the same news. If I called the same manufacturers today, I'd hear much more encouraging news, but my options were to start writing shorter articles, go into hock, or take my chances again with Darlene.

The way out of my dilemma came from an unexpected quarter. My father-in-law often dealt with inventors who put together computer systems to monitor various industrial processes, and he thought that one of them might have the answer. On his advice, I followed a trail of leads and suggestions that eventually led me to a converted church in the farmlands of central Ohio. There, Bill Cavage, Marv Monroe, and Bill Jones, three young engineers doing business as the Optek Corporation, tinkered with disk drives, photo-sensors, and other devices in hopes of making the big sale. Optek's specialty is making machines like the one they produce for drug companies, which counts pills as they pass by at a rate of 24,000 per minute and kicks out any bottles that receive the wrong number of pills. For men who can do all this, I thought, turning a small computer into a word-processing system should be a cinch.

For a while, I was a little worried about what they would come up with, especially after my father-in-law called to ask how important it was that I be able to use both upper- and lower-case letters. But finally, for a total of about $4,000, Optek gave me the machinery I have used happily to this day.

The ingredients were the basic four of any word-processing system. First was the computer itself, the Processor Technology SOL-20. Its detailed specifications—its 48K of random access memory, its Intel 8080 microprocessing chip—are now of antiquarian interest, since Processor Technology went out of business several months after I bought my computer.

The second element in my system was the monitor, a twelve-inch TV screen. Some monitors are like black-and-white TVs; mine—which, oddly enough, was produced by the same company, Ball Corporation, that makes home-canning supplies, displays light-green letters against a background of dark green and is supposed to be easier on the eyes. Third was the external storage device—the equipment that saves the documents you've written when the computer is turned off. The equipment I chose, two small tape recorders, was such a complete disaster that I must discuss it separately later on. Fourth was the printer, a ponderous machine, built like a battleship, which had been an IBM Selectric typewriter before it was converted to accept printing instructions from a computer.

These four machines, and the yards and yards of multi-strand cable that connected them, were the hardware of my system. The software consisted of a program called The Electric Pencil, with a manual explaining the mysteries of "block move," "home cursor," and "global search and replace."

I skip past the day during which I thought the computer didn't work at all (missing fuse) and the week or two it took me to understand all the moves The Electric Pencil could make. From that point on, I knew there was a heaven.

What was so exciting? Merely the elimination of all drudgery, except for the fundamental drudgery of figuring out what to say, from the business of writing. The process works this way.

When I sit down to write a letter or start the first draft of an article, I simply type on the keyboard and the words appear on the screen. For six months, I found it awkward to compose first drafts on the computer. Now I can hardly do it any other way. It is faster to type this way than with a normal typewriter, because you don't need to stop at the end of the line for a carriage return (the computer automatically "wraps" the words onto the next line when you reach the right-hand margin), and you never come to the end of the page, because the material on the screen keeps sliding up to make room for each new line. It is also more satisfying to the soul, because each maimed and misconceived passage can be made to vanish instantly, by the word or by the paragraph, leaving a pristine green field on which to make the next attempt.

My computer has a 48K memory. Since each K represents 1,024 bytes of information—each byte representing one character or digit—the machine can manipulate more than 49,000 items of information at a time. In practice, after allowing for the space that The Electric Pencil's programming instructions occupy in the computer's memory, the machine can handle documents 6,500 to 7,500 words long, or a little longer than this article. I break anything longer into chunks or chapters and work with them one at a time.

When I've finished with such a chunk, I press another series of buttons and store what I have written on my disk drive. This is a cigar-box-shaped unit that sits next to my computer, connected through a shocking-pink ribbon cable containing thirty-four separate strands. Inside the drive is the floppy disk, which is essentially magnetic recording tape pressed into the shape of a small record and then enclosed in a square cardboard envelope, 5 1/4 inches on each side. The system transfers data from the computer to the disk, or vice versa, at about 1,000 words per second, so it is no nuisance to pause after each fifteen or twenty minutes of writing to store what I've just done. Each of the disks in my system can hold about 100K of information, or more than twice as much as a full load from the computer memory. If one disk is full, I pull it out and snap another in.

