Unlike other stage performers and political figures, Stallman has no "off" mode. Aside from the St. Ignucius character, the ideologue you see onstage is the ideologue you meet backstage. Later that evening, during a dinner conversation in which a programmer mentions his affinity for "open source" programs, Stallman, between bites, upbraids his tablemate: "You mean free software. That"s the proper way to refer to it."
During the question-and-answer session, Stallman admits to playing the pedagogue at times. "There are many people who say, "Well, first let"s invite people to join the community, and then let"s teach them about freedom." And that could be a reasonable strategy, but what we have is almost everybody"s inviting people to join the community, and hardly anybody"s teaching them about freedom once they come in."
The result, Stallman says, is something akin to a third-world city. People move in, hoping to strike it rich or at the very least to take part in a vibrant, open culture, and yet those who hold the true power keep evolving new tricks and strategies-i.e., software patents-to keep the ma.s.ses out. "You have millions of people moving in and building shantytowns, but n.o.body"s working on step two: getting them out of those shantytowns. If you think talking about software freedom is a good strategy, please join in doing step two. There are plenty working on step one. We need more people working on step two."
Working on "step two" means driving home the issue that freedom, not acceptance, is the root issue of the free software movement. Those who hope to reform the proprietary software industry from the inside are on a fool"s errand. "Change from the inside is risky,"
Stallman stays. "Unless you"re working at the level of a Gorbachev, you"re going to be neutralized."
Hands pop up. Stallman points to a member of the golf shirt-wearing contingent. "Without patents, how would you suggest dealing with commercial espionage?"
"Well, those two questions have nothing to do with each other, really," says Stallman.
"But I mean if someone wants to steal another company"s piece of software."
Stallman"s recoils as if hit by a poisonous spray.
"Wait a second," Stallman says. "Steal? I"m sorry, there"s so much prejudice in that statement that the only thing I can say is that I reject that prejudice.
Companies that develop nonfree software and other things keep lots and lots of trade secrets, and so that"s not really likely to change. In the old days-even in the 1980s-for the most part programmers were not aware that there were even software patents and were paying no attention to them. What happened was that people published the interesting ideas, and if they were not in the free software movement, they kept secret the little details. And now they patent those broad ideas and keep secret the little details. So as far as what you"re describing, patents really make no difference to it one way or another."
"But if it doesn"t affect their publication," a new audience member jumps in, his voice trailing off almost as soon as he starts speaking.
"But it does," Stallman says. "Their publication is telling you that this is an idea that"s off limits to the rest of the community for 20 years. And what the h.e.l.l good is that? Besides, they"ve written it in such a hard way to read, both to obfuscate the idea and to make the patent as broad as possible, that it"s basically useless looking at the published information to learn anything anyway. The only reason to look at patents is to see the bad news of what you can"t do."
The audience falls silent. The speech, which began at 3:15, is now nearing the 5:00 whistle, and most listeners are already squirming in their seats, antsy to get a jump start on the weekend. Sensing the fatigue, Stallman glances around the room and hastily shuts things down. "So it looks like we"re done," he says, following the observation with an auctioneer"s "going, going, gone" to flush out any last-minute questioners. When n.o.body throws their hand up, Stallman signs off with a traditional exit line.
"Happy hacking," he says. Endnotes
1. See "Grateful Dead Time Capsule: 1985-1995 North American Tour Grosses."
2. See Evan Leibovitch, "Who"s Afraid of Big Bad Wolves," ZDNet Tech Update (December 15, 2000).
3. For narrative purposes, I have hesitated to go in-depth when describing Stallman"s full definition of software "freedom." The GNU Project web site lists four fundamental components: The freedom to run a program, for any purpose (freedom 0). The freedom to study how a program works, and adapt it to your needs (freedom 1).
The freedom to redistribute copies of a program so you can help your neighbor (freedom 2). The freedom to improve the program, and release your improvements to the public, so that the whole community benefits (freedom 3). For more information, please visit "The Free Software Definition" at 4. See Eric Raymond, "Shut Up and Show Them the Code," online essay, (June 28, 1999). 5. See "Guest Interview: Eric S. Raymond," Linux.com (May 18, 1999).
The GNU General Public License
By the spring of 1985, Richard Stallman had settled on the GNU Project"s first milestone-a Lisp-based free software version of Emacs. To meet this goal, however, he faced two challenges. First, he had to rebuild Emacs in a way that made it platform independent. Second, he had to rebuild the Emacs Commune in a similar fashion.
The dispute with UniPress had highlighted a flaw in the Emacs Commune social contract. Where users relied on Stallman"s expert insight, the Commune"s rules held. In areas where Stallman no longer held the position of alpha hacker-pre-1984 Unix systems, for example-individuals and companies were free to make their own rules.
