next up previous contents
Next: KDE, Qt, and Troll Up: The Open Source Definition Previous: The Open Source Definition


The concept of free software is an old one. When computers first reached universities, they were research tools. Software was freely passed around, and programmers were paid for the act of programming, not for the programs themselves. Only later on, when computers reached the business world, did programmers begin to support themselves by restricting the rights to their software and charging fees for each copy. Free Software as a political idea has been popularized by Richard Stallman since 1984, when he formed the Free Software Foundation and its GNU Project. Stallman's premise is that people should have more freedom, and should appreciate their freedom. He designed a set of rights that he felt all users should have, and codified them in the GNU General Public License or GPL. Stallman punningly christened his license the copyleft because it leaves the right to copy in place. Stallman himself developed seminal works of free software such as the GNU C Compiler, and GNU Emacs, an editor so alluring to some that it is spoken of as it were a religion. His work inspired many others to contribute free software under the GPL. Although it is not promoted with the same libertarian fervor, the Open Source Definition includes many of Stallman's ideas, and can be considered a derivative of his work.

The Open Source Definition started life as a policy document of the Debian GNU/Linux Distribution. Debian, an early Linux system and one still popular today, was built entirely of free software. However, since there were other licenses than the copyleft that purported to be free, Debian had some problem defining what was free, and they had never made their free software policy clear to the rest of the world. I was the leader of the Debian project, at that time, and I addressed these problems by proposing a Debian Social Contract and the Debian Free Software Guidelines in July 1997. Many Debian developers had criticisms and improvements that I incorporated into the documents. The Social Contract documented Debian's intent to compose their system entirely of free software, and the Free Software Guidelines made it possible to classify software into free and non-free easily, by comparing the software license to the guidelines.

Debian's guidelines were lauded in the free software community, especially among Linux developers, who were working their own free software revolution at the time in developing the first practical free operating system. When Netscape decided to make their web browser free software, they contacted Eric Raymond. Raymond is the Margaret Meade of free software: he has written several anthropological articles explaining the free software phenomenon and the culture that has grown around it, works that are the first of their kind and have shown a spotlight on this formerly little-known phenomenon. Netscape management was impressed with Raymond's essay ``The Cathedral and the Bazaar,'' a chronicle of a successful free software development using unpaid volunteer contributors, and asked him to consult, under a non-disclosure agreement, while they developed a license for their free software. Raymond insisted that Netscape's license comply with Debian's guidelines for it to be taken seriously as free software.

Raymond and I had met occasionally at the Hacker's Conference, an invitation-only gathering of creative and unconventional programmers. We had corresponded on various subjects via email. He contacted me in February of 1997 with the idea for Open Source. Raymond was concerned that conservative business people were put off by Stallman's freedom pitch, which was, in contrast, very popular among the more liberal programmers. He felt this was stifling the development of Linux in the business world while it flourished in research. He met with business people in the fledgling Linux industry, and together they conceived of a program to market the free software concept to people who wore ties. Larry Augustin of VA Research and Sam Ockman (who later left VA to form Penguin Computing) were involved, as well as others who aren't known to me.

Some months before Open Source, I had conceived of the idea of Open Hardware, a similar concept but for hardware devices and their interfaces rather than software programs. Open Hardware has not been as successful as Open Source to date, but it is still operating and you can find information on it at

Raymond felt that the Debian Free Software Guidelines were the right document to define Open Source, but that they needed a more general name and the removal of Debian-specific references. I edited the Guidelines to form the Open Source Definition. I had formed a corporation for Debian called Software in the Public Interest, and I offered to register a trademark for Open Source so that we could couple its use to the definition. Raymond agreed, and I registered a certification mark, a special form of trademark meant to be applied to other people's products, on the term. About a month after I registered the mark, it became clear that Software in the Public Interest might not be the best home of the Open Source mark, and I transferred ownership of the mark to Raymond. Raymond and I have since formed the Open Source Initiative, an organization exclusively for managing the Open Source campaign and its certification mark. At this writing, the Open Source Initiative is governed by a six-person board chosen from well-known free software contributors, and seeks to expand its board to about ten people.

At the time of its conception there was much criticism for the Open Source campaign, even among the Linux contingent who had already bought-in to the free software concept. Many pointed to the existing use of the term `` Open Source'' in the political intelligence industry. Others felt the term `` Open'' was already overused. Many simply preferred the established name Free Software. I contended that the overuse of `` Open'' could never be as bad as the dual meaning of ``Free'' in the English language -- either liberty or price, with price being the most oft-used meaning in the commercial world of computers and software. Richard Stallman later took exception to the campaign's lack of an emphasis on freedom, and the fact that as Open Source became more popular, his role in the genesis of free software, and that of his Free Software Foundation, were being ignored -- he complained of being ``written out of history.'' This situation was made worse by a tendency for people in the industry to compare Raymond and Stallman as if they were proponents of competing philosophies rather than people who were using different methods to market the same concept. I probably exacerbated the situation by pitting Stallman and Raymond against each other in debates at Linux Expo and Open Source Expo. It became so popular to type-cast the two as adversaries that an email debate, never intended for publication, appeared the online journal Salon. At that point, I asked Raymond to tone down a dialogue that it had never been his intent to enter.

When the Open Source Definition was written, there were already a large number of products that fit the definition. The problem was with programs that did not meet the definition, yet were seductive to users.

next up previous contents
Next: KDE, Qt, and Troll Up: The Open Source Definition Previous: The Open Source Definition

Download this document: [src.tar.gz][ps.gz][html.tar.gz][dvi.gz]

Open Resources (
Last updated: 1999-08-06