When I finish what I'm working on, I switch on my printer. If I'm sending a letter, I load the stationery into the printer and push the print button, and then fish each piece of paper out of the printer when it is done. There are machines that automatically feed single sheets of paper into the printer, but that takes us back to big slices of the income pie. If I am printing a draft of an article, I can hook up my tractor feed, push the print button, and go out for a beer. The tractor pulls an endless sheet of paper through the printer—and the perforated paper can be separated into pages when the printing is done, so it looks like a normal manuscript.

The system prints about thirty characters per second, which means it takes less than a minute per double-spaced page. When it has completed its work, I take the manuscript and start working it over with a pencil, just as I did in days of old. The difference is that after I've made my changes, I have only to type in the changes I have made and start the printer up again—rather than retype the whole mess.

None of this may sound impressive to those who have fleets of secretaries at their disposal, or to writers who can say precisely what they mean the first time through. Isaac Asimov recently complained in Popular Computing that his word-processor didn't save him much time on revisions, since he composes at ninety words per minute and "95 per cent of what I write in the first draft stays in the second [and final] draft." My first-draft survival ratio is closer to one percent, so for me the age of painless revisions is a marvel.

You won't catch me saying that my machine has made me a better writer, but I don't think it has made me any worse. Since I now spend less time and energy retyping, I have more left over for editing and rewriting.
I think he's basically right, and if anything he's understating how much of a boon computers have been to the writing and editing process. They don't make writing quicker at the expense of quality; they improve the quality by allowing you to write and edit more quickly.

He then describes a strange way to use technology to improve editing, which I had never heard of:
When I think I'm finished with an article, I set the print speed to Slow. This runs the printer at about 100 words per minute, or roughly the pace of reading aloud. I stuff my ears with earplugs and then lean over the platen as the printing begins. Watching the article printed at this speed is like hearing it read; infelicities are more difficult to ignore than when you are scooting your eye over words on a page.
I have not yet stooped to the politician's trick of programming the computer to write standard letters of reply. I have, however, discovered a few other sneaky word-processing feats. Suppose you are writing an article in which an unusual word appears frequently—let us choose "Brzezinski" once again. When writing the draft, you simply type a certain character, say * or + , each time Brzezinski should appear, and then when you're ready to print you signal the computer to insert "Brzezinski" in place of the character.
Imagine thinking that this keyboard shortcut was so devious it was worth mentioning in the same breath as having an artificial-intelligence ghostwriter!

He also used a precursor to TurboTax:
In addition to The Electric Pencil, I bought the software for a computer-programming language known as BASIC . . . . When I want to know how many prime numbers there are between one million and two million, or how quickly my mortgage payments would bankrupt me if interest rates rose to 35 percent—that is, when I don't want to do my work—I can kill ten minutes writing programs to tell me the answer. Getting down to business, I use the computer to do my income tax. My economic life is a mess of $2.75 parking-lot tickets and $13.89 lunch receipts, which used to pile up like fall leaves until I spent a week burrowing through them at income-tax time. Now all I do is sit down at the machine for five minutes every few nights and type in all transactions of interest to the tax man—so much in from my employers, so much out to the credit-card company. At the end of the year, I load the income-tax program into the computer, push the button marked "Run," and watch as my tax return is prepared. Since it took me only about six months to learn BASIC (and the tax laws) well enough to write the program, I figure this approach will save me time by 1993.
Fallows goes on to discuss a long litany of technical issues that are now happily irrelevant to us. Should you use tapes or disks to store your information? If you use disks, should they be 5 1/2 or 8 inches in diameter? And should they have "single-density" or "double-density storage"? Fallows reports that he bought two 5 1/2-inch disks, but was frustrated that "each disk fills up too quickly and I have to keep rotating different disks in and out of the drive."

He admits that his computer has caused serious problems:
For one, it creates yet another reason to feel vulnerable to the workings of fate.

