Free For All - How Linux and the Free Software Movement Undercut the High Tech Titans
Peter Wayner (2002-12-22)

The list should also include the dozens of journalists at places like Slashdot.org, LinuxWorld, Linux magazine, Linux Weekly News, Kernel Traffic, Salon, and the New York Times. I should specifically mention the work of Joe Barr, Jeff Bates, Janelle Brown, Zack Brown, Jonathan Corbet, Elizabeth Coolbaugh, Amy Harmon, Andrew Leonard, Rob Malda, John Markoff, Mark Nielsen, Nicholas Petreley, Harald Radke, and Dave Whitinger. They wrote wonderful pieces that will make a great first draft of the history of the open source movement. Only a few of the pieces are cited directly in the footnotes, largely for practical reasons. The entire body of websites like Slashdot, Linux Journal, Linux World, Kernel Notes, or Linux Weekly News should be required reading for anyone interested in the free software movement.

There are hundreds of folks at Linux trade shows who took the time to show me their products, T-shirts, or, in one case, cooler filled with beer. Almost everyone I met at the conferences was happy to speak about their experiences with open source software. They were all a great source of information, and I don't even know most of their names.

See http://www.wayner.org/books/ffa/ for the FIRST PDF EDITION Page layout for this and the original paper edition designed by William Ruoto, see Not printed on acid-free paper. Library of Congress Cataloging-in-Publication Data Wayner, Peter, 1964 Free for all : how Linux and the free software movement undercut the high-tech titans / Peter Wayner. p. cm. ISBN 0-06-662050-3 1. Linux. 2. Operating systems (Computers) 3. Free computer software. I. Title. QA76.76.063 W394 2000 005.4'469 dc21 00-023919 00 01 02 03 04 V/RRD 10 9 8 7 6 5 4 3 2 1

The last competitor, though, was the most surprising to everyone. Schmalensee saw Linux, a program given away for free, as a big potential competitor. When he said Linux, he really meant an entire collection of programs known as “open source” software. These were written by a loose-knit group of programmers who shared all of the source code to the software over the Internet.

Schmalensee didn't mention that most people thought of Linux as a strange tool created and used by hackers in dark rooms lit by computer monitors. He didn't mention that many people had trouble getting Linux to work with their computers. He forgot to mention that Linux manuals came with subheads like “Disk Druid-like 'fstab editor' available.” He didn't delve into the fact that for many of the developers, Linux was just a hobby they dabbled with when there was nothing interesting on television. And he certainly didn't mention that most people thought the whole Linux project was the work of a mad genius and his weirdo disciples who still hadn't caught on to the fact that the Soviet Union had already failed big-time. The Linux folks actually thought sharing would make the world a better place. Fat-cat programmers who spent their stock-option riches on Porsches and balsamic vinegar laughed at moments like this.

Schmalensee didn't mention these facts. He just offered Linux as an alternative to Windows and said that computer manufacturers might switch to it at any time. Poof. Therefore, Microsoft had competitors. At the trial, the discourse quickly broke down into an argument over what is really a worthy competitor and what isn't. Were there enough applications available for Linux or the Mac? What qualifies as “enough”? Were these really worthy?

Under cross-examination, Schmalensee explained that he wasn't holding up the Mac, BeOS, or Linux as competitors who were going to take over 50 percent of the marketplace. He merely argued that their existence proved that the barriers produced by the so-called Microsoft monopoly weren't that strong. If rational people were investing in creating companies like BeOS, then Microsoft's power wasn't absolute.

Afterward, most people quickly made up their minds. Everyone had heard about the Macintosh and knew that back then conventional wisdom dictated that it would soon fail. But most people didn't know anything about BeOS or Linux. How could a company be a competitor if no one had heard of it? Apple and Microsoft had TV commercials. BeOS, at least, had a charismatic chairman. There was no Linux pitchman, no Linux jingle, and no Linux 30-second spot in major media. At the time, only the best-funded projects in the Linux community had enough money to buy spots on late-night community-access cable television. How could someone without money compete with a company that hired the Rolling Stones to pump excitement into a product launch?

When people heard that Microsoft was offering a free product as a worthy competitor, they began to laugh even louder at the company's chutzpah. Wasn't money the whole reason the country was having a trial? Weren't computer programmers in such demand that many companies couldn't hire as many as they needed, no matter how high the salary? How could Microsoft believe that anyone would buy the supposition that a bunch of pseudo-communist nerds living in their weird techno-utopia where all the software was free would ever come up with software that could compete with the richest company on earth? At first glance, it looked as if Microsoft's case was sinking so low that it had to resort to laughable strategies. It was as if General Motors were to tell the world “We shouldn't have to worry about fixing cars that pollute because a collective of hippies in Ithaca, New York, is refurbishing old bicycles and giving them away for free.” It was as if Exxon waved away the problems of sinking oil tankers by explaining that folksingers had written a really neat ballad for teaching birds and otters to lick themselves clean after an oil spill. If no one charged money for Linux, then it was probably because it wasn't worth buying.

But as everyone began looking a bit deeper, they began to see that Linux was being taken seriously in some parts of the world. Many web servers, it turned out, were already running on Linux or another free cousin known as FreeBSD. A free webserving tool known as Apache had controlled more than 50 percent of the web servers for some time, and it was gradually beating out Microsoft products that cost thousands of dollars. Many of the web servers ran Apache on top of a Linux or a FreeBSD machine and got the job done. The software worked well, and the nonexistent price made it easy to choose.

Linux was also winning over some of the world's most serious physicists, weapons designers, biologists, and hard-core scientists. Some of the nation's top labs had wired together clusters of cheap PCs and turned them into supercomputers that were highly competitive with the best machines on the market. One upstart company started offering “supercomputers” for $3,000. These machines used Linux to keep the data flowing while the racks of computers plugged and chugged their way for hours on complicated simulations.

There were other indications. Linux users bragged that their system rarely crashed. Some claimed to have machines that had been running for a year or more without a problem. Microsoft (and Apple) users, on the other hand, had grown used to frequent crashes. The “Blue Screen of Death” that appears on Windows users' monitors when something goes irretrievably wrong is the butt of many jokes.

Linux users also bragged about the quality of their desktop interface. Most of the uninitiated thought of Linux as a hacker's system built for nerds. Yet recently two very good operating shells called GNOME and KDE had taken hold. Both offered the user an environment that looked just like Windows but was better. Linux hackers started bragging that they were able to equip their girlfriends, mothers, and friends with Linux boxes without grief. Some people with little computer experience were adopting Linux with little trouble.

To the rest of the world, this urge to putter and fiddle with machines is more than a source of marital comedy. Cox is one of the great threats to the continued dominance of Microsoft, despite the fact that he found a way to weld spaghetti to a nonstick pan. He is one of the core developers who help maintain the Linux kernel. In other words, he's one of the group of programmers who helps guide the development of the Linux operating system, the one Richard Schmalensee feels is such a threat to Microsoft. Cox is one of the few people whom Linus Torvalds, the creator of Linux, trusts to make important decisions about future directions. Cox is an expert on the networking guts of the system and is responsible for making sure that most of the new ideas that people suggest for Linux are considered carefully and integrated correctly. Torvalds defers to Cox on many matters about how Linux-based computers talk with other computers over a network. Cox works long and hard to find efficient ways for Linux to juggle multiple connections without slowing down or deadlocking.

The group that works with Cox and Torvalds operates with no official structure. Millions of people use Linux to keep their computers running, and all of them have copies of the source code. In the 1980s, most companies began keeping the source code to their software as private as possible because they worried that a competitor might come along and steal the ideas the source spelled out. The source code, which is written in languages like C, Java, FORTRAN, BASIC, or Pascal, is meant to be read by programmers. Most companies didn't want other programmers understanding too much about the guts of their software. Information is power, and the companies instinctively played their cards close to their chests.

When Linus Torvalds first started writing Linux in 1991, however, he decided to give away the operating system for free. He included all the source code because he wanted others to read it, comment upon it, and perhaps improve it. His decision was as much a radical break from standard programming procedure as a practical decision. He was a poor student at the time, and this operating system was merely a hobby. If he had tried to sell it, he wouldn't have gotten anything for it. He certainly had no money to build a company that could polish the software and market it. So he just sent out copies over the Internet.

Today, about a thousand people regularly work with people like Alan Cox on the development of the Linux kernel, the official name for the part of the operating system that Torvalds started writing back in 1991. That may not be an accurate estimate because many people check in for a few weeks when a project requires their participation. Some follow everything, but most people are just interested in little corners. Many other programmers have contributed various pieces of software such as word processors or spreadsheets. All of these are bundled together into packages that are often called plain Linux or GNU/Linux and shipped by companies like Red Hat or more ad hoc groups like Debian. ¹ While Torvalds only wrote the core kernel, people use his name, Linux, to stand for a whole body of software written by thousands of others. It's not exactly fair, but most let it slide. If there hadn't been the Linux kernel, the users wouldn't have the ability to run software on a completely free system. The free software would need to interact with something from Microsoft, Apple, or IBM. Of course, if it weren't for all of the other free software from Berkeley, the GNU project, and thousands of other garages around the world, there would be little for the Linux kernel to do.

^1.Linux Weekly News keeps a complete list of distributors. These range from the small, one- or two-man operations to the biggest, most corporate ones like Red Hat: Alzza Linux, Apokalypse, Armed Linux, Bad Penguin Linux, Bastille Linux, Best Linux (Finnish/Swedish), Bifrost, Black Cat Linux (Ukrainian/Russian), Caldera OpenLinux, CCLinux, Chinese Linux Extension, Complete Linux, Conectiva Linux (Brazilian), Debian GNU/Linux, Definite Linux, DemoLinux, DLD, DLite, DLX, DragonLinux, easyLinux, Enoch, Eridani Star System, Eonova Linux, e-smith server and gateway, Eurielec Linux (Spanish), eXecutive Linux, floppyfw, Floppix, Green Frog Linux, hal91, Hard Hat Linux, Immunix, Independence, Jurix, Kha0s Linux, KRUD, KSI-Linux, Laetos, LEM, Linux Cyrillic Edition, LinuxGT, Linux-Kheops (French), Linux MLD (Japanese), LinuxOne OS, LinuxPPC, LinuxPPP (Mexican), Linux Pro Plus, Linux Router Project, LOAF, LSD, Mandrake, Mastodon, MicroLinux, MkLinux, muLinux, nanoLinux II, NoMad Linux, OpenClassroom, Peanut Linux, Plamo Linux, PLD, Project Ballantain, PROSA, QuadLinux, Red Hat, Rock Linux, RunOnCD, ShareTheNet, Skygate, Slackware, Small Linux, Stampede, Stataboware, Storm Linux, SuSE, Tomsrtbt, Trinux, TurboLinux, uClinux, Vine Linux, WinLinux 2000, Xdenu, XTeamLinux, and Yellow Dog Linux.

All of these people work at their own pace. Some work in their homes, like Alan Cox. Some work in university labs. Others work for businesses that use Linux and encourage their programmers to plug away so it serves their needs.

The team is united by mailing lists. The Linux Kernel mailing list hooks up Cox in Britain, Torvalds in Silicon Valley, and the others around the globe. They post notes to the list and discuss ideas. Sometimes verbal fights break out, and sometimes everyone agrees. Sometimes people light a candle by actually writing new code to make the kernel better, and other times they just curse the darkness.

Cox is now one of several people responsible for coordinating the addition of new code. He tests it for compatibility and guides Linux authors to make sure they're working together optimally. In essence, he tests every piece of incoming software to make sure all of the gauges work with the right system of measurement so there will be no glitches. He tries to remove the incompatibilities that marred Zorro.

Other features are not so popular, and they're tackled by the people who need the features. Some people want to hook their Linux boxes up to Macintoshes. Doing that smoothly can require some work in the kernel. Others may want to add special code to enable a special device like a high-speed camera or a strange type of disk drive. These groups often work on their own but coordinate their solutions with the main crowd. Ideally, they'll be able to come up with some patches that solve their problem without breaking some other part of the system.

