The draft of the new GPL is very important and much is at stake in the revision process. The Linux kernel, the vast majority of the GNU project's development tool-chain, and more than seven of 10 programs under any free software license are subject to the terms of the the GPLv2 and, in the vast majority of cases, "any later version."
The GPL is so widespread that it is frequently referred to as "the Constitution of the free software movement." As it introduces changes, any discussion draft creates a potentially dangerous moment for the free software movement. While this danger is real, it does not exist to the extent or for the reasons that many in the community believe. In a way, the GPLv3 is both more and less important than many of us think.
One the one hand, those who have called the GPL "the Constitution of the free software movement" have overstated the importance of the license. A constitution is a legal document that sets the fundamental political principles of a government; it is a law to which all other laws are held.
The free software movement has such a document but it is not the GNU GPL. It is the Free Software Definition (FSD). In the FSD, free software is defined as software that satisfies Richard Stallman's now-famous four freedoms: the freedoms to use, examine, distribute, and develop software as a community. The FSD embodies the principles to which any free software license must be held -- even the FSF's own licenses and even the GPL must satisfy the FSD.
While the GNU project advocates the principles in the FSD for ethical and moral reasons, the same principles are articulated in the Open Source Definition (OSD), and the Debian Free Software Guidelines (DFSG). These documents form the Constitution of our movement. None of these documents are undergoing significant revision as part of a new GPL. There is no discussion draft for a new set of free software principles.
However, while our principles are not threatened by the current GPL revision process, our ability to act as a cohesive community through and after the GPL revision process is less secure. It is for this less obvious reason that the GPL discussion draft requires our attention and effort. After all, free and open source software is a collection of strange bedfellows: capitalists and anti-capitalists; hackers of all sorts on opposite ends of the world, the political spectrum, and any imaginable linguistic or cultural divide.
Our community has come together in ways that have been surprising to everyone but ourselves. As a result, we have been able to achieve more than anyone, including ourselves, imagined. We have come together around a common set of principles and, equally importantly, around a shared body of free, primarily GPLed, software.
The GPL is one thing that almost everyone in the free and open source software communities have in common. For that reason, the revision has the potential to highlight disagreements, differences in opinion, differences in business models, and differences in tactics. These differences are not new, but will gain a new venue for expression.
Not everyone involved will get their way. For this reason, it's important to remember that 15 years ago, when the GPLv2 was drafted, not everyone got their way either. And that was fine. Other licenses exist, and our movement is big enough for more than more one opinion and for more than one license. No number of GPL revisions will change that. We would be wise to remember that the potential for the GPL to hinder our ability to work together is far more dangerous than the even the most radical change textual change the FSF might suggest.
For the last week and for the next year, we, as a community, are faced with the task of working together to make our most important license the best license it can be. Our first goal must be to ensure that the license stays within the spirit of our shared principles.
Secondly, we must recognize that the GPL is designed to embody a set of tactical decisions that are over and beyond any definition of free software; copyleft is only one such tactic. We don't need to agree with the the FSF's tactical decisions to agree with its license. We must respect that the GPL will attempt to stay consistent with its own previous tactics and with its own historical goals.
Along these lines, and in the interest of working together toward a better license, we should proceed by asking ourselves these questions while considering the text of the license:
- Does the GPL discussion draft live up the four freedoms as we understand them?
- Does the GPL discussion draft live up the principles and goals enshrined in previous version of the GPL?
- Does the GPL discussion draft introduce practical problems, inconveniences, or annoyances that will keep free software from succeeding?
- Does the GPL discussion draft introduce problems that might prevent people currently contributing to GPL software from doing so? Similarly, might future or potential contributors be hindered?
- Could things be said more clearly in the text of the license?
These categories are not cleanly divided, or always distinguishable. That is fine. Attempting to consider issues in these terms will help us all focus our energy on the things that really matter.
By approaching problems in these terms, and by clearly noting where each problem lies, we'll be able to prioritize problems both for ourselves and for our community and we'll be able to remember exactly what is at stake. We can live with a tactically clumsy license, or an ill-wrought phrase but we cannot operate without essential freedoms. In debate and examination, it can be easy to lose sight and track of the broader issues and the larger goals.
Above all, we must remember that our community and its goals are more important than any single license -- no matter how widespread. By focusing on our goals and the community through which they are achieved, we'll be able to ensure that we have the best GPL through which to continue our work.