Programmed for Success
By Tom Mueller, HemispheresThe digital dynamics of creating great software don't have to depend on corporate empires with cadres of developers. The Internet is letting independent programmers share their wares with customers one-to-one.
Tom Davis lives in Vermont's Green Mountains and writes gorgeous code. Mornings he arises anywhere between 5 and 10, brews a pint of wickedly strong Sumatra, powers on his wall-to-wall stereo system, and boots up. He sips and types and toe-taps through the day, deep in Code Mode, his name for the ideal programming state, with the golden guitar of Frank Zappa and Joni Mitchell's rich vibrato filling the air. When the flow of ideas slackens, Davis goes for long walks in the woods that surround his house or sits on the porch and watches the clouds drift through the surrounding peaks and on out to the distant horizon. Then, refreshed, he returns to the computer and forges ahead, often long into the night. When his mind is exhausted, the workday finally ends. This has been going on, with a few brief interruptions, every day for the last seven years. "Sometimes I don't even know what day it is," he says. "One time I discovered I didn't know what month it was."
The object of Davis' obsession is Zoot (www.zootsoftware.com), a computer program that is uncannily good at gathering, organizing, and synthesizing knowledge. It is a Personal Information Manager of sorts, though calling Zoot a PIM is like calling the QE2 a boat: It's vastly more elegant and powerful than the digital daybooks typically sold under this rubric. Over the last four years I've watched Zoot evolve through perhaps 200 versions, milestones in Davis' tireless quest for the Holy Grail of information processing. Version 4.0, released this August, can do almost anything. With Zoot running in the background, I highlight text in my browser, word processor, or other Windows program and clip it straight into Zoot. Zoot synchronizes automatically with Microsoft Outlook, pulling in e-mails, contacts, and tasks. Once inside Zoot, all this information can be sifted, searched, filtered, and viewed in a dazzling number of ways. I can manually sort each item into storage folders or set up "Smart" folders that automatically categorize, reformat, or otherwise process my data. Zoot passes the ultimate test of good software: It helps you think better.
Yet this remarkable tool, and Davis' equally remarkable lifestyle, would be impossible without one crucial concept: shareware. In broad terms, shareware is a computer program that people can download free from the Internet, to try for a limited time. If they want to use the program beyond the trial period they must purchase it; otherwise they simply delete it from their system. The concept of shareware arose in the late 1970s, when computer enthusiasts began to swap clever programs over early electronic bulletin boards. In 1982 shareware proper was born when some programmers began asking their users for money. Back then, a popular program might get 1,000 downloads.
Now popular programs are downloaded 50,000 times a day, and the chat program ICQ ("I Seek You"), just topped 100 million total downloads from one shareware site alone. Scores of Web sites now specialize in collecting, reviewing, and disseminating shareware programs, some reporting 20 percent annual growth, others offering as many as 50,000 programs for immediate download. The shareware explosion is due to a convergence of factors, like increasing bandwidth and the growing acceptance of credit cards on the Internet. But the driving force behind shareware is the expansion of the World Wide Web. Estimates of Web use run as high as 106 million in the United States -- 53 percent of the nation's adults -- and 350 million worldwide, climbing to 766 million over the next five years.
Launch a Web site, and the shareware author gains immediate access to 350 million potential customers. (So far people from 30 countries have visited Tom Davis' site and bought a copy of Zoot.) The site becomes a 24-hour store with no staff to pay, no rent, zero costs for marketing, sales, and distribution. "Try before you buy," a popular marketing strategy in many sectors, is particularly compelling in software, where aesthetics and system compatibility are equally decisive. Software is also the ideal Web production packaging, no logistical logjams, just a brief, blissfully clean stream of data.
The shareware model makes such good sense, in fact, that big software companies have begun to adopt it. Symantec, the Silicon Valley colossus, now offers many of its products for sale at its Web site (www.symantec.com) on the same terms as Tom Davis and other shareware authors. Microsoft and Netscape do Symantec one better, offering their Web browsers for free. "Adware" has emerged, software that's yours for free if you can bear the steady stream of banner ads (and in some cases, the discreet monitoring of your browsing habits for marketing purposes). But perhaps the most striking changes have taken place at the level of small, independent software developers. "The Internet has increased the pace of creativity in shareware incredibly," says Preston Gralla, Executive Editor for ZDNet (www.zdnet.com/downloads), one of the Web's premier download sites. "Little-known authors, kids in their basement, can now distribute their programs to millions of people for free. New kinds of programs are being written; entirely new business models are emerging." A perfect example of shareware's revolutionary potential is Napster (www.napster.com). Last year Shawn Fanning, a freshman at Northeastern University, decided he wanted to share music with other people over the Internet. He wrote a program to do it, posted it free on the Net. Today Napster is a company worth many millions of dollars, which threatens to transform the entire recording industry.