Each day, Cox and his virtual colleagues pore through the lists trying to figure out how to make Linux better, faster, and more usable. Sometimes they skip out to watch a movie. Sometimes they go for hikes. But one thing they don't do is spend months huddled in conference rooms trying to come up with legal arguments. Until recently, the Linux folks didn't have money for lawyers, and that means they didn't get sidetracked by figuring out how to get big and powerful people like Richard Schmalensee to tell a court that there's no monopoly in the computer operating system business.

The battle between Linux and Microsoft is lining up to be the classic fight between the people like Schmalensee and the people like Cox. On one side are the armies of lawyers, lobbyists, salesmen, and expensive executives who are armed with patents, lawsuits, and legislation. They are skilled at moving the levers of power until the gears line up just right and billions of dollars pour into their pockets. They know how to schmooze, toady, beg, or even threaten until they wear the mantle of authority and command the piety and devotion of the world. People buy Microsoft because it's “the standard.” No one decreed this, but somehow it has come to be.

That's an idyllic picture, and the early success of Linux, FreeBSD, and other free packages makes it tempting to think that the success will build. Today, open source servers power more than 50 percent of the web servers on the Internet, and that is no small accomplishment. Getting thousands, if not millions, of programmers to work together is quite amazing given how quirky programmers can be. The ease of copying makes it possible to think that Alan Cox could get up late and still move the world.

Right now, the free software movement stands at a crucial moment in its history. In the past, a culture of giving and wide-open sharing let thousands of programmers build a great operating system that was, in many ways, better than anything coming from the best companies. Many folks began working on Linux, FreeBSD, and thousands of other projects as hobbies, but now they're waking up to find IBM, HewlettPackard, Apple, and all the other big boys pounding on their door. If the kids could create something as nice as Linux, everyone began to wonder whether these kids really had enough good stuff to go the distance and last nine innings against the greatest power hitters around.

Linus Torvalds may be on the cover of magazines, but he can't do anything with the wave of a hand. He must charm and cajole the thousands of folks on the Linux mailing list to make a change. Many of the free software projects may generate great code, but they have to beg for computers. The programmers might even surprise him and come up with an even better solution. They've done it in the past. But no money means that no one has to do what anyone says.

But shows that are charming and fresh in a barn can become thin and weak on a big stage on Broadway. The glitches and raw functionality of Linux and free software don't seem too bad if you know that they're built by kids in their spare time. Building real tools for real companies, moms, police stations, and serious users everywhere is another matter. Everyone may be hoping that sharing, caring, and curiosity are enough, but no one knows for certain. Maybe capital will end up winning. Maybe it won't. It's freedom versus assurance; it's wide-open sharing versus stock options; it's cooperation versus intimidation; it's the geeks versus the suits, all in one knockdown, hack-till-you-drop, winner-take-everything fight.

FreeBSD is a close cousin to the Linux kernel and one that predates it in some ways. It descends from a long tradition of research and development of operating systems at the University of California at Berkeley. The name BSD stands for “Berkeley Software Distribution,” the name given to one of the first releases of operating system source code that Berkeley made for the world. That small package grew, morphed, and absorbed many other contributions over the years.

Referring to Linux and FreeBSD as cousins is an apt term because they share much of the same source code in the same way that cousins share some of the same genes. Both borrow source code and ideas from each other. If you buy a disk with FreeBSD, which you can do from companies like Walnut Creek, you may get many of the same software packages that you get from a disk from Red Hat Linux. Both include, for instance, some of the GNU compilers that turn source code into something that can be understood by computers.

On that January 14, a new member of the WINE list was learning just how volunteering works. The guy posted a note to the list that described his Diamond RIO portable music device that lets you listen to MP3 files whenever you want. “I think the WINE development team should drop everything and work on getting this program to work as it doesn't seem like Diamond wants to release a Linux utility for the Rio,” he wrote.

The WINE clone of the Win32 is a fascinating example of how open source starts slowly and picks up steam. Bob Amstadt started the project in 1993, but soon turned it over to Alexandre Julliard, who has been the main force behind it. The project, although still far from finished, has produced some dramatic accomplishments, making it possible to run major programs like Microsoft Word or Microsoft Excel on a Linux box without using Windows. In essence, the WINE software is doing a good enough job acting like Windows that it's fooling Excel and Word. If you can trick the cousins, that's not too bad.

The WINE home page (www.winehq.com) estimates that more than 90,000 people use WINE regularly to run programs for Microsoft Windows without buying Windows. About 140 or more people regularly contribute to the project by writing code or fixing bugs. Many are hobbyists who want the thrill of getting their software to run without Windows, but some are corporate programmers. The corporate programmers want to sell their software to the broadest possible marketplace, but they don't want to take the time to rewrite everything. If they can get their software working well with WINE, then people who use Linux or BSD can use the software that was written for Microsoft Windows.

The new user who wanted to get his RIO player working with his Linux computer soon got a rude awakening. Andreas Mohr, a German programmer, wrote back,

Mohr's suggestion was to file a bug report that ranks the usability of the software so the programmers working on WINE can tweak it. This is just the first step in the free software experience. Someone has to notice the problem and fix it. In this case, someone needs to hook up their Diamond RIO MP3 player to a Linux box and try to move MP3 files with the software written for Windows. Ideally, the software will work perfectly, and now all Linux users will be able to use RIO players. In reality, there might be problems or glitches. Some of the graphics on the screen might be wrong. The software might not download anything at all. The first step is for someone to test the product and write up a detailed report about what works and what doesn't.

WINE can't pay anyone, and that means that great ideas sometimes get ignored. The free software community, however, doesn't necessarily see this as a limitation. If the RIO player were truly important, someone else would come along and pick up the project. Someone else would do the work and file a bug report so everyone could use the software. If there's no one else, then maybe the RIO software isn't that important to the Linux community. Work gets done when someone really cares enough to do it.

Still, he told me, “At the time Toy Story was coming out, there was a space shuttle flying with the Debian GNU/Linux distribution on it controlling a biological experiment. People would say 'Are you proud of working at Pixar?' and then I would say my hobby software was running on the space shuttle now. That was a turnaround point when I realized that Linux might become my career.”

In fact, it's a bad idea to see the free software revolution as having much to do with Microsoft. Even if Linux, FreeBSD, and other free software packages win, Microsoft will probably continue to fly along quite happily in much the same way that IBM continues to thrive even after losing the belt of the Heavyweight Computing Champion of the World to Microsoft. Anyone who spends his or her time focused on the image of a ragtag band of ruffians and orphans battling the Microsoft leviathan is bound to miss the real story.

Anyone who tunes in to the battle between Microsoft and the world expecting to see a good old-fashioned fight for marketplace domination is going to miss the real excitement. Sure, Linux, FreeBSD, OpenBSD, NetBSD, Mach, and the thousands of other free software projects are going to come out swinging. Microsoft is going to counterpunch with thousands of patents defended by armies of lawyers. Some of the programmers might even be a bit weird, and a few will be entitled to wear the adjective “ragtag.” But the real revolution has nothing to do with whether Bill Gates keeps his title as King of the Hill. It has nothing to do with whether the programmers stay up late and work in the nude. It has nothing to do with poor grooming, extravagant beards, Coke-bottle glasses, black trench coats, or any of the other stereotypes that fuel the media's image.

Meanwhile, on the other coast, the lawsuit tied up Berkeley and the BSD project for several years, and the project lost valuable energy and time by devoting them to the legal fight. In the meantime, several other completely free software projects started springing up around the globe. These began in basements and depended on machines that the programmer owned. One of these projects was started by Linus Torvalds and would eventually grow to become Linux, the unstoppable engine of hype and glory. He didn't have the money of the Berkeley computer science department, and he didn't have the latest machines that corporations gave them. But he had freedom and the pile of source code that came from unaffiliated, free projects like GNU that refused to compromise and cut intellectual corners. Although Torvalds might not have realized it at the time, freedom turned out to be most valuable of all.

One of the people who wanted UNIX was the Finnish student Linus Torvalds, who couldn't afford this tithe. He was far from the first one, and the conflict began long before he started to write Linux in 1991.

The first move to separate Berkeley's version of UNIX from AT&T's control wasn't really a revolution. No one was starting a civil war by firing shots at Fort Sumter or starting a revolution by dropping tea in the harbor. In fact, it started long before the lawsuit and Linux. In 1989, some people wanted to start hooking their PCs and other devices up to the Internet, and they didn't want to use UNIX.

While news traveled quickly to some corners, it didn't reach Finland. Network Release 2 came in June 1991, right around the same time that Linus Torvalds was poking around looking for a high-grade OS to use in experiments. Jolitz's 386BSD came out about six months later as Torvalds began to dig into creating the OS he would later call Linux. Soon afterward, Jolitz lost interest in the project and let it lie, but others came along. In fact, two groups called NetBSD and FreeBSD sprang up to carry the torch.

Any grown-up should take one look at this battle and understand just how the free software movement got so far. While the Berkeley folks were meeting with lawyers and worrying about whether the judges were going to choose the right side, Linus Torvalds was creating his own kernel. He started Linux on his own, and that made him a free man.

In June 1991, soon after Torvalds ³ started his little science project, the Computer Systems Research Group at Berkeley released what they thought was their completely unencumbered version of BSD UNIX known as Network Release 2. Several projects emerged to port this to the 386, and the project evolved to become the FreeBSD and NetBSD versions of today. Torvalds has often said that he might never have started Linux if he had known that he could just download a more complete OS from Berkeley.

^3.Everyone in the community, including many who don't know him, refers to him by his first name. The rules of style prevent me from using that in something as proper as a book.

The core of an OS is often called the “kernel,” which is one of the strange words floating around the world of computers. When people are being proper, they note that Linus Torvalds was creating the Linux kernel in 1991. Most of the other software, like the desktop, the utilities, the editors, the web browsers, the games, the compilers, and practically everything else, was written by other folks. If you measure this in disk space, more than 95 percent of the code in an average distribution lies outside the kernel. If you measure it by user interaction, most people using Linux or BSD don't even know that there's a kernel in there. The buttons they click, the websites they visit, and the printing they do are all controlled by other programs that do the work.

In 1991, Torvalds had a short list of features he wanted to add to the kernel. The Internet was still a small network linking universities and some advanced labs, and so networking was a small concern. He was only aiming at the 386, so he could rely on some of the special features that weren't available on other chips. High-end graphics hardware cards were still pretty expensive, so he concentrated on a text-only interface. He would later fix all of these problems with the help of the people on the Linux kernel mailing list, but for now he could avoid them.

When Torvalds started crafting the Linux kernel, he decided he was going to create a bigger, more integrated version that he called a “monolithic kernel.” This was something of a bold move because the academic community was entranced with what they called “microkernels.” The difference is partly semantic and partly real, but it can be summarized by analogy with businesses. Some companies try to build large, smoothly integr the steps of production. Others try to create smaller operations that subcontract much of the production work to other companies. One is big, monolithic, and all-encompassing, while the other is smaller, fragmented, and heterogeneous. It's not uncommon to find two companies in the same industry taking different approaches and thinking they're doing the right thing.

By the beginning of 1992, Linux was no longer a Finnish student's part-time hobby. Several influential programmers became interested in the code. It was free and relatively usable. It ran much of the GNU code, and that made it a neat, inexpensive way to experiment with some excellent tools. More and more people downloaded the system, and a significant fraction started reporting bugs and suggestions to Torvalds. He rolled them back in and the project snowballed.

This talent for organizing the work of others is a rare commodity, and Torvalds had a knack for it. He was gracious about sharing his system with the world and he never lorded it over anyone. His messages were filled with jokes and self-deprecating humor, most of which were carefully marked with smiley faces (:-)) to make sure that the message was clear. If he wrote something pointed, he would apologize for being a “hothead.” He was always gracious in giving credit to others and noted that much of Linux was just a clone of UNIX. All of this made him easy to read and thus influential.

His greatest trick, though, was his decision to avoid the mantle of power. He wrote in 1992, “Here's my standing on 'keeping control,' in 2 words (three?): I won't. The only control I've effectively been keeping on Linux is that I know it better than anybody else.”