The tension between the freedom to modify and the freedom to exert authorial privilege had been building before GOSMACS. The Copyright Act of 1976 had overhauled U.S. copyright law, extending the legal protection of copyright to software programs. According to Section 102(b) of the Act, individuals and companies now possessed the ability to copyright the "expression"
of a software program but not the "actual processes or methods embodied in the program."See Hal Abelson, Mike Fischer, and Joanne Costello, "Software and Copyright Law," updated version (1998).
Translated, programmers and companies had the ability to treat software programs like a story or song. Other programmers could take inspiration from the work, but to make a direct copy or nonsatirical derivative, they first had to secure permission from the original creator. Although the new law guaranteed that even programs without copyright notices carried copyright protection, programmers quickly a.s.serted their rights, attaching coypright notices to their software programs.
At first, Stallman viewed these notices with alarm.
Rare was the software program that didn"t borrow source code from past programs, and yet, with a single stroke of the president"s pen, Congress had given programmers and companies the power to a.s.sert individual authorship over communally built programs. It also injected a dose of formality into what had otherwise been an informal system. Even if hackers could demonstrate how a given program"s source-code bloodlines stretched back years, if not decades, the resources and money that went into battling each copyright notice were beyond most hackers" means. Simply put, disputes that had once been settled hacker-to-hacker were now settled lawyer-to-lawyer. In such a system, companies, not hackers, held the automatic advantage.
Proponents of software copyright had their counter-arguments: without copyright, works might otherwise slip into the public domain. Putting a copyright notice on a work also served as a statement of quality. Programmers or companies who attached their name to the copyright attached their reputations as well. Finally, it was a contract, as well as a statement of ownership. Using copyright as a flexible form of license, an author could give away certain rights in exchange for certain forms of behavior on the part of the user. For example, an author could give away the right to suppress unauthorized copies just so long as the end user agreed not to create a commercial offshoot.
It was this last argument that eventually softened Stallman"s resistance to software copyright notices.
Looking back on the years leading up to the GNU Project, Stallman says he began to sense the beneficial nature of copyright sometime around the release of Emacs 15.0, the last significant pre-GNU Project upgrade of Emacs. "I had seen email messages with copyright notices plus simple "verbatim copying permitted" licenses," Stallman recalls. "Those definitely were [an] inspiration."
For Emacs 15, Stallman drafted a copyright that gave users the right to make and distribute copies. It also gave users the right to make modified versions, but not the right to claim sole ownership of those modified versions, as in the case of GOSMACS.
Although helpful in codifying the social contract of the Emacs Commune, the Emacs 15 license remained too "informal" for the purposes of the GNU Project, Stallman says. Soon after starting work on a GNU version of Emacs, Stallman began consulting with the other members of the Free Software Foundation on how to sh.o.r.e up the license"s language. He also consulted with the attorneys who had helped him set up the Free Software Foundation.
Mark Fischer, a Boston attorney specializing in intellectual-property law, recalls discussing the license with Stallman during this period. "Richard had very strong views about how it should work," Fischer says, "He had two principles. The first was to make the software absolutely as open as possible. The second was to encourage others to adopt the same licensing practices."
Encouraging others to adopt the same licensing practices meant closing off the escape hatch that had allowed privately owned versions of Emacs to emerge. To close that escape hatch, Stallman and his free software colleagues came up with a solution: users would be free to modify GNU Emacs just so long as they published their modifications. In addition, the resulting "derivative" works would also have carry the same GNU Emacs License.
The revolutionary nature of this final condition would take a while to sink in. At the time, Fischer says, he simply viewed the GNU Emacs License as a simple contract. It put a price tag on GNU Emacs" use. Instead of money, Stallman was charging users access to their own later modifications. That said, Fischer does remember the contract terms as unique.
"I think asking other people to accept the price was, if not unique, highly unusual at that time," he says.
The GNU Emacs License made its debut when Stallman finally released GNU Emacs in 1985. Following the release, Stallman welcomed input from the general hacker community on how to improve the license"s language. One hacker to take up the offer was future software activist John Gilmore, then working as a consultant to Sun Microsystems. As part of his consulting work, Gilmore had ported Emacs over to SunOS, the company"s in-house version of Unix. In the process of doing so, Gilmore had published the changes as per the demands of the GNU Emacs License. Instead of viewing the license as a liability, Gilmore saw it as clear and concise expression of the hacker ethos. "Up until then, most licenses were very informal," Gilmore recalls.
As an example of this informality, Gilmore cites a copyright notice for trn, a Unix utility. Written by Larry Wall, future creator of the Perl programming language, patch made it simple for Unix programmers to insert source-code fixes-" patches" in hacker jargon-into any large program. Recognizing the utility of this feature, Wall put the following copyright notice in the program"s accompanying README file:
Copyright (c) 1985, Larry Wall You may copy the trn kit in whole or in part as long as you don"t try to make money off it, or pretend that you wrote it.See Trn Kit README.
