I've been thinking about why non-commercial software written for academic research is so rarely free software, and why this is a problem. On the surface, you would think that the creators of non-commercial software often funded by philanthropies or taxpayers and distributed at no cost would be early adopters of free and open source licenses, since, unlike a commercial vendor, they've got no income from proprietary licensing that would be sacrificed in a license change. However, there has been very little progress toward software freedom in this space.
This is a problem because even if we don't want or need to modify the core functionality of an application, we need the freedoms offered by free software to do things like translate applications into other human languages, to integrate them into larger distributions of educational resources, to include them in services that are commercially funded if that makes the most sense in a given case, or port them to new operating systems or programming languages. People who think that it is sufficient to provide these freedoms on a case by case basis ("just ask!") don't understand the scale of the free software movement. If, as the world of free software grows, at each point we had to remember to ask more and more people for permission to do more things, it simply wouldn't work. Granting freedoms universally through licensing has worked, and amazingly well.
Part of the reason we have not seen much free licensing is that conducting professional research is by its nature about control, as in creating a controlled experiment. You can't be testing the effect of a piece of software that is being modified by its users or incorporating changes by a third party during the course of the experiment. Also, in the academic career path you need to be able to take credit for your work. You've got to worry about people taking credit for your work or that if your work integrates too much from others you won't be seen as original. I'm an outsider on this world, but I think that's reasonably close to correct.
On the opposite side of the balance scale, researchers are in no way accountable to educators. We are not their customers. They do not work for us, and whether we do or don't use their work may have no direct bearing on their careers. We have no direct leverage to push them to freely license their work.
Another complication is that the rhetoric in the academy around open source software, open content, access to knowledge, etc. has tended to stick to themes of collaboration and access. If, as you often hear, open source is not about licensing, it is about collaboration, then if you have no intention of creating a collaborative process, you don't need an open source license. If people only need access to a resource, then access is enough. Freedom to redistribute and modify aren't necessary.
What we have not seen in the academy is an understanding of the role of freedom. Not in the "it is morally wrong to use proprietary software" sense, but that recognizes that if the products of research in educational software are made truly free, there is a growing pool of people, governments, and companies around the world ready to leverage that freedom to create utility, to improve education for students everywhere. It doesn't matter if the researcher has no interest in collaboration and will never accept a patch from anyone. It doesn't matter if the researcher wants to control a trademark. It isn't about the open source development process. It is about freedom.