He made it clear that people could vote to depose him at any time. “If people feel I do a bad job, they can do it themselves.” They could just take all of his Linux code and start their own version using Torvalds's work as a foundation.

Torvalds's burgeoning kernel dovetailed nicely with the tools that the GNU project created. All of the work by Stallman and his disciples could be easily ported to work with the operating system core that Torvalds was now calling Linux. This was the power of freely distributable source code. Anyone could make a connection, and someone invariably did. Soon, much of the GNU code began running on Linux. These tools made it easier to create more new programs, and the snowball began to roll.

This freedom also attracted others to the party. They knew that Linux would always be theirs, too. They could write neat features and plug them into the Linux kernel without worrying that Torvalds would yank the rug out from under them. The GPL was a contract that lasted long into the future. It was a promise that bound them together.

The Linux kernel also succeeded because it was written from the ground up for the PC platform. When the Berkeley UNIX hackers were porting BSD to the PC platform, they weren't able to make it fit perfectly. They were taking a piece of software crafted for older computers like the VAX, and shaving off corners and rewriting sections until it ran on the PC.

During the early months of Torvalds's work, the BSD group was stuck in a legal swamp. While the BSD team was involved with secret settlement talks and secret depositions, Linus Torvalds was happily writing code and sharing it with the world on the Net. His life wasn't all peaches and cream, but all of his hassles were open. Professor Andy Tanenbaum, a fairly well-respected and famous computer scientist, got in a long, extended debate with Torvalds over the structure of Linux. He looked down at Linux and claimed that Linux would have been worth two F's in his class because of its design. This led to a big flame war that was every bit as nasty as the fight between Berkeley and AT&T's USL. In fact, to the average observer it was even nastier. Torvalds returned Tanenbaum's fire with strong words like “fiasco,” “brain-damages,” and “suck.” He brushed off the bad grades by pointing out that Albert Einstein supposedly got bad grades in math and physics. The highpriced lawyers working for AT&T and Berkeley probably used very expensive and polite words to try and hide the shivs they were trying to stick in each other's back. Torvalds and Tanenbaum pulled out each other's virtual hair like a squawkfest on the Jerry Springer show.

But Torvalds's flame war with Tanenbaum occurred in the open in an Internet newsgroup. Other folks could read it, think about it, add their two cents' worth, and even take sides. It was a wide-open debate that uncovered many flaws in the original versions of Linux and Tanenbaum's Minix. They forced Torvalds to think deeply about what he wanted to do with Linux and consider its flaws. He had to listen to the arguments of a critic and a number of his peers on the Net and then come up with arguments as to why his Linux kernel didn't suck too badly.

The fight between Torvalds and Tanenbaum, however, drew people into the project. Other programmers like David Miller, Ted T'so, and Peter da Silva chimed in with their opinions. At the time, they were just interested bystanders. In time, they became part of the Linux brain trust. Soon they were contributing source code that ran on Linux. The argument's excitement forced them to look at Torvalds's toy OS and try to decide whether his defense made any sense. Today, David Miller is one of the biggest contributors to the Linux kernel. Many of the original debaters became major contributors to the foundations of Linux.

To this day, all of the devotees of the various BSDs grit their teeth when they hear about Linux. They think that FreeBSD, NetBSD, and OpenBSD are better, and they have good reasons for these beliefs. They know they were out the door first with a complete running system. But Linux is on the cover of the magazines. All of the great technically unwashed are now starting to use “Linux” as a synonym for free software. If AT&T never sued, the BSD teams would be the ones reaping the glory. They would be the ones to whom Microsoft turned when it needed a plausible competitor. They would be more famous.

McKusick says, “If you plot the installation base of Linux and BSD over the last five years, you'll see that they're both in exponential growth. But BSD's about eighteen to twenty months behind. That's about how long it took between Net Release 2 and the unencumbered 4.4BSD-Lite. That's about how long it took for the court system to do its job.”

Through the 1990s, the little toy operating system grew slowly and quietly as more and more programmers were drawn into the vortex. At the beginning, the OS wasn't rich with features. You could run several different programs at once, but you couldn't do much with the programs. The system's interface was just text. Still, this was often good enough for a few folks in labs around the world. Some just enjoyed playing with computers. Getting Linux running on their PC was a challenge, not unlike bolting an aftermarket supercharger onto a Honda Civic. But others took the project more seriously because they had serious jobs that couldn't be solved with a proprietary operating system that came from Microsoft or others.

In time, more people started using the system and started contributing their additions to the pot. Someone figured out how to make MIT's free X Window System run on Linux so everyone could have a graphical interface. Someone else discovered how to roll in technology for interfacing with the Internet. That made a big difference because everyone could hack, tweak, and fiddle with the code and then just upload the new versions to the Net.

It goes without saying that all the cool software coming out of Stallman's Free Software Foundation found its way to Linux. Some were simple toys like GNU Chess, but others were serious tools that were essential to the growth of the project. By 1991, the FSF was offering what might be argued were the best text editor and compiler in the world. Others might have been close, but Stallman's were free. These were crucial tools that made it possible for Linux to grow quickly from a tiny experimental kernel into a full-featured OS for doing everything a programmer might want to do.

James Lewis-Moss, one of the many programmers who devote some time to Linux, says that GCC made it possible for programmers to create, revise, and extend the kernel. “GCC is integral to the success of Linux,” he says, and points out that this may be one of the most important reasons why “it's polite to refer to it as GNU/Linux.”

Lewis-Moss points out one of the smoldering controversies in the world of free software: all of the tools and games that came from the GNU project started becoming part of what people simply thought of as plain “Linux.” The name for the small kernel of the operating system soon grew to apply to almost all the free software that ran with it. This angered Stallman, who first argued that a better name would be“Lignux.”When that failed to take hold, he moved to “GNU/Linux.” Some ignored his pleas and simply used “Linux,” which is still a bit unfair. Some feel that“GNU/Linux”is too much of a mouthful and, for better or worse, just plain Linux is an appropriate shortcut. Some, like Lewis-Moss, hold firm to GNU/Linux.

Soon some people were bundling together CD-ROMs with all this software in one batch. The group would try to work out as many glitches as possible so that the purchaser's life would be easier. All boasted strange names like Yggdrasil, Slackware, SuSE, Debian, or Red Hat. Many were just garage projects that never made much money, but that was okay. Making money wasn't really the point. People just wanted to play with the source. Plus, few thought that much money could be made. The GPL, for instance, made it difficult to differentiate the product because it required everyone to share their source code with the world. If Slackware came up with a neat fix that made their version of Linux better, then Debian and SuSE could grab it. The GPL prevented anyone from constraining the growth of Linux.

But only greedy businessmen see sharing and competition as negatives. In practice, the free flow of information enhanced the market for Linux by ensuring that it was stable and freely available. If one key CDROM developer gets a new girlfriend and stops spending enough time programming, another distribution will pick up the slack. If a hurricane flattened Raleigh, North Carolina, the home of Red Hat, then another supplier would still be around. A proprietary OS like Windows is like a set of manacles. An earthquake in Redmond, Washington, could cause a serious disruption for everyone.

The competition and the GPL meant that the users would never feel bound to one OS. If problems arose, anyone could always just start a splinter group and take Linux in that direction. And they did. All the major systems began as splinter groups, and some picked up enough steam and energy to dominate. In time, the best splinter groups spun off their own splinter groups and the process grew terribly complicated.

Hall remembers well the moment he discovered Linux. He told Linux Today,

I didn't even know I was involved with Linux at first. I got a copy of Dr. Dobb's Journal, and in there was an advertisement for “get a UNIX operating system, all the source code, and run it on your PC.” And I think it was $99. And I go, “Oh, wow, that's pretty cool. For $99, I can do that.” So I sent away for it, got the CD. The only trouble was that I didn't have a PC to run it on. So I put it on my Ultrix system, took a look at the main pages, directory structure and stuff, and said, “Hey, that looks pretty cool.” Then I put it away in the filing cabinet. That was probably around January of 1994.

At the meeting, Torvalds helped Hall and his boss set up a PC with Linux. This was the first time that Hall actually saw Linux run, and he was pleasantly surprised. He said, “By that time I had been using UNIX for probably about fifteen years. I had used System V, I had used Berkeley, and all sorts of stuff, and this really felt like UNIX. You know . . . I mean, it's kind of like playing the piano. You can play the piano, even if it's a crappy piano. But when it's a really good piano, your fingers just fly over the keys. That's the way this felt. It felt good, and I was really impressed.”

This experience turned Hall into a true convert and he went back to Digital convinced that the Linux project was more than just some kids playing with a toy OS. These so-called amateurs with no centralized system or corporate backing had produced a very, very impressive system that was almost as good as the big commercial systems. Hall was an instant devotee. Many involved in the project recall their day of conversion with the same strength. A bolt of lightning peeled the haze away from their eyes, and they saw.

Hall set out trying to get Torvalds to rewrite Linux so it would work well on the Alpha. This was not a simple task, but it was one that helped the operating system grow a bit more. The original version included some software that assumed the computer was designed like the Intel 386. This was fine when Linux only ran on Intel machines, but removing these assumptions made it possible for the software to run well on all types of machines.

Hall went sailing with Torvalds to talk about the guts of the Linux OS. Hall told me, "I took him out on the Mississippi River, went up and down the Mississippi in the river boat, drinking Hurricanes, and I said to him, 'Linus, did you ever think about porting Linux to a 64-bit processor, like the Alpha?' He said, 'Well, I thought about doing that, but the Helsinki office has been having problems getting me a system, so I guess I'll have to do the PowerPC instead.'

“I knew that was the wrong answer, so I came back to Digital (at the time), and got a friend of mine, named Bill Jackson, to send out a system to Linus, and he received it about a couple weeks after that. Then I found some people inside Digital who were also thinking about porting Linux to an Alpha. I got the two groups together, and after that, we started on the Alpha Linux project.”

Hall also helped start a group called Linux International, which works to make the corporate world safe for Linux. “We help vendors understand the Linux marketplace,” Hall told me. “There's a lot of confusion about what the GPL means. Less now, but still there's a lot of confusion. We helped them find the markets.”

Today, Linux International helps control the trademark on the name Linux and ensures that it is used in an open way. “When someone wanted to call themselves something like 'Linux University,' we said that's bad because there's going to be more than one. 'Linux University of North Carolina' is okay. It opens up the space.”

In the beginning, Torvalds depended heavily on the kindness of strangers like Hall. He didn't have much money, and the Linux project wasn't generating a huge salary for him. Of course, poverty also made it easier for people like Hall to justify giving him a machine. Torvalds wasn't rich monetarily, but he became rich in machines.

By 1994, when Hall met Torvalds, Linux was already far from just a one-man science project. The floppy disks and CD-ROMs holding a version of the OS were already on the market, and this distribution mechanism was one of the crucial unifying forces. Someone could just plunk down a few dollars and get a version that was more or less ready to run. Many simply downloaded their versions for free from the Internet.

In 1994, getting Linux to run was never really as simple as putting the CD-ROM in the drive and pressing a button. Many of the programs didn't work with certain video cards. Some modems didn't talk to Linux. Not all of the printers communicated correctly. Yet most of the software worked together on many standard machines. It often took a bit of tweaking, but most people could get the OS up and running on their computers.

This was a major advance for the Linux OS because most people could quickly install a new version without spending too much time downloading the new code or debugging it. Even programmers who understood exactly what was happening felt that installing a new version was a long, often painful slog through technical details. These CDROMs not only helped programmers, they also encouraged casual users to experiment with the system.

Other CD-ROM groups became more commercial. Debian sold its disks to pay for Internet connection fees and other expenses, but they were largely a garage operation. So were groups with names like Slackware, FreeBSD, and OpenBSD. Other groups like Red Hat actually set out to create a burgeoning business, and to a large extent, they succeeded. They took the money and used it to pay programmers who wrote more software to make Linux easier to use.

Slowly but surely, more and more people became aware of Linux, the GNU project, and its cousins like FreeBSD. No one was making much money off the stuff, but the word of mouth was spreading very quickly. The disks were priced reasonably, and people were curious. The GPL encouraged people to share. People began borrowing disks from their friends. Some companies even manufactured cheap rip-off copies of the CD-ROMs, an act that the GPL encouraged.

