T1#open-source#research
The GNU Project Begins — Stallman's Free-Software Declaration

Metadata
- Date
- Decade
- 1980s
- Tier
- T1
- Sources
- 07
- Connections
- 01
- Tags
- #open-source#research
On 27 September 1983, Richard Stallman, a hacker at the MIT Artificial Intelligence Laboratory, posted a message to the Usenet newsgroups net.unix-wizards and net.usoft. The subject line was flat—new UNIX implementation. The body opened with two words: "Free Unix!"
Starting this Thanksgiving I am going to write a complete Unix-compatible software system called GNU (for Gnu's Not Unix), and give it away free to everyone who can use it.
This was the founding declaration of the GNU Project—the moment a forty-year movement crystallised around a single conviction: that the freedom to redistribute and modify software should be defended not by philosophy alone, but by working code paired with a legal framework. gnu.org attaches its own caveat to the text: the actual history "differs in many ways from this initial plan"—the start slipped to January 1984, and several of the philosophical concepts of free software were not clarified until years later.
Why Stallman Was Angry
The MIT AI Lab of the 1970s was the symbolic centre of hacker culture. Programmers shared a PDP-10, read each other's source, and traded modifications. The premise was unstated but firm: code is a form of knowledge, and knowledge is to be shared.
Two events broke that premise for Stallman.
The Lisp Machine exodus (around 1981). Members of the MIT AI Lab left to commercialise the Lisp Machine and locked internal code away as trade secrets. In Stallman's own account, "in 1981, the spin-off company Symbolics had hired away nearly all of the hackers from the AI Lab, and the depopulated community was unable to maintain itself."
The printer control program. Stallman asked for the source of the control program for the lab's printer and was refused on the grounds of a nondisclosure agreement. He cites it as his first experience of being on the receiving end of an NDA: the right to fix the fault in front of him had been taken away by the legal construct of corporate ownership.
In January 1984 Stallman quit his job at MIT and began writing GNU software—some three months after the announcement, and partly to deny MIT any future claim of intellectual property over the code.
The GNU Manifesto (1985)
In March 1985, Stallman published the GNU Manifesto in Dr. Dobb's Journal, asking for support in developing the GNU operating system; part of the text was carried over from the 1983 announcement.
It is worth being exact here, because the point is often garbled: the four freedoms of free software are not in the Manifesto. The Manifesto enumerates nothing of the kind. By the FSF's own account, the freedoms were originally numbered 1, 2, and 3 around 1990, and freedom 0 was added afterwards because running the program was judged more fundamental than the rest. The settled wording is:
- Freedom 0 — the freedom to run the program as you wish, for any purpose
- Freedom 1 — the freedom to study how the program works, and change it so it does your computing as you wish
- Freedom 2 — the freedom to redistribute copies so you can help others
- Freedom 3 — the freedom to distribute copies of your modified versions to others
The word "free" here means "freedom", not "free of charge". The definition's own phrasing: "free as in 'free speech,' not as in 'free beer.'"
In October of the same year, he founded the Free Software Foundation (FSF) to serve as the legal and financial vessel of the GNU project.
The GPL — Freedom Encoded as Licence
Philosophy alone could not prevent companies from absorbing GNU code, modifying it, and shipping it as closed binaries. Stallman engineered a legal mechanism.
Copyleft inverts copyright. By writing into the licence the condition that any redistribution must be on the same terms, copyleft makes derivative works inherit the freedom as a matter of contract. The word was not his invention: in 1984 or 1985 Don Hopkins mailed him an envelope scribbled with the joke "Copyleft—all rights reversed", and Stallman borrowed it as the name for the distribution scheme he was then designing.
- GPLv1 (1989) — first unified GNU General Public License; the core bargain, source made available whenever binaries are distributed, is already here
- GPLv2 (1991) — added the section 7 "Liberty or Death" clause: if patents or other obligations make it impossible to honour the GPL's terms, you may not distribute at all
- GPLv3 (2007) — addressed "tivoisation" (locking down hardware to deny the user freedom over modified code); strengthened patent provisions further
The GPL is sometimes called "viral" because derivatives must remain GPL. That property is precisely what makes it a legal shield for the open commons.
GNU/Linux — A Fusion by Accident
By the early 1990s, the GNU project had completed nearly the entire UNIX userland: compiler (GCC), editor (Emacs), shell (bash), the core utilities. What it lacked was a working kernel—the GNU Hurd's microkernel design proved hard to finish.
In 1991, Linus Torvalds in Finland released the Linux kernel. Its first licence was his own, forbidding distribution for a fee; he switched it to the GNU GPL with version 0.12 in January 1992, prompted by requests to drop the no-money condition. The GNU userland and the Linux kernel snapped together. GNU/Linux existed as a fully functioning free operating system.
Stallman still insists the system should be called GNU/Linux rather than Linux. The wider industry mostly does not. As a technical matter, however, a substantial share of the code in a typical Linux distribution traces to the GNU project.
Where the 1983 Post Lives Today
The terrain forty years on:
- Cloud infrastructure — most VMs on AWS, Azure, and GCP are Linux, which is to say GNU/Linux
- Android — the kernel is GPLv2 Linux, but everything above it was deliberately built to avoid the GPL: Bionic, its libc, is BSD-licensed, and the framework is Apache 2.0
- Containers and Kubernetes — they depend on namespaces and cgroups in the GPLv2 kernel, though Docker and Kubernetes are themselves Apache 2.0, not copyleft
- TiVo and locked-down embedded devices — the immediate practical reason for GPLv3
- AI training data — GPL-licensed source code is at the centre of ongoing debate over what may legally be used to train models
The GNU project's invention was securing the freedom of code through licence. It is not, however, the ancestor of every open-source licence: the permissive line—the BSD licences, and MIT/X11 from 1987—predates GPLv1 (1989) and ran alongside it. What GNU created was copyleft, a lineage that runs on through the MPL, the AGPL, the EUPL, and outside software into Creative Commons' ShareAlike condition. Stallman's 1983 Usenet post was not the announcement of a personal project. It was the first architectural drawing for the legal and economic structure that governs code as a public good.
Questions this page answers
- When did the GNU Project begin?
- On 27 September 1983, when Richard Stallman posted to the Usenet groups net.unix-wizards and net.usoft under the subject 'new UNIX implementation'. Actual work on the system began in January 1984.
- What does GNU stand for?
- It stands for 'GNU's Not Unix' — a recursive acronym that contains itself, naming a system compatible with UNIX but not UNIX.
- How are GNU and Linux related?
- GNU set out to write a complete UNIX-compatible system that could be redistributed freely, and for years it lacked a kernel. When the Linux kernel adopted GNU's GPLv2 in 1992 the two fused, forming the substrate of the modern open-source economy.
Sources
TertiaryGNU Project — Wikipedia
Last updated: