A colleague asked me a question about licensing. In this case of learning objects, but linking in to software and other works. In writing my specific answer to his query I found I needed to explain some of my pragmatic views on licensing, and realised that this part of my response was worth preserving and making available to others as well.

There are a number of significant points to remember when considering development of new licenses, whether it be for software, other functional material (such as learning objects or game engines) or for content (such as music, game levels or game characters).

Does the license actually meet the legitimate needs and both licensors and licensees?

If it does not, then it either won’t be used by licensors (no content) or it won’t be used by licensees (no usage), or it will be used but only in ways which violate the license (legal headaches, also see below).

Is the license actually enforceable by any means other than costly litigation?

If it’s too difficult to tell when someone is violating the license, or very difficult to prove, then to enforce it you will need costly litigation. One of the common criticisms of the GPL is that it’s never been tested in court. To my mind this shows a great deal of strength of the GPL on the following basis: there are huge amounts of GPL’d code now released, and significant use is made of it. There have been quite a lot of examples of allegations of violations of the license and in all cases the matter has been decided without need to go to a court case. Hence the GPL is a very strong legal instrument – so strong and clear that no one has been advised by lawyers that it’s worth fighting a claim of license infringement – they just negotiate a reasonable settlement, usually complying with the license from this point on – note that the enforcers of the GPL aren’t like your usual grasping company looking for punitive damages (it’s difficult to show monetary damages from GPL violation) but they’re looking for actual compliance and so are happy to settle for compliance.

Is there a simpler, easier to understand license that can achieve most of the existing goals without complicating the “license-o-spehere” any further?

There are thousands of free, open and partially open licenses out there for many types of work, without adding to it unnecessarily. Plus if an existing system does what’s needed, why re-invent the wheel?

Is the problem the license seeks to address a real “market failure” or simply a failure of courage on behalf of licensors?

If there really is a gap in the license-o-sphere which needs plugging then that’s a market failure. On the other hand, a failure of courage to release under an existing license is unlikely to be solved by fiddling around the edges of existing licenses. Any new system that requires significant investment in technology to provide usability/enforcability isn’t worth the effort if the failure of courage won’t be solved for a significant number of people. You’d be much better off spending your time advocating the benefits of open-ness and co-operation to the faint-hearted instead of developing the license and the technology needed to support it.

Does the new license assume that most people are evil and grasping and want to rip off originators, or does it assume that users (often under these sorts of license re-users rather than simply users) are honest and co-operative?

Any license that even implies that originators using the license use it because they think people are trying to rip them off will likely lead to people trying to rip them off (treat us like children and we act like children; treat us like adults and most of us will act like adults).