At the top of the pyramid was Linus Torvalds. Many Linux developers treated him like the king of all he surveyed, but he was like the monarchs who were denuded by a popular constitutional democracy. He had always focused on building a fast, stable kernel, and that was what he continued to do. The rest of the excitement, the packaging, the features, and the toys, were the dominion of the volunteers and contributors.

Torvalds moved to Silicon Valley and took a job with the very secret company Transmeta in order to help design the next generation of computer chips. He worked out a special deal with the company that allowed him to work on Linux in his spare time. He felt that working for one of the companies like Red Hat would give that one version of Linux a special imprimatur, and he wanted to avoid that. Plus, Transmeta was doing cool things.

In January 1999, the world caught up with the pioneers. Schmalensee mentioned Linux on the witness stand during the trial and served official notice to the world that Microsoft was worried about the growth of Linux. The system had been on the company's radar screen for some time. In October 1998, an internal memo from Microsoft describing the threat made its way to the press. Some thought it was just Microsoft's way of currying favor during the antitrust investigation. Others thought it was a serious treatment of a topic that was difficult for the company to understand.

The media followed Schmalensee's lead. Everyone wanted to know about Linux, GNU, open source software, and the magical effects of widespread, unconditional sharing. The questions came in tidal waves, and Torvalds tried to answer them again and again. Was he sorry he gave it all away? No. If he charged anything, no one would have bought his toy and no one would have contributed anything. Was he a communist? No, he was rather apolitical. Don't programmers have to eat? Yes, but they will make their money selling a service instead of getting rich off bad proprietary code. Was Linux going to overtake Microsoft? Yes, if he had his way. World Domination Soon became the motto.

But there were also difficult questions. How would the Linux world resist the embrace of big companies like IBM, Apple, Hewlett-Packard, and maybe even Microsoft? These were massive companies with paid programmers and schedules to meet. All the open source software was just as free to them as anyone else. Would these companies use their strength to monopolize Linux?

Many wanted to know when Linux would become easier to use for nonprogrammers. Programmers built the OS to be easy to take apart and put back together again. That's a great feature if you like hacking the inside of a kernel, but that doesn't excite the average computer user. How was the open source community going to get the programmers to donate their time to fix the mundane, everyday glitches that confused and infuriated the nonprogrammers? Was the Linux community going to be able to produce something that a nonprogrammer could even understand?

Others wondered if the Linux world could ever agree enough to create a software package with some coherence. Today, Microsoft users and programmers pull their hair out trying to keep Windows 95, Windows 98, and Windows NT straight. Little idiosyncrasies cause games to crash and programs to fail. Microsoft has hundreds of quality assurance engineers and thousands of support personnel. Still, the little details drive everyone crazy.

New versions of Linux appear as often as daily. People often create their own versions to solve particular problems. Many of these changes won't affect anyone, but they can add up. Is there enough consistency to make the tools easy enough to use?

Many wondered if Linux was right for world domination. Programmers might love playing with source code, but the rest of the world just wants something that delivers the e-mail on time. More important, the latter are willing to pay for this efficiency.

During this time, the relationship between AT&T and the universities was cordial. AT&T owned the commercial market for UNIX and Berkeley supplied many of the versions used in universities. While the universities got BSD for free, they still needed to negotiate a license with AT&T, and companies paid a fortune. This wasn't too much of a problem because universities are often terribly myopic. If they share their work with other universities and professors, they usually consider their sharing done. There may be folks out there without university appointments, but those folks are usually viewed as cranks who can be safely ignored. Occasionally, those cranks write their own OS that grows up to be Linux. The BSD version of freedom was still a far cry from Stallman's, but then Stallman hadn't articulated it yet. His manifesto was still a few years off.

Daniel is basically correct. The BSD code has evolved, or forked, into many different versions with names like FreeBSD, OpenBSD, and NetBSD while the Linux UNIX kernel released under Stallman's GPL is limited to one fairly coherent package. Still, there is plenty of crosspollination between the different versions of BSD UNIX. Both NetBSD 1.0 and FreeBSD 2.0, for instance, borrowed code from 4.4 BSD-Lite. Also, many versions of Linux come with tools and utilities that came from the BSD project.

But Daniel's point is also clouded with semantics. There are dozens if not hundreds of different Linux distributions available from different vendors. Many differ in subtle points, but some are markedly different. While these differences are often as great as the ones between the various flavors of BSD, the groups do not consider them psychologically separate. They haven't forked politically even though they've split off their code.

Sam Ockman, a Linux enthusiast and the founder of Penguin Computing, remembers the day of the meeting just before Netscape announced it was freeing its source code. “Eric Raymond came into town because of the Netscape thing. Netscape was going to free their software, so we drove down to Transmeta and had a meeting so we could advise Netscape,” he said.

The definition of what was open source grew out of the Debian project, one of the different groups that banded together to press CDROMs of stable Linux releases. Groups like these often get into debates about what software to include on the disks. Some wanted to be very pure and only include GPL'ed software. In a small way, that would force others to contribute back to the project because they wouldn't get their software distributed by the group unless it was GPL'ed. Others wanted less stringent requirements that might include quasi-commercial projects that still came with their source code. There were some cool projects out there that weren't protected by GPL, and it could be awfully hard to pass up the chance to integrate them into a package.

Stallman saw this secrecy as a great crime. Computer users should be able to share the source code so they can share ways to make it better. This trade should lead to more information-trading in a great feedback loop. Some folks even used the word “bloom” to describe the explosion of interest and cross-feedback. They're using the word the way biologists use it to describe the way algae can just burst into existence, overwhelming a region of the ocean. Clever insights, brilliant bug fixes, and wonderful new features just appear out of nowhere as human curiosity is amplified by human generosity in a grand explosion of intellectual synergy. The only thing missing from the picture is a bunch of furry Ewoks dancing around a campfire. ⁸

^8.Linux does have many marketing opportunities. Torvalds chose a penguin named Tux as the mascot, and several companies actually manufacture and sell stuffed penguins to the Linux realm. The BSD world has embraced a cute demon, a visual pun on the fact that BSD UNIX uses the word “daemon” to refer to some of the faceless background programs in the OS.

Raymond pointed out that the free source world can do a great job with these nasty bugs. He characterized this with the phrase, “Given enough eyeballs, all bugs are shallow,” which he characterized as “Linus's Law.” That is, eventually some programmer would start printing and using the Internet at the same time. After the system crashed a few times, some programmer would care enough about the problem to dig into the free source, poke around, and spot the problem. Eventually somebody would come along with the time and the energy and the commitment to diagnose the problem. Raymond named this “Linus's Law” after Linus Torvalds. Raymond is a great admirer of Torvalds and thinks that Torvalds's true genius was organizing an army to work on Linux. The coding itself was a distant second.

The comparison to software was simple. Corporations gathered the tithes, employed a central architect with a grand vision, managed the team of programmers, and shipped a product every once and a bit. The Linux world, however, let everyone touch the Source. People would try to fix things or add new features. The best solutions would be adopted by oth ers and the mediocre would fall by the wayside. Many different Linux versions would proliferate, but over time the marketplace of software would coalesce around the best standard version.

Linus Torvalds changed his mind by increasing the speed of sharing, which Raymond characterized as the rule of “release early and often, delegate everything you can, be open to the point of promiscuity.” Torvalds ran Linux as openly as possible, and this eventually attracted some good contributors. In the past, the FSF was much more careful about what it embraced and brought into the GNU project. Torvalds took many things into his distributions and they mutated as often as daily. Occasionally, new versions came out twice a day.

Raymond mixed this experience with his time watching Torvalds's team push the Linux kernel and used them as the basis for his essay on distributing the Source. “Mostly I was trying to pull some factors that I had observed as unconscious folklore so people could take them out and reason about them,” he said.

There is a good empirical reason for the faith in the strength of free source. After all, a group of folks who rarely saw each other had assembled a great pile of source code that was kicking Microsoft's butt in some corners of the computer world. Linux servers were common on the Internet and growing more common every day. The desktop was waiting to be conquered. They had done this without stock options, without corporate jets, without secret contracts, and without potentially illegal alliances with computer manufacturers. The success of the software from the GNU and Linux world was really quite impressive.

Part of this problem is the success of Raymond's metaphor. He said he just wanted to give the community some tools to understand the success of Linux and reason about it. But his two visions of a cathedral and a bazaar had such a clarity that people concentrated more on dividing the world into cathedrals and bazaars. In reality, there's a great deal of blending in between. The most efficient bazaars today are the suburban malls that have one management company building the site, leasing the stores, and creating a unified experience. Downtown shopping areas often failed because there was always one shop owner who could ruin an entire block by putting in a store that sold pornography. On the other side, religion has always been something of a bazaar. Martin Luther effectively split apart Christianity by introducing competition. Even within denominations, different parishes fight for the hearts and souls of people.

The same blurring holds true for the world of open source software. The Linux kernel, for instance, contains many thousands of lines of source code. Some put the number at 500,000. A few talented folks like Alan Cox or Linus Torvalds know all of it, but most are only familiar with the corners of it that they need to know. These folks, who may number in the thousands, are far outnumbered by the millions who use the Linux OS daily.

Second, no one really knows who reads the Linux source code for the opposite reason. The GNU/Linux source is widely available and frequently downloaded, but that doesn't mean it's read or studied. The Red Hat CDs come with one CD full of pre-compiled binaries and the second full of source code. Who knows whoever pops the second CDROM in their computer? Everyone is free to do so in the privacy of their own cubicle, so no records are kept.

If I were to bet, I would guess that the ratios of cognoscenti to uninformed users in the Linux and Microsoft worlds are pretty close. Reading the Source just takes too much time and too much effort for many in the Linux world to take advantage of the huge river of information available to them.

The average population, however, is aging quickly. As the software becomes better, it is easier for working stiffs to bring it into the corporate environments. Many folks brag about sneaking Linux into their office and replacing Microsoft on some hidden server. As more and more users find a way to make money with the free software, more and more older people (i.e., over 25) are able to devote some time to the revolution.

I suppose I would like to report that there's a healthy contingent of women taking part in the free source world, but I can't. It would be nice to isolate the free software community from the criticism that usually finds any group of men. By some definition or legal reasoning, these guys must be practicing some de facto discrimination. Somebody will probably try to sue someone someday. Still, the women are scarce and it's impossible to use many of the standard explanations. The software is, after all, free. It runs well on machines that are several generations old and available from corporate scrap heaps for several hundred dollars. Torvalds started writing Linux because he couldn't afford a real version of UNIX. Lack of money or the parsimony of evil, gender-nasty parents who refuse to buy their daughters a computer can hardly be blamed.

This may change in the future if organizations like LinuxChix (www.linuxchix.org) have their way. They run a site devoted to celebrating women who enjoy the open source world, and they've been trying to start up chapters around the world. The site gives members a chance to post their names and biographical details. Of course, several of the members are men and one is a man turning into a woman. The member writes, “I'm transsexual (male-to-female, pre-op), and at the moment still legally married to my wife, which means that if we stay together we'll eventually have a legal same-sex marriage.”

Racial politics, however, are more complicated. Much of the Linux community is spread out throughout the globe. While many members come from the United States, major contributors can be found in most countries. Linus Torvalds, of course, came from Finland, one of the more technically advanced countries in the world. Miguel de Icaza, the lead developer of the GNOME desktop, comes from Mexico, a country perceived as technically underdeveloped by many in the United States.

In general, the free source revolution is worldwide and rarely encumbered by racial and national barricades. Europe is just as filled with Linux developers as America, and the Third World is rapidly skipping over costly Microsoft and into inexpensive Linux. Interest in Linux is booming in China and India. English is, of course, the default language, but other languages continue to live thanks to automatic translation mechanisms like Babelfish.

When Linux began to take off, Torvalds moved to Silicon Valley and took a job with the supersecret research firm Transmeta. At Comdex in November 1999, Torvalds announced that Transmeta was working on a low-power computing chip with the nickname “Crusoe.”