Shortly after I got my machine, I was typing away in the basement as a summer thunderstorm moved into town. Knowing what I now know, these days I immediately shut off the machine and unplug it from the wall whenever thunder is reported any nearer than West Virginia. But I was not so wise then. I had turned on the printer and gone upstairs when a bolt of lightning struck the house. There was a huge boom, and a white flash outside windows on all sides of our house. Several million volts coursed through our wiring and blew out nearly every electrical appliance that was plugged in. The blast burned out a clothes iron, and if it had that effect on a big hunk of steel, you can imagine what went on in the computer's delicate interior. For a month I was machine-less, thrown back on my Smith-Corona, while computer repairmen replaced one silicon chip after another that had melted in its casing.

Computers cause another, more insidious problem, by forever distorting your sense of time. When I first saw the system in the back room at Optek, I was so dazzled by the instantaneous deletion of sentences and movement of paragraphs that I thought I could never want anything more. When the scientists at Optek warned me about certain bottlenecks, I had to stifle my laughter. In particular, they warned me that I might grow impatient with tape recorders as a way to store data. You have to understand, they told me, it can take five or ten minutes to load a long draft into the computer from tapes, whereas a disk drive (which would add a thousand dollars to the cost) could do the job in seconds. Typical vulgarians of the machine age, I told myself. How could they imagine that I would object to five or ten minutes, when I had been spared Darlene?

Three weeks later, I was griping constantly about the tapes and scanning the pages of Byte magazine, looking for a good deal on a disk drive. Ten minutes was intolerable when everything else happened in a flash. Worse, the tapes had the fatal defect of unreliability; even after waiting ten minutes, you were never quite sure that the information was safely stored. The only way to tell was to try to feed the data back from the tapes into the computer, which took another ten minutes and often led to the infuriating message "Tape Error." After one article disappeared forever behind a thicket of Tape Errors, I scraped up $800 for a cut-rate disk drive. Now my discontent is awakened only when I read stories about the new disks—larger ones that hold twice as much data as mine, and double-density models that hold twice as much as that.

I CAN HARDLY BRING myself to mention the true disadvantage of computers, which is that I have become hopelessly addicted to them. To the outside world, I present myself as a man with a business need for a word-processing machine. Sure, I have a computer: I'd have a drill press if I were in the machine-tool business. This is the argument I make frequently to my wife. The truth, which she has no doubt guessed, is that I love to see them work.

I nearly destroyed my health, to say nothing of my marriage, during the months when I switched off The Electric Pencil at ten or eleven at night-and then switched on BASIC and spent a few hours refining a tax-and-accounting system. At first my goal was merely to design a program that would work, that wouldn't print "Syntax Error in Line 2140" when I tried to run it. Then I started playing around, seeing if I could work out a scheme for financial projections that would take care of estimated tax payments to the IRS. Would it have been easier to mail in the $150 each quarter and then square accounts with the IRS at the end of the year? Of course—but that was not the point. Eventually, I aspired even to elegant programming, designing the matrices and the nested loops in a way that added the beauty of simplicity to the scheme.

When I contemplate my future with computers, my emotions are mixed. Because time and progress have passed my machine by, I simply can't buy any new programs for the SOL. They don't exist. This is a source of unending frustration: how I'd love to use a new word-processing program, one that could insert footnotes at the bottom of the proper page or automatically prepare an index for a book. How I'd love to get VisiCalc or SuperCalc or one of the other accounting systems that can turn a home computer into a miniature version of the National Bureau of Economic Research. How deprived I feel as I read the fliers for CompuServe and The Source, the over-the-phone services that enable you to make airline reservations, call up old newspaper articles, and send computer mail, all from the privacy of your home. How I wish my employers would install computers in their headquarters, so I could submit articles over the telephone, one computer to another, instead of fighting the crowds at the Express Mail window.

Yet even as I think these thoughts, I fear their fulfillment. My computer already competes with wife and children for my affection: can our family stand anything more? The question will remain moot until the price of replacement computers comes down a little more—or until I succeed in convincing my wife that she, too, needs a computer, so I can give her mine and rush out to buy a new one for myself (for business purposes, of course).
He ends by telling us of his shockingly modest hope "for a world in which my sons can grow up to have a better computer than their father had." Uh, yeah!


(Photo of my brother by me.)