A Licence to Print...Licences

Licensing lies at the heart of free software. Indeed, it could be argued that Richard Stallman's greatest legacy is the GNU GPL, since that first showed how to preserve the essential liberty of free software, and how to deal with free-riders. But as well as a boon, licences are also a bane: there are too many of the damn things, which is why I was a little unkind to the Honest Public Licence idea, good in itself.

In a way, it's surprising that it has taken the open source world so long to do some navel-gazing and look closely at the state of open source licences. The result, a draft of the License Proliferation Committee Report, makes fascinating reading.

Originally, the LP Committee started to divide the OSI approved licenses into "recommended," "non-recommended" and "other" tiers. As we met and discussed, however, it became apparent that there is no one open source license that serves everyone's needs equally well. Some people like copyleft. Some don't. Governmental bodies have specific needs concerning copyright rights. As we discussed which licenses should be "recommended," it became clear that the recommended licenses were really the same as licenses that were either widely used (for example the GPL), or that had a strong community (for example Eclipse). Thus, we switched from the "recommended"/"non-recommended" terminology to a more descriptive terminology of:

-Licenses that are popular and widely used or with strong communities

-Special purpose licenses

-Licenses that are redundant with more popular licenses

-Non-reusable licenses

-Other/Miscellaneous licenses

We thought that these more descriptive categories may help people initially picking a license to use one of the more popular licenses, thereby helping to reduce the numbers of different licenses commonly used. We realize that the majority of open source projects currently use the GPL and that the GPL does not always play well with other licenses. We also realize that the GPL is a great license choice for some people and not so great a license choice for others. Thus, we can't just recommend that everybody use the GPL.. While such a recommendation would solve the license proliferation problem, it is not realistic.

We encourage new licensors to use licenses in the "popular and strong communities" group if any licenses in that group fit their needs. There are only nine licenses in this group and if everyone considered these licenses first when choosing a license for their project, some of the issues relating to license proliferation would diminish.

What's particularly interesting is that there are just nine licences in the "popular and strong communities" group, and that they are mainly the ones you'd expect:

- Apache License, 2.0

- New BSD license

- GNU General Public License (GPL)

- GNU Library or "Lesser" General Public License (LGPL)

- MIT license

- Mozilla Public License 1.1 (MPL)

- Common Development and Distribution License

- Common Public License

- Eclipse Public License

Most of these are well known; the only "strange" ones are the Common Public License, an early IBM choice, and Sun's Common Development and Distribution License.

Also of note is the Wizard Project:

The wizard assists new licensors in choosing which licenses meet their goals. The wizard also lets licensors find licenses that almost meet their goals. We hope that being able to generate a list of existing licenses that meet defined goals will lessen the need for people to create their own new licenses.

This is very similar to a tool available on the Creative Commons site. Indeed, it's hard not to get the feeling that on this occasion the open source world is generally following developments in the open content world - not necessarily a bad thing, and a sign of the growing maturity of the latter.