There are, of course, some conspiracy theories. Transmeta is funded by a number of big investors including Microsoft cofounder Paul Allen. The fact that they chose to employ Torvalds may be part of a plan, some think, to distract him from Linux development. After all, version 2.2 of the kernel took longer than many expected, although it may have been because its goals were too ambitious. When Microsoft needed a coherent threat to offer up to the Department of Justice, Transmeta courteously made Torvalds available to the world. Few seriously believe this theory, but it is constantly whispered as a nervous joke.

This freedom also extended to programmers at work. In many companies, the computer managers are doctrinaire and officious. They often quickly develop knee-jerk reactions to technologies and use these stereotypes to make technical decisions. Free software like Linux was frequently rejected out of hand by the gatekeepers, who thought something must be wrong with the software if no one was charging for it. These attitudes couldn't stop the engineers who wanted to experiment with the free software, however, because it had no purchase order that needed approval.

The people on the side of BSD-style license, on the other hand, seem pragmatic, organized, and focused. There are three major free versions of BSD UNIX alone, and they're notable because they each have centrally administered collections of files. The GPL-protected Linux can be purchased from at least six major groups that bundle it together, and each of them includes packages and pieces of software they find all over the Net.

The BSD-license folks are also less cultish. The big poster boys, Torvalds and Stallman, are both GPL men. The free versions of BSD, which helped give Linux much of its foundation, are largely ignored by the press for all the wrong reasons. The BSD teams appear to be fragmented because they are all separate political organizations who have no formal ties. There are many contributors, which means that BSD has no major charismatic leader with a story as compelling as that of Linus Torvalds.

The Apache web server is protected by a BSD-style license that permits commercial reuse of the software without sharing the source code. It is a separate program, however, and many Linux users run the software on Linux boxes. Of course, this devotion to business and relatively quiet disposition isn't always true. Theo de Raadt, the leader of the OpenBSD faction, is fond of making bold proclamations. In his interview with me, he dismissed the Free Software Foundation as terribly misnamed because you weren't truly free to do whatever you wanted with the software.

Someone might point out that Alan Cox, one of the steadfast keepers of the GPL-protected Linux kernels, is not particularly flashy nor given to writing long manifestos on the Net. Others might say that Brian Behlendorf has been a great defender of the Apache project. He certainly hasn't avoided defending the BSD license, although not in the way that Stallman might have liked. He was, after all, one of the members of the Apache team who helped convince IBM that they could use the Apache web server without danger.

The three BSD projects are well known for keeping control of all the source code for all the software in the distribution. They're very centrally managed and brag about keeping all the source code together in one build tree. The Linux distributions, on the other hand, include software from many different sources. Some include the KDE desktop. Others choose GNOME. Many include both.

Some groups have become very effective marketing forces. Red Hat is a well-run company that has marketing teams selling people on upgrading their software as well as engineering teams with a job of writing improved code to include in future versions. Red Hat packages their distribution in boxes that are sold through normal sales channels like bookstores and catalogs. They have a big presence at trade shows like LinuxExpo, in part because they help organize them.

In many cases, there is no clear spectrum defined between order and anarchy. The groups just have their own brands of order. OpenBSD brags about stopping security leaks and going two years without a rootlevel intrusion, but some of its artwork is a bit scruffy. Red Hat, on the other hand, has been carefully working to make Linux easy for everyone to use, but they're not as focused on security details.

This disorder is changing a bit now that serious companies like Red Hat and VA Linux are entering the arena. These companies pay fulltime programmers to ensure that their products are bug free and easy to use. If their management does a good job, the open source software world may grow more ordered and actually anticipate more problems instead of waiting for the right person to come along with the time and the inclination to solve them.

Most people quickly become keenly aware of this competition. Each of the different teams creating distributions flags theirs as the best, the most up-to-date, the easiest to install, and the most plush. The licenses mean that each group is free to grab stuff from the other, and this ensures that no one builds an unstoppable lead like Microsoft did in the proprietary OS world. Sure, Red Hat has a large chunk of the mindshare and people think their brand name is synonymous with Linux, but anyone can grab their latest distribution and start making improvements on it. It takes little time at all.

But Stallman is right to distance himself from Soviet-style communism because there are few similarities. There's little central control in Stallman's empire. All Stallman can do to enforce the GNU General Public License is sue someone in court. He, like the Pope, has no great armies ready to keep people in line. None of the Linux companies have much power to force people to do anything. The GNU General Public License is like a vast disarmament treaty. Everyone is free to do what they want with the software, and there are no legal cudgels to stop them. The only way to violate the license is to publish the software and not release the source code.

Many people who approach the free software world for the first time see only communism. Bob Metcalfe, an entrepreneur, has proved himself several times over by starting companies like 3Com and inventing the Ethernet. Yet he looked at the free software world and condemned it with a derisive essay entitled “Linux's 60's technology, open-sores ideology won't beat W2K, but what will?”

The essay makes more confounding points equating Richard Stallman to Karl Marx for his writing and Linus Torvalds to Vladimir Lenin because of his aim to dominate the software world with his OS. For grins, he compares Eric Raymond to “Trotsky waiting for The People's ice pick” for no clear reason. Before this gets out of hand, he backpedals a bit and claims, “OK, communism is too harsh on Linux. Lenin too harsh on Torvalds [sic].”Then he sets off comparing the world of open source to the tree-hugging, back-to-the-earth movement.

“How about Linux as organic software grown in utopia by spiritualists?” he wonders. “If North America actually went back to the earth, close to 250 million people would die of starvation before you could say agribusiness. When they bring organic fruit to market, you pay extra for small apples with open sores--the Open Sores Movement.”

But numbers like this can't really capture the depth of the gift. Linus Torvalds always likes to say that he started writing Linux because he couldn't afford a decent OS for his machine so he could do some experiments. Who knows how many kids, grown-ups, and even retired people are hacking Linux now and doing some sophisticated computer science experiments because they can? How do we count this beneficence?

Free source code has none of these inefficiencies. Websites like Slashdot, Freshmeat, Linux Weekly News, LinuxWorld, KernelTraffic, and hundreds of other Linux or project-specific portals do a great job moving the software to the people who can use its value. People write the code and then other folks discover the value in it. Bad or unneeded code isn't foisted on anyone.

The comparison does offer some insight into life in the free software community. Some conventions like LinuxExpo and the hundreds of install-fests are sort of like parties. One company at a LinuxExpo was serving beer in its booth to attract attention. Of course, Netscape celebrated its decision to launch the Mozilla project with a big party. They then threw another one at the project's first birthday.

But the giving goes beyond the parties and the conferences. Giving great software packages creates social standing in much the same way that giving a lavish feast will establish you as a major member of the tribe. There is a sort of pecking order, and the coders of great systems like Perl or Linux are near the top. The folks at the top of the pyramid often have better luck calling on other programmers for help, making it possible for them to get their jobs done a little better. Many managers justify letting their employees contribute to the free software community because they build up a social network that they can tap to finish their official jobs.

The free source world, on the other hand, is a big free-for-all in both senses of the phrase. The code circulates for everyone to grab, and only those who need it dig in. There's no great connection between programmer and user. People grab software and take it without really knowing to whom they owe any debt. I only know a few of the big names who wrote the code running the Linux box on my desk, and I know that there are thousands of people who also contributed. It would be impossible for me to pay back any of these people because it's hard to keep them straight.

Of course, there's also a certain element of selfishness to the charity. The social prestige that comes from writing good free software is worth a fair amount in the job market. People like to list accomplishments like “wrote driver” or “contributed code to Linux Kernel 2.2” on their résumé. Giving to the right project is a badge of honor because serious folks doing serious work embraced the gift. That's often more valuable and more telling than a plaque or an award from a traditional boss.

Newberry is also a Linux fan. He reads the Kernel list but rarely contributes much to it. He runs various versions of Linux around the house, and none of them were working as well as he wanted with his Macintosh. So he poked around in the software, fixed it, and sent his code off to Alan Cox, who watches over the part of the kernel where his fixes belonged.

“I contributed some changes to the Appletalk stack that's in the Linux Kernel that make it easier for a Linux machine to offer dial-in services for Macintosh users,” he said in an article published in Salon. “As it stands, Mac users have always been able to dial into a Linux box and use IP protocols, but if they wanted to use Appletalk over PPP, the support wasn't really there.”

Of course, all of this justification and rationalization aren't the main reason why Newberry spends so much of his time hacking on Linux. Sure, it may help his company's bottom line. Sure, it might beef up his résumé by letting him brag that he got some code in the Linux kernel. But he also sees this as a bit of charity.

“I get a certain amount of satisfaction from the work . . . but I get a certain amount of satisfaction out of helping people. Improving Linux and especially its integration with Macs has been a pet project of mine for some time,” he says. Still, he sums up his real motivation by saying, “I write software because I just love doing it.” Perhaps we're just lucky that so many people love writing open source software and giving it away.

It's not hard to find bad stories about people who write good code. One person at a Linux conference told me, “The strange thing about Linus Torvalds is that he hasn't really offended everyone yet. All of the other leaders have managed to piss off someone at one time or another. It's hard to find someone who isn't hated by someone else.” While he meant it as a compliment for Torvalds, he sounded as if he wouldn't be surprised if Torvalds did a snotty, selfish, petulant thing. It would just be par for the course.

Occasionally, the fights get interesting. Eric Raymond and Bruce Perens are both great contributors to the open source movement. In fact, both worked together to try to define the meaning of the term. Perens worked with the community that creates the Debian distribution of Linux to come up with a definition of what was acceptable for the community. This definition morphed into a more official version used by the Open Source Initiative. When they got a definition they liked, they published it and tried to trademark the term “open source” in order to make sure it was applied with some consistency. It should be no surprise that all of that hard work brought them farther apart.

The free software world, of course, removes these barriers. If the Hotmail folks had joined the Linux team instead of Microsoft, they would be free to do whatever they wanted with their website even if it annoyed Linus Torvalds, Richard Stallman, and the pope. They wouldn't be rich, but there's always a price.

This love also has a more traditional effect on the hackers who create the free source code. They do it because they love what they're doing. Many of the people in the free source movement are motivated by writing great software, and they judge their success by the recognition they get from equally talented peers. A “nice job” from the right person--like Richard Stallman, Alan Cox, or Linus Torvalds--can be worth more than $100,000 for some folks. It's a strange way to keep score, but for most of the programmers in the free source world it's more of a challenge than money. Any schmoe in Silicon Valley can make a couple of million dollars, but only a few select folks can rewrite the network interface code of the Linux kernel to improve the throughput of the Apache server by 20 percent.

And of course there are thousands of free software projects that are going to get left behind hanging out at the same old pizza joint. There were always going to be thousands left behind. People get excited about new projects, better protocols, and neater code all the time. The old code just sort of withers away. Occasionally someone rediscovers it, but it is usually just forgotten and superseded. But this natural evolution wasn't painful until the successful projects started ending up on the covers of magazines and generating million-dollar deals with venture capitalists. People will always be wondering why their project isn't as big as Linux.

This split is already growing. Red Hat software employs some of the major Linux contributors like Alan Cox. They get a salary while the rest of the contributors get nothing. Sun, Apple, and IBM employees get salaries, but folks who work on Apache or the open versions of BSD get nothing but the opportunity to hack cool code.

Jeff Bates, an editor at Slashdot, says that Mozilla may have suffered because Netscape was so successful. The Netscape browser was already available for free for Linux. “There wasn't a big itch to scratch,” he says. “We already had Netscape, which was fine for most people. This project interested a smaller group than if we'd not had Netscape-hence why it didn't get as much attention.”

In most cases, the flow is not particularly novel. The companies just choose FreeBSD or some version of Linux for their machines like any normal human being. Many web companies use a free OS like Linux or FreeBSD because they're both cheap and reliable. This is going to grow much more common as companies realize they can save a substantial amount of money over buying seat licenses from companies like Microsoft.

What happens if a bug emerges in some version of the Linux kernel and it makes it into several distributions? It's not really the fault of the distribution creators, because they were just shipping the latest version of the kernel. And it's not really the kernel creators' fault, because they weren't marketing the kernel as ready for everyone to run. They were just floating some cool software on the Net for free. Who's responsible for the bug? Who gets sued?