The proliferation of Web-distributed software leads many experts to discard the term "shareware" as obsolete. "I don't see any difference at all between trial software from Symantec and 'shareware' from a one-man show," says Gralla. Programmers whose sole income is from one program object to what they perceive as the term's negative associations. "The word shareware is misleading," says John Daleiden, creator of the EDGE Diagrammer (www.pacestar.com). "It implies 'Hey, I've done this incredibly simple thing that you can all use for free, and it would be really neat if you could give me a buck or two for it." Tom Davis agrees. "Zoot is just commercial software that you can order over the Net."
I beg, with all due respect, to differ. Zoot is not "just commercial software." There is a fundamental difference between Zoot and trial software from Symantec. And thank goodness. By its very nature, shareware can be more creative, efficient, and adaptable than mainstream programs. Corporate software is typically written for other corporations and must satisfy the conservative tastes of the business mass market. Programs written at large companies must also shoulder their fair share of high labor costs, marketing budgets, and overhead, or risk cancellation. The resulting software is normally solid, workmanlike, unsurprising, and safe, but has little room for innovation, no taste for risks. Take Agenda and Improv, two of the most brilliant programs I have ever used. Both were written by large teams at Lotus Development Corporation. Agenda was a free-form database with some similarities to Zoot, Improv a highly flexible spreadsheet application. They are, alas, no longer with us. Despite $40 million in gross sales, Agenda was discontinued for financial underperformance. Though widely applauded by software critics, Improv was cancelled for similar reasons and an alleged turf war with 1-2-3, the Lotus flagship spreadsheet.
Tom Davis has slightly different priorities. "As long as there is $5,000 in the bank, I don't worry about running out of money." He has no conflicts of interest or quarterly earnings reports to worry about. Like many other solo developers, what really interests him is perfecting his program. Sales of $40 million, so disappointing to Lotus, would keep Davis and an army of shareware authors happily churning out code. Shareware allows a level of software meritocracy than could never exist in the cautious world of corporate software.
Now that people can earn a respectable living with shareware, they are able to develop unique talents and indulge eccentricities as they never could in a corporate setting. When I asked Eric Isaacson (www.eji.com), author of several successful programs, how much he'd earned with shareware, there was a pause of about a nanosecond. "Total gross over 14 years is $1,274,507.39, after deducting bank fees for credit-card sales, but before deducting postage, printing, and other direct expenses." He has recently taken up walking. "I am trying to walk every street, alley, road, trail, railroad, shoreline, and thoroughfare in the 400-square-mile county in which I live." This is a systematic man. Not surprisingly, it shows in his code. When he set out to write one of his top sellers, the A86 assembler computer software program that translates human commands into the computer's machine language, Isaacson found that no existing software was precise enough for him. So he wrote everything from scratch: all elements of A86, as well as the BIOS (Basic Input-Output System), the operating system, the text editor, the works. Isaacson's penchant for efficient code has produced the finest assembler on earth for Intel computers.
It is hard to imagine Isaacson, the walker, the code fetishist, working on a big development team. Few enterprises are flexible enough to put up with Tom Davis' aversion to schedules, his love of loud guitar riffs, the jokes he scatters throughout his program, his insistence on naming it after a Frank Zappa song ("Zoot Allures"). There is a causal link here: A86 and Zoot are brilliant not in spite of, but because of, the idiosyncrasies of their creators. These people are a 21st-century version of the classical philosopher or medieval monk; their withdrawal from the world and absolute, laser-like focus on one shining ideal produces unique results, in program and programmer alike. "Scientists say that one's activities can actually change the physical structure of the brain," says Tom Davis, "and I'm convinced that seven years of continuous coding have changed the structure of my brain considerably. I find structured code very compelling to look at." This is much more than a job.
Shareware allows people with a remarkable range of skills and specializations to crystallize their experience in a program, then give me that distilled experience at a bargain price. I have an astronomy program called SkyMap (www.skymap.com), written by Chris Marriott, a distinguished British astrophysicist. My music notation program, Mozart (www.mozart.co.uk), was written by David Webber, a mathematician with a lifelong passion for the saxophone. My file manager, Singularity (www.winability.com), is by an honest-to-goodness rocket scientist named Andrei Belogortseff, who until 1992 was Senior Research Scientist at a top-secret nuclear missile laboratory in the Ukraine. When the USSR dissolved, Belogortseff moved to Chicago, where he now earns his living with shareware. Shareware is beginning to tap a huge, unrealized reservoir of talent.
