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 nobody throws their hand up, Stallman signs off with a traditional exit line.
“Happy hacking”, he says.
Chapter 9. The GNU General Public License
By the spring of 1985, Richard Stallman had settled on the 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”.[1] 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 asserted 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 assert 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 shore 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