The free OS also puts Intel's lion's share up for grabs. Linux runs well on Intel chips, but it also runs on chips made by IBM, Motorola, Compaq, and many others. The NetBSD team loves to brag that its software runs on almost all platforms available and is dedicated to porting it to as many as possible. Someone using Linux or NetBSD doesn't care who made the chip inside because the OS behaves similarly on all of them.

This threat shows that the emergence of the free OSs ensures that hardware companies will also face increased competitive pressure. Sure, they may be able to get Microsoft off their back, but Linux may make things a bit worse.

He has a point. Linux is a lot of fun to play with and it is now a very stable OS, but it took a fair number of years to get to this point. Many folks in the free source world like to say things like, “It used to be that the most fun in Linux was just getting it to work.” Companies like Morgan Stanley, Schwab, American Airlines, and most others live and die on the quality of their computer systems. They're quite willing to pay money if it helps ensure that things don't go wrong.

Red Hat has managed to sell enough CD-ROM disks to fund the development of new projects. They've created a good selection of installation tools that make it relatively easy for people to use Linux. They also help pay salaries for people like Alan Cox who contribute a great deal to the evolution of the kernel. They do all of this while others are free to copy their distribution disks verbatim.

McVoy doesn't argue with these facts, but feels that they're just a temporary occurrence. The huge growth of interest in Linux means that many new folks are exploring the operating system. There's a great demand for the hand-holding and packaging that Red Hat offers. In time, though, everyone will figure out how to use the product and the revenue stream should disappear as competition drives out the ability to charge $50 for each disk.

CoSource says that it will try to put together the bounties of many small groups and allow people to pay them with credit cards. It uses the example of a group of Linux developers who would gather together to fund the creation of an open source version of their favorite game. They would each chip in $10, $20, or $50 and when the pot got big enough, someone would step forward. Creating a cohesive political group that could effectively offer a large bounty is a great job for these sites.

On the other hand, forking can hurt the community by duplicating efforts, splitting alliances, and sowing confusion in the minds of users. If Bob starts writing and publishing his own version of Linux out of his house, then he's taking some energy away from the main version. People start wondering if the version they're running is the Missouri Synod version of Emacs or the Christian Baptist version. Where do they send bug fixes? Who's in charge? Distribution groups like Debian or Red Hat have to spend a few moments trying to decide whether they want to include one version or the other. If they include both, they have to choose one as the default. Sometimes they just throw up their hands and forget about both. It's a civil war, and those are always worse than a plain old war.

Of course, good software can have anti-forking effects. Linus Torvalds said in one interview, “Actually, I have never even checked 386BSD out; when I started on Linux it wasn't available (although Bill Jolitz's series on it in Dr. Dobbs Journal had started and were interesting), and when 386BSD finally came out, Linux was already in a state where it was so usable that I never really thought about switching. If 386BSD had been available when I started on Linux, Linux would probably never have happened.” So if 386BSD had been easier to find on the Net and better supported, Linux might never have begun.

While the three forks of BSD may cooperate more than they compete, the Linux world still likes to look at the BSD world with a bit of contempt. All of the forks look somewhat messy, even if having the freedom to fork is what Stallman and GNU are ostensibly fighting to achieve. The Linux enthusiasts seem to think, “We've got our ducks in a single row. What's your problem?” It's sort of like the Army mentality. If it's green, uniform, and the same everywhere, then it must be good.

The BSD lacks the monomaniacal cohesion of Linux, and this seems to hurt their image. The BSD community has always felt that Linux is stealing the limelight that should be shared at least equally between the groups. Linux is really built around a cult of Linus Torvalds, and that makes great press. It's very easy for the press to take photos of one man and put him on the cover of a magazine. It's simple, clean, neat, and perfectly amenable to a 30-second sound bite. Explaining that there's FreeBSD, NetBSD, OpenBSD, and who knows what smaller versions waiting in the wings just isn't as manageable.

Eric Raymond, a true disciple of Linus Torvalds and Linux, sees it in technical terms. The BSD community is proud of the fact that each distribution is built out of one big source tree. They get all the source code for all the parts of the kernel, the utilities, the editors, and whatnot together in one place. Then they push the compile button and let people work. This is a crisp, effective, well-managed approach to the project.

The Linux groups, however, are not that coordinated at all. Torvalds only really worries about the kernel, which is his baby. Someone else worries about GCC. Everyone comes up with their own source trees for the parts. The distribution companies like Red Hat worry about gluing the mess together. It's not unusual to find version 2.0 of the kernel in one distribution while another is sporting version 2.2.

“In BSD, you can do a unified make. They're fairly proud of that,” says Raymond. “But this creates rigidities that give people incentives to fork. The BSD things that are built that way develop new spin-off groups each week, while Linux, which is more loosely coupled, doesn't fork.”

But this distinction may be semantic. Forking does occur in the Linux realm, but it happens as small diversions that get explained away with other words. Red Hat may choose to use GNOME, while another distribution like SuSE might choose KDE. The users will see a big difference because both tools create virtual desktop environments. You can't miss them. But people won't label this a fork. Both distributions are using the same Linux kernel and no one has gone off and said, “To hell with Linus, I'm going to build my own version of Linux.” Everyone's technically still calling themselves Linux, even if they're building something that looks fairly different on the surface.

Jason Wright, one of the developers on the OpenBSD team, sees the organization as a good thing. “The one thing that all of the BSDs have over Linux is a unified source tree. We don't have Joe Blow's tree or Bob's tree,” he says. In other words, when they fork, they do it officially, with great ceremony, and make sure the world knows of their separate creations. They make a clear break, and this makes it easier for developers.

Wright says that this single source tree made it much easier for them to turn OpenBSD into a very secure OS.“We've got the security over Linux. They've recently been doing a security audit for Linux, but they're going to have a lot more trouble. There's not one place to go for the source code.”

To extend this to political terms, the Linux world is like the 1980s when Ronald Reagan ran the Republican party with the maxim that no one should ever criticize another Republican. Sure, people argued internally about taxes, abortion, crime, and the usual controversies, but they displayed a rare public cohesion. No one criticizes Torvalds, and everyone is careful to pay lip service to the importance of Linux cohesion even as they're essentially forking by choosing different packages.

John Gilmore, one of the founders of the free software company Cygnus and a firm believer in the advantages of the GNU General Public License, says, “In Linux, each package has a maintainer, and patches from all distributions go back through that maintainer. There is a sense of cohesion. People at each distribution work to reduce their differences from the version released by the maintainer. In the BSD world, each tree thinks they own each program--they don't send changes back to a central place because that violates the ego model.”

Jordan Hubbard, the leader of FreeBSD, is critical of Raymond's characterization of the BSD world. “I've always had a special place in my heart for that paper because he painted positions that didn't exist,” Hubbard said of Raymond's piece “The Cathedral and the Bazaar.” "You could point to just the Linux community and decide which part was cathedral-oriented and which part was bazaar-oriented.

When it comes right down to it, there's even plenty of forking going on about the definition of a fork. When some of the Linux team point at the BSD world and start making fun about the forks, the BSD team gets defensive. The BSD guys always get defensive because their founder isn't on the cover of all the magazines. The Linux team hints that maybe, if they weren't forking, they would have someone with a name in lights, too.

Hubbard is right. Linux forks just as much, they just call it a distribution or an experimental kernel or a patch kit. No one has the chutzpah to spin off their own rival political organization. No one has the political clout.

The most prevalent form of government in these communities is the benign dictatorship. Richard Stallman wrote some of the most important code in the GNU pantheon, and he continues to write new code and help maintain the old software. The world of the Linux kernel is dominated by Linus Torvalds. The original founders always seem to hold a strong sway over the group. Most of the code in the Linux kernel is written by others and checked out by a tight circle of friends, but Torvalds still has the final word on many changes.

The two of them are, of course, benign dictators, and the two of them don't really have any other choice. Both have a seemingly absolute amount of power, but this power is based on a mixture of personal affection and technical respect. There are no legal bounds that keep all of the developers in line. There are no rules about intellectual property or non-disclosure. Anyone can grab all of the Linux kernel or GNU source code, run off, and start making whatever changes they want. They could rename it FU, Bobux, Fredux, or Meganux and no one could stop them. The old threats of lawyers, guns, and money aren't anywhere to be seen.

The Debian group has a wonderful pedigree and many praise it as the purest version of Linux around, but it began as a bunch of outlaws who cried mutiny and tossed Richard Stallman overboard. Well, it wasn't really so dramatic. In fact, “mutiny” isn't really the right word when everyone is free to use the source code however they want.

This army is a diverse bunch. At a recent Linux conference, Jeff Bates, one of the editors of the influential website Slashdot (www.slashdot.org), pointed me toward the Debian booth, which was next to theirs. “If you look in the booth, you can see that map. They put a pushpin in the board for every developer and project leader they have around the world. China, Netherlands, Somalia, there are people coming from all over.”

Lewis-Moss's job isn't exactly programming, but it's close. He has to download the source code, compile the program, run it, and make sure that the latest version of the source works correctly with the latest version of the Linux kernel and the other parts of the OS that keep a system running. The packager must also ensure that the program works well with the Debian-specific tools that make installation easier. If there are obvious bugs, he'll fix them himself. Otherwise, he'll work with the author on tracking down and fixing the problems.

The Linux development effort moves slowly forward with thousands of stories like Lewis-Moss's. Folks come along, check out the code, and toss in a few contributions that make it a bit better for themselves. The mailing list debates some of the changes if they're controversial or if they'll affect many people. It's a very efficient system in many ways, if you can stand the heat of the debates.

While the mailing list looks like an idealized notion of a congress for the Linux kernel development, it is not as perfect as it may seem. Not all comments are taken equally because friendships and political alliances have evolved through time. The Debian group elected a president to make crucial decisions that can't be made by deep argument and consensus. The president doesn't have many other powers in other cases.

While the Linux and GNU worlds are dominated by their one great Sun King, many other open source projects have adopted a more modern government structure that is more like Debian. The groups are still fairly ad hoc and unofficial, but they are more democratic. There's less idolatry and less dependence on one person.

This seriousness and corporatization are probably the only possible steps that the Apache group could take. They've always been devoted to advancing the members' interests. Many of the other open source projects like Linux were hobbies that became serious. The Apache project was always filled with people who were in the business of building the web. While some might miss the small-town kind of feel of the early years, the corporate structure is bringing more certainty and predictability to the realm. The people don't have to wear suits now that it's a corporation. It just ensures that tough decisions will be made at a predictable pace.

The next induction ceremony for this pantheon should include Robert Young, the CEO of Red Hat Software, who helped the Linux and the open source world immeasurably by finding a way to charge people for something they could get for free. This discovery made the man rich, which isn't exactly what the free software world is supposed to do. But his company also contributed a sense of stability and certainty to the Linux marketplace, and that was sorely needed. Many hard-core programmers, who know enough to get all of the software for free, are willing to pay $70 to Red Hat just because it is easier. While some may be forever jealous of the millions of dollars in Young's pocket, everyone should realize that bringing Linux to a larger world of computer illiterates requires good packaging and hand-holding. Free software wouldn't be anywhere if someone couldn't find a good way to charge for it.

The best way to understand why Young ranks with the folks who discovered how to sell sugar water is to go to a conference like LinuxExpo. In the center of the floor is the booth manned by Red Hat Software, the company Young started in Raleigh, North Carolina, after he got through working in the computer-leasing business. Young is in his fifties now and manages to survive despite the fact that most of his company's devotees are much closer to 13. Red Hat bundles together some of the free software made by the community and distributed over the Net and puts it on one relatively easy-to-use CD-ROM. Anyone who wants to install Linux or some of its packages can simply buy a disk from Red Hat and push a bunch of keys. All of the information is on one CD-ROM, and it's relatively tested and pretty much ready to go. If things go wrong, Red Hat promises to answer questions by e-mail or telephone to help people get the product working.