The code these people write is predictably quite different from that produced at a major software house. Big companies typically divide a developing program into teams, each team responsible for a "module," a specific feature set. If one team changes its module, this can break every other team's module. "Many times, software development teams don't even work in the same building, much less the same room, so a project can become very cumbersome," says Paul Mayer, manager of Microsoft Network's Computing Central shareware forum (http://computingcentral.msn.com/topics/shareware/default.asp). "But get one good programmer to do the entire project, and his left hand always knows what his right hand is doing." Shareware is frequently cleaner and more efficient as a result.
Shareware can also be more adaptable in the future. "Authors tend to listen to the users and in many cases fix bugs and add features the same day the suggestions roll in," says Mayer. "You'd never see that in the corporate software world. Too many meetings and too many chiefs to appease before anything can be done." If you find a bug in Zoot, drop Tom Davis an e-mail; when you log on the next morning, nine times out of 10 you will find an update. Likewise with new features. The Net is a two-way street: Just as I can download programs, I can upload ideas, frequently through a direct line with the creator of the software. "If someone makes the effort to write me a message describing something they'd like Zoot to do," Davis says, "it must be pretty important to them, so I'm going to pay close attention." John Daleiden likewise implements several user ideas each month in his EDGE diagrammed. "In fact, it's uncanny majority of the suggestions I receive at any given time tend to focus on one area of the program, suggesting that users have a sense of what features are cutting edge in the industry right now." For the smart shareware author, customers are an invaluable source of fresh perspectives and concrete uses, an informal advisory board that guards against the tunnel vision that might otherwise afflict a lone programmer. Tom Davis's advisory board meets on "Zooters", a mailing list where users ask him questions, discuss with one another the unique ways they employ Zoot, suggest improvements. From this creative free-for-all, Davis frequently sees new directions for his program.
This symbiotic relationship between creator and customer is one of the most compelling benefits of the shareware model. An example of the creative energies released by such an intellectual swap meet is Linux, spearhead of the "open source" software movement. In 1991, the year after Microsoft released Windows 3.0, a 21-year-old Finn named Linus Torvalds wrote an entire operating system from scratch and put it on the Net for free, complete with its source code. (A normal program is sold in binary form, a string of 0s and 1s that runs fine on computers, but is incomprehensible to humans. Give programmers the source code, however, and they can understand its inner workings, to rewrite and improve it.) A handful, then hundreds, now thousands of programmers around the world have pitched in to improve Linux. The result is an operating system more stable and efficient than Windows, which runs for free on millions of computers.
Torvalds noticed that the more eclectic his circle of critic-collaborators became, the more deficiencies were weeded out of Linux. "The more strange uses there are," he has said, "the more likely it is to be a good system." Tom Davis harnesses these same diverse, corrective forces in Zooters. His correspondents are preachers and professors and private detectives, writers and economists, trial lawyers and film producers, all of whom use Zoot differently. People have described ways in which they've tackled problems in swimwear, corporate competitive intelligence, and lumbosacral spinal fusion. Everyone has a voice in how Zoot is to evolve in the future. Not long ago I asked Davis what new features he would include in the next major version of Zoot. "There will be a rich text editor and further database enhancements. For the other features, I may call a vote of the Zooters." Can you see that happening with Microsoft Word?
Not now, perhaps, but shareware may eventually change the way even the largest companies develop and support their software. Big enterprises have already bought into the shareware approach to sales and marketing; as Web use and general computer literacy continue to grow, customers spoiled by the responsiveness of shareware may come to expect similar treatment from mainstream applications. They will demand faster bug fixes, more regular and comprehensive updates. If big companies fail to deliver, shareware may be able to provide an alternative, just as Linux has already done for millions of users.
Right now, however, one thing is certain: shareware is a lot more fun. In the bubbly, collegial atmosphere of the Zooters mailing list, where jokes and playful aliases are the order of the day and the "share" in shareware is anything but rhetorical, Zoot has evolved more in the last year than any mainstream program I know has in five. As the great Dutch historian Jan Huizinga once observed, human play can in many ways be more creative, constructive, even serious, than conventional work. Huizinga's acute insight fits Zooters, and countless other watering holes scattered across the vast, trackless savanna of the Net, where people congregate to pursue common interests in the bright, clear light of unfeigned fascination. These are charmed worlds, where a liberating sense of semi-anonymity blends with a strong, surprisingly personal feeling of community. Here ideas flow freely, and work of a higher order can be done.
Tom Mueller is a novelist and freelance journalist based in Italy who secretly wishes he could write killer code. This article originally appeared in the September, 2000 issue of Hemispheres magazine.