Such statements, while reflective of the hacker ethic, also reflected the difficulty of translating the loose, informal nature of that ethic into the rigid, legal language of copyright. In writing the GNU Emacs License, Stallman had done more than close up the escape hatch that permitted proprietary offshoots. He had expressed the hacker ethic in a manner understandable to both lawyer and hacker alike.
It wasn"t long, Gilmore says, before other hackers began discussing ways to "port" the GNU Emacs License over to their own programs. Prompted by a conversation on Usenet, Gilmore sent an email to Stallman in November, 1986, suggesting modification: You should probably remove "EMACS" from the license and replace it with "SOFTWARE" or something. Soon, we hope, Emacs will not be the biggest part of the GNU system, and the license applies to all of it.See John Gilmore, quoted from email to author.
Gilmore wasn"t the only person suggesting a more general approach. By the end of 1986, Stallman himself was at work with GNU Project"s next major milestone, a source-code deb.u.g.g.e.r, and was looking for ways to revamp the Emacs license so that it might apply to both programs. Stallman"s solution: remove all specific references to Emacs and convert the license into a generic copyright umbrella for GNU Project software. The GNU General Public License, GPL for short, was born.
In fashioning the GPL, Stallman followed the software convention of using decimal numbers to indicate prototype versions and whole numbers to indicate mature versions. Stallman published Version 1.0 of the GPL in 1989 (a project Stallman was developing in 1985), almost a full year after the release of the GNU Deb.u.g.g.e.r, Stallman"s second major foray into the realm of Unix programming. The license contained a preamble spelling out its political intentions:
The General Public License is designed to make sure that you have the freedom to give away or sell copies of free software, that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things.
To protect your rights, we need to make restrictions that forbid anyone to deny you these rights or to ask you to surrender the rights. These restrictions translate to certain responsibilities for you if you distribute copies of the software, or if you modify it.See Richard Stallman, et al., "GNU General Public License: Version 1," (February, 1989).
In fashioning the GPL, Stallman had been forced to make an additional adjustment to the informal tenets of the old Emacs Commune. Where he had once demanded that Commune members publish any and all changes, Stallman now demanded publication only in instances when programmers circulated their derivative versions in the same public manner as Stallman. In other words, programmers who simply modified Emacs for private use no longer needed to send the source-code changes back to Stallman. In what would become a rare compromise of free software doctrine, Stallman slashed the price tag for free software. Users could innovate without Stallman looking over their shoulders just so long as they didn"t bar Stallman and the rest of the hacker community from future exchanges of the same program.
Looking back, Stallman says the GPL compromise was fueled by his own dissatisfaction with the Big Brother aspect of the original Emacs Commune social contract.
As much as he liked peering into other hackers"
systems, the knowledge that some future source-code maintainer might use that power to ill effect forced him to temper the GPL.
"It was wrong to require people to publish all changes," says Stallman. "It was wrong to require them to be sent to one privileged developer. That kind of centralization and privilege for one was not consistent with a society in which all had equal rights."
As hacks go, the GPL stands as one of Stallman"s best.
It created a system of communal ownership within the normally proprietary confines of copyright law. More importantly, it demonstrated the intellectual similarity between legal code and software code.
Implicit within the GPL"s preamble was a profound message: instead of viewing copyright law with suspicion, hackers should view it as yet another system begging to be hacked.
"The GPL developed much like any piece of free software with a large community discussing its structure, its respect or the opposite in their observation, needs for tweaking and even to compromise it mildly for greater acceptance," says Jerry Cohen, another attorney who helped Stallman with the creation of the license. "The process worked very well and GPL in its several versions has gone from widespread skeptical and at times hostile response to widespread acceptance."
In a 1986 interview with Byte magazine, Stallman summed up the GPL in colorful terms. In addition to proclaiming hacker values, Stallman said, readers should also "see it as a form of intellectual jujitsu, using the legal system that software h.o.a.rders have set up against them."See David Betz and Jon Edwards, "Richard Stallman discusses his public-domain [sic] Unix-compatible software system with BYTE editors," BYTE (July, 1996).
(Reprinted on the GNU Project web site: This interview offers an interesting, not to mention candid, glimpse at Stallman"s political att.i.tudes during the earliest days of the GNU Project. It is also helpful in tracing the evolution of Stallman"s rhetoric.
Describing the purpose of the GPL, Stallman says, "I"m trying to change the way people approach knowledge and information in general. I think that to try to own knowledge, to try to control whether people are allowed to use it, or to try to stop other people from sharing it, is sabotage." Contrast this with a statement to the author in August 2000: "I urge you not to use the term "intellectual property" in your thinking. It will lead you to misunderstand things, because that term generalizes about copyrights, patents, and trademarks.