To make matters worse for Red Hat, the potential competitors don't have to go out onto the Net and reassemble the collection of software for themselves. The GPL specifically forbids people from placing limitations on redistributing the source code. That means that a potential competitor doesn't have to do much more than buy a copy of Red Hat's disk and send it off to the CD-ROM pressing plant. People do this all the time. One company at the exposition was selling copies of all the major Linux distributions like Red Hat, Slackware, and OpenBSD for about $3 per disk. If you bought in bulk, you could get 11 disks for $25. Not a bad deal if you're a consumer.

Red Hat also added a custom installation utility to make life easier for people who want to add Red Hat to their computer. ¹² They could have made this package installation tool proprietary. After all, Red Hat programmers wrote the tool on company time. But Young released it with the GNU General Public License, recognizing that the political value of giving something back was worth much more than the price they could charge for the tool.

^12.Er, I mean to say “add Linux” or “add GNU/Linux.” “Red Hat” is now one of the synonyms for free software.

This is part of a deliberate political strategy to build goodwill among the programmers who distribute their software. Many Linux users compare the different companies putting together free source software CDROMs and test their commitment to the free software ideals. Debian, for instance, is very popular because it is a largely volunteer project that is careful to only include certified free source software on their CD-ROMs. Debian, however, isn't run like a business and it doesn't have the same attitude. This volunteer effort and enlightened pursuit of the essence of free software make the Debian distribution popular among the purists.

Distributors like Caldera, on the other hand, include nonfree software with their disk. You pay $29.95 to $149.95 for a CD-ROM and get some nonfree software like a word processor tossed in as a bonus. This is a great deal if you're only going to install the software once, but the copyright on the nonfree software prevents you from distributing the CD-ROM to friends. Caldera is hoping that the extras it throws in will steer people toward its disk and get them to choose Caldera's version of Linux. Many of the purists, like Richard Stallman, hate this practice and think it is just a not very subtle way to privatize the free software. If the average user isn't free to redistribute all the code, then there's something evil afoot. Of course, Stallman or any of the other software authors can't do anything about this because they made their software freely distributable.

Several companies are already making PCs with Linux software installed at the factory. While they could simply download the software from the Net themselves and create their own package, several have chosen to bundle Red Hat's version with their machines. Sam Ockman, the president of Penguin Computing, runs one of those companies.

Ockman is a recent Stanford graduate in his early twenties and a strong devotee of the Linux and GPL world. He says he started his company to prove that Linux could deliver solid, dependable servers that could compete with the best that Sun and Microsoft have to offer.

Ockman has mixed feelings about life at Stanford. While he fondly remembers the “golf course-like campus,” he says the classes were too easy. He graduated with two majors despite spending plenty of time playing around with the Linux kernel. He says that the computer science department's hobbled curriculum drove him to Linux. “Their whole CS community is using a stupid compiler for C on the Macintosh,” he says.“Why don't they start you off on Linux? By the time you get to [course] 248, you could hack on the Linux kernel or your own replacement kernel. It's just a tragedy that you're sitting there writing virtual kernels on a Sun system that you're not allowed to reboot.”

When Ockman had to choose a version of Linux for his Penguin computers, he chose Red Hat. Bob Young's company made the sale because it was playing by the rules of the game and giving software back with a GPL. Ockman says, “We actually buy the box set for every single one. Partially because the customers like to get the books, but also to support Red Hat. That's also why we picked Red Hat. They're the most free of all of the distributions.”

Debian, Ockman concedes, is also very free and politically interesting, but says that his company is too small to support multiple distributions. “We only do Red Hat. That was a very strategic decision on our part. All of the distributions are pretty much the same, but there are slight differences in this and that. We could have a twelve-person Debian group, but it would just be a nightmare for us to support all of these different versions of Linux.”

At the LinuxExpo, Red Hat was selling T-shirts, too. One slick number retailing for $19 just said “The Revolution of Choice” in Red Hat's signature old typewriter font. Others for sale at the company's site routinely run for $15 or more. They sucked me in. When I ordered my first Red Hat disk from them, I bought an extra T-shirt to go with the mix.

Many of the other groups are part of the game. The OpenBSD project sold out of their very fashionable T-shirts with wireframe versions of its little daemon logo soon after the beginning of the LinuxExpo. They continue to sell more T-shirts from their website. Users can also buy CD-ROMs from OpenBSD.

The most expensive T-shirt at the show came with a logo that imitated one of the early marketing images of the first Star Wars movie. The shirt showed Torvalds and Stallman instead of Han Solo and Luke Skywalker under a banner headline of “OS Wars.” The shirt cost only $100, but “came with free admission to the upcoming Linux convention in Atlanta.”

Ockman looks at this market competition for T-shirts and sees a genius. He says, "I think Bob Young's absolutely brilliant. Suddenly he realized that there's no future in releasing mainframes. He made a jump after finding college kids in Carolina [using Linux]. For him to make that jump is just amazing. He's a marketing guy. He sat down and figured it out.

Young's plan to brand the OS with a veneer of cool produced more success than anyone could imagine. Red Hat is by far the market leader in providing Linux to the masses, despite the fact that many can and do “steal” a low-cost version. Of course, “steal” isn't the right word, because Red Hat did the same thing. “Borrow” isn't right, “grab” is a bit casual, and “join in everlasting communion with the great free software continuum” is just too enthusiastic to be cool.

The GPL is a powerful force that prevents Red Hat from making many unilateral decisions. There are plenty of distributions that would like to take over the mantle of the most popular version of Linux. It's not hard. The source code is all there.

There are parts of this conspiracy theory that are already true. Red Hat does dominate the United States market for Linux and it controls a great deal of the mindshare. Their careful growth supported by an influx of cash ensured a strong position in the marketplace.

Can they squeeze their partners by charging different rates for Linux? Microsoft is known to offer lower Windows prices to their friends. This is unlikely. Anyone can just buy a single Red Hat CDROM from a duplicator like CheapBytes. This power play won't work.

Can they force people to pay a “Red Hat tax” just to upgrade to the latest software? Not likely. Red Hat is going to be a service company, and they're going to compete on having the best service for their customers. Their real competitor will be companies that sell support contracts like LinuxCare. Service industries are hard work. Every customer needs perfect care or they'll go somewhere else next time. Red Hat's honeymoon with the IPO cash will only last so long. Eventually, they're going to have to earn the money to get a return on the investment. They're going to be answering a lot of phone calls and e-mails.

Still, most users including the best programmers end up paying a company like Red Hat, Caldera, or a group like OpenBSD to do some of the basic research in building a Linux system. All of the distribution companies charge for a copy of their software and throw in some support. While the software is technically free, you pay for help to get it to work.

Of course, the cost of this is debatable. Tivo, for instance, is a company that makes a set-top box for recording television content on an internal hard disk. The average user just sees a fancy, easy-to-use front end, but underneath, the entire system runs on the Linux operating system. Tivo released a copy of the stripped-down version of Linux it ships on its machines on its website, fulfilling its obligation to the GNU GPL. The only problem I've discovered is that the web page (www.tivo.com/linux/) is not particularly easy to find from the home page. If I hadn't known it was there, I wouldn't have found it.

Free source folks are just as free to share ideas. Many of the rival Linux and BSD distributions often borrow code from each other. While they compete for the hearts and minds of buyers, they're forced by the free source rules to share the code. If someone writes one device driver for one platform, it is quickly modified for another.

The Linux movement isn't really about nations and it's not really about war in the old-fashioned sense. It's about nerds building software and letting other nerds see how cool their code is. It's about empowering the world of programmers and cutting out the corporate suits. It's about spending all night coding on wonderful, magnificent software with massive colonnades, endless plazas, big brass bells, and huge steam whistles without asking a boss “Mother, may I?” It's very individualistic and peaceful.

That stirring romantic vision may be moving the boys in the trenches, but the side effects are beginning to be felt in the world of global politics. Every time Linux, FreeBSD, or OpenBSD is installed, several dollars don't go flowing to Seattle. There's a little bit less available for the Microsoft crowd to spend on mega-mansions, SUVs, and local taxes. The local library, the local police force, and the local schools are going to have a bit less local wealth to tax. In essence, the Linux boys are sacking Seattle without getting out of their chairs or breaking a sweat. You won't see this battle retold on those cable channels that traffic in war documentaries, but it's unfolding as we speak.

The difference in treatment probably did not result from any secret love for Linux or OpenBSD lurking in the hearts of the regulators in the Bureau of Export Affairs at the Department of Commerce. The regulators are probably more afraid of losing a lawsuit brought by Daniel Bernstein. In the latest decision released in May 1999, two out of three judges on an appeals panel concluded that the U.S. government's encryption regulations violated Bernstein's rights of free speech. The government argued that source code is a device not speech. The case is currently being appealed. The new regulations seem targeted to specifically address the problems the court found with the current regulations.

Most folks in the free source world may not have big bank accounts. Those are just numbers in a computer anyway, and everyone who can program knows how easy it is to fill a computer with numbers. But the free source world has good software and the source code that goes along with it. How many times a day must Bill Gates look at the blue screen of death that splashes across a Windows computer monitor when the Windows software crashes? How many times does Torvalds watch Linux crash? Who's better off? Who's wealthier?

There's no question that people like Stallman love life with source code. A deeper question is whether the free source realm offers a wealthier lifestyle for the average computer user. Most people aren't programmers, and most programmers aren't even the hard-core hackers who love to fiddle with the UNIX kernel. I've rarely used the source code to Linux, Emacs, or any of the neat tools on the Net, and many times I've simply recompiled the source code without looking at it. Is this community still a better deal?

Some grouse that comparing features like this isn't fair to the Mac or Windows world. The GNOME toolkit, they point out, didn't come out of years of research and development. The start button and the toolbar look the same because the GNOME developers were merely copying. The GNU/Linux world didn't create their own OS, they merely cloned all of the hard commercial research that produced UNIX. It's always easier to catch up, but pulling ahead is hard. The folks who want to stay on the cutting edge need to be in the commercial world. It's easy to come up with a list of commercial products and tools that haven't been cloned by an open source dude at the time of this writing: streaming video, vector animation, the full Java API, speech recognition, three dimensional CAD programs, speech synthesis, and so forth. The list goes on and on. The hottest innovations will always come from well capitalized start-ups driven by the carrot of wealth.

Most Linux users don't need to rewrite the source, but they can still benefit from the freedom. If everyone has the freedom, then someone will come along with the ability to do it and if the problem is big enough, someone probably will. In other words, only one person has to fly the X-wing fighter down the trench and blow up the Death Star.

Which is a better world? A polished Disneyland where every action is scripted, or a pile of Lego blocks waiting for us to give them form? Do we want to be entertained or do we want to interact? Many free software folks would point out that free software doesn't preclude you from settling into the bosom of some corporation for a long winter's nap. Companies like Caldera and Linuxcare are quite willing to hold your hand and give you the source code. Many other corporations are coming around to the same notion. Netscape led the way, and many companies like Apple and Sun will follow along. Microsoft may even do the same thing by the time you read this.

While Stallman didn't have monetary capital, he did have plenty of intellectual capital. By 1991, his GNU project had built many well respected tools that were among the best in their class. Torvalds had a great example of what the GPL could do before he chose to protect his Linux kernel with the license. He also had a great set of tools that the GNU project created.

Stallman's reputation also can be worth more than money when it opens the right doors. He continues to be blessed by the implicit support of MIT, and many young programmers are proud to contribute their work to his projects. It's a badge of honor to be associated with either Linux or the Free Software Foundation. Programmers often list these details on their résumés, and the facts have weight.

Of course, companies like Red Hat lie in a middle ground. The company charges money for support and plows this money back into improving the product. It pays several engineers to devote their time to improving the entire Linux product. It markets its work well and is able to charge a premium for what people are able to get for free.

He's done his part. The open source movement thrives on the GCC compiler, and Cygnus managed to find a way to make money on the process of keeping the compiler up to date. The free operating systems like Linux or FreeBSD are great alternatives for people today. They're small, fast, and very stable, unlike the best offerings of Microsoft or Apple. If the open software movement continues to succeed and grow, his child could grow up into a world where the blue screen of death that terrorizes Microsoft users is as foreign to them as manual typewriters.

One group that is locked out of the fray is the Linux community. While software for playing DVD movies exists for Macintoshes and PCs, there's none for Linux. DeCSS should not be seen as a hacker's tool, but merely a device that allows Linux users to watch the legitimate copies of the DVDs that they bought. Locking out Linux is like locking in Apple and Microsoft.

The battle between the motion picture community and the Linux world is just heating up as I write this. There will be more lawsuits and prehaps more jail time ahead for the developers who produced DeCSS and the people who shared it through their websites.

Most of the battles are not so dramatic. They're largely technical, and the free source world should win these easily. Open source solutions haven't had the same sophisticated graphical interface as Apple or Windows products. Most of the programmers who enjoy Linux or the various versions of BSD don't need the graphical interface and may not care about it. The good news is that projects like KDE and GNOME are great tools already. The open source world must continue to tackle this area and fight to produce something that the average guy can use.

Microsoft's greatest asset is the installed base of Windows, and it will try to use this to the best of its ability to defeat Linux. At this writing, Microsoft is rolling out a new version of the Domain Name Server (DNS), which acts like a telephone book for the Internet. In the past, many of the DNS machines were UNIX boxes because UNIX helped define the Internet. Windows 2000 includes new extensions to DNS that practically force offices to switch over to Windows machines to run DNS. Windows 2000 just won't work as well with an old Linux or UNIX box running DNS.

This is a typical strategy for Microsoft and one that is difficult, but not impossible, for open source projects to thwart. If the cost of these new servers is great enough, some group of managers is going to create its own open source clone of the modified DNS server. This has happened time and time again, but not always with great success. Linux boxes come with Samba, a program that lets Linux machines act as file servers. It works well and is widely used. Another project, WINE, started with the grand design of cloning all of the much more complicated Windows API used by programmers. It is a wonderful project, but it is far from finished. The size and complexity make a big difference.

The first cracks are already obvious. Microsoft lost the server market to Apache and Linux on the basis of price and performance. Web server managers are educated computer users who can make their own decisions without having to worry about the need to train others. Hidden computers like this are easy targets, and the free software world will gobble many of them up. More users mean more bug fixes and propagations of better code.

Of course, free software really isn't free. A variety of companies offering Linux support need to charge something to pay their bills. Distributions like Red Hat or FreeBSD may not cost much, but they often need some customization and hand-holding. Is a business just trading one bill for another? Won't Linux support end up costing the same thing as Microsoft's product?

Of course, there are also hard numbers. An article in Wired by Andrew Leonard comes with numbers originally developed by the Gartner Group. A 25-person office would cost $21,453 to outfit with Microsoft products and $5,544.70 to outfit with Linux. This estimate is a bit conservative. Most of the Linux cost is debatable because it includes almost $3,000 for 10 service calls to a Linux consultant and about $2,500 for Applixware, an office suite that does much of the same job as Microsoft Office. A truly cheap and technically hip office could make do with the editor built into Netscape and one of the free spreadsheets available for Linux. It's not hard to imagine someone doing the same job for about $3, which is the cost of a cheap knockoff of Red Hat's latest distribution.

Of course, it's important to realize that free software still costs money to support. But so does Microsoft's. The proprietary software companies also charge to answer questions and provide reliable information. It's not clear that Linux support is any more expensive to offer.

Also, many offices large and small keep computer technicians on hand. There's no reason to believe that Linux technicians will be any more or less expensive than Microsoft technicians. Both answer questions. Both keep the systems running. At least the Linux tech can look at the source code.

These users will be the most loyal to Microsoft because they will find it harder than anyone else to move. They can't afford to hire their own Linux gurus to redo the office, and they don't have the time to teach themselves.

These are the main weaknesses for Microsoft, and the company is already taking them seriously. I think many underestimate how bloody the battle is about to become. If free source software is able to stop and even reverse revenue growth for Microsoft, there are going to be some very rich people with deep pockets who feel threatened. Microsoft is probably going to turn to the same legal system that gave it such grief and find some wedge to drive into the Linux community. Their biggest weapon will be patents and copyright to stop the cloners.

Suddenly, brands like Hewlett-Packard or IBM can mean something when they're slapped on a PC. Any goofball in a garage can put a circuit board in a box and slap on Microsoft Windows. A big company like HP or IBM could do extra work to make sure the Linux distribution on the box worked well with the components and provided a glitch-free existence for the user.

Despite these gifts, free software will continue to grow on the campuses. Students often have little cash and Microsoft doesn't get any great tax deduction by giving gifts to individual students (that's income). The smartest kids in the dorms will continue to run Linux. Many labs do cutting-edge work that requires customized software. These groups will naturally be attracted to free source code because it makes their life easier. It will be difficult for Microsoft to counteract the very real attraction of free software.

If things go perfectly for Microsoft, the company will be able to pull out one or two patents from its huge portfolio and use these to sue Red Hat, Walnut Creek, and a few of the other major distributors. Ideally, this patent would cover some crucial part of the Linux or BSD operating system. After the first few legal bills started arriving on the desk of the Red Hat or Walnut Creek CEO, the companies would have to settle by quitting the business. Eventually, all of the distributors of Linux would crumble and return to the small camps in the hills to lick their wounds. At least, that's probably the dream of some of Microsoft's greatest legal soldiers.

This maneuver is far from a lock for Microsoft because the free software world has a number of good defenses. The first is that the Linux and BSD world do a good job of publicizing their advances. Any patent holder must file the patent before someone else publishes their ideas. The Linux discussion groups and source distributions are a pretty good public forum. The ideas and patches often circulate publicly long before they make their way into a stable version of the kernel. That means that the patent holders will need to be much farther ahead than the free software world.

Linux and the free software world are often the cradle of new ideas. University students use open source software all the time. It's much easier to do way cool things if you've got access to the source. Sure, Microsoft has some smart researchers with great funding, but can they compete with all the students?

The second defense is adaptability. The free software distributions can simply strip out the offending code. The Linux and BSD disks are very modular because they come from a variety of different sources. The different layers and tools come from different authors, so they are not highly integrated. This makes it possible to remove one part without ruining the entire system.

It will be pretty difficult for a company like Microsoft to find a patent that will allow it to deal a fatal blow to either the Linux or BSD distributions. The groups will just clip out the offending code and then work around it.

Microsoft's greatest hope is to lock up the next generation of computing with patents. New technologies like streaming multimedia or Internet audio are still up for grabs. While people have been studying these topics in universities for some time, the Linux community is further behind. Microsoft will try to dominate these areas with crucial patents that affect how operating systems deal with this kind of data. Their success at this is hard to predict. In any event, while they may be able to cripple the adoption of some new technologies like streaming multimedia, they won't be able to smash the entire world.

This does not preclude the free software world from using some ideas or software. There's no reason why Linux can't run proprietary application software that costs money. Perhaps people will sell licenses for some distributions and patches. Still, the users must shift mental gears when they encounter these packages.

One of the biggest challenges for the free software community will be developing the leadership to undertake these battles. It is one thing to mess around in a garage with your buddies and hang out in some virtual he-man/Microsoft-haters clubhouse cooking up neat code. It's a very different challenge to actually achieve the world domination that the Linux world muses about. When I started writing the book, I thought that an anthem for the free software movement might be Spinal Tap's “Flower People.” Now I think it's going to be Buffalo Springfield's “For What It's Worth,” which warns, “There's something happening here / What it is ain't exactly clear.”

driver Most computers are designed to work with optional devices like modems, disk drives, printers, cameras, and keyboards. A driver is a piece of software that translates the signals sent by the device into a set of signals that can be understood by the operating system. Most operating systems are designed to be modular, so these drivers can be added as an afterthought whenever a user connects a new device. They are usually designed to have a standard structure so other software will work with them. The driver for each mouse, for instance, translates the signals from the mouse into a standard description that includes the position of the mouse and its direction. Drivers are an important point of debate in the free software community because volunteers must often create the drivers. Most manufacturers write the drivers for Windows computers because these customers make up the bulk of their sales. The manufacturers often avoid creating drivers for Linux or BSD systems because they perceive the market to be small. Some manufacturers also cite the GNU GPL as an impediment because they feel that releasing the source code to their drivers publishes important competitive information.

GNOME The GNU Network Object Model Environment, which might be summarized as “All of the functionality of Microsoft Windows for Linux.” It's actually more. There are many enhancements that make the tool easier to use and more flexible than the prototype from Redmond. See also KDE, another package that accomplishes much of the same. (www.gnome.org)

GNU/Linux The name some people use for Linux as a way of giving credit to the GNU project for its leadership and contribution of code.

Linux The name given to the core of the operating system started by Linus Torvalds in 1991. The word is now generally used to refer to an entire bundle of free software packages that work together. Red Hat Linux, for instance, is a large bundle of software including packages written by many other unrelated projects.

C. Scott Ananian "Questions Not to Ask on Linux-Kernel", 1998-05, [http://lwn.net/980521/a/nonfaq.html].

C. Scott Ananian "A Linux Lament: As Red Hat Prepares to Go Public, One Linux Hacker's Dreams of IPO Glory Are Crushed by the Man", 1999-07-30, [http://www.salon.com/tech/feature/1999/07/30/redhat_shares/index.html].

Zack Brown "The 'Linux' vs. 'GNU/Linux' Debate", 1999-04-13, [http://www.kt.opensrc.org/kt19990408_13.html#editorial].

Rachel Chalmers "Challenges Ahead for the Linux Standards Base", 1999-04, [http://www.linuxworld.com/linuxworld/lw-1999-04/lw-04-lsb.html].

James Coates "A Rebellious Reaction to the Linux Revolution", 1999-04-25, [http://www.chicagotribune.com/business/printedition/article/0,1051,SA-Vo9904250051,00.html].

Mary Lisbeth D'Amico ""erman Division of Microsoft Protests 'Where Do You Want to Go Tomorrow' Slogan: Linux Site Holds Contest for New Slogan While Case Is Pending", 1999-04-13, [http://www.linuxworld.com/linuxworld/lw-1999-04/lw-04-german.html].

Eric Kidd "Why You Might Want to Use the Library GPL for Your Next Library", 1999-03, [http://www.linuxgazette.com/issue38/kidd.html].

LWN "Linux Beat Windows NT Handily in an Oracle Performance Benchmark", 1999-04-29, [http://rpmfind.net/veillard/oracle/].

Robert McMillan, Nora Mikes "After the 'Sweet Sixteen': Linus Torvalds's Take on the State of Linux", 1999-03, [http://www.linuxworld.com/linuxworld/lw-1999-03/lw03-torvalds.html].

Bob Metcalfe "Linux's '60s Technology: Open-Sores Ideology Won't Beat W2K, but What Will?", 1999-06-19, [http://www.infoworld.com/articles/op/xml/990621opmetcalfe.xml].

Mindcraft "Web and File Server Comparison: Microsoft Windows NT Server 4.0 and Red Hat Linux 5.2 Upgraded to the Linux 2.2.2 Kernel", 1999-04-13, [http://www.mindcraft.com/whitepapers/nts4rhlinux.html].

Eric Raymond "The Cathedral and the Bazaar: Musings on Linux and Open Source by an Accidental Revolutionary", 1999, O'Reilly.

Alessandro Rubini "Tour of the Linux Kernel Source", .

David A Rusling "The Linux Kernel", , [http://metalab.unc.edu/mdw/LDP/tlk/tlk-title.html].

Doc Searles "It's an Industry", Linux Journal, 1999-05-21, [http://www.linuxresources.com/articles/conversations/001.html].

Victoria Slind-Flor "Linux May Alter IP Legal Landscape: Some Predict More Contract Work if Alternative to Windows Catches On", National Law Journal, 1999-03-12, [http://www.lawnewsnetwork.com/stories/mar/e030899q.html].

Linus Torvalds "Linus Torvalds: Leader of the Revolution", Linux's History, 1992-07-31, [http://www.li.org/li/linuxhistory.shtml].

Dave Whitenger "Words of a Maddog", 1999-04-19, [http://linuxtoday.com/stories/5118.html].

Riley Williams "Linux Kernel Version History", , [http://ps.cus.umist.ac.uk/~rhw/kernel.versions.html].