As a developer of new language technologies, it is your goal to create something either scientifically interesting, practically useful, or both. Your success at attracting a market (either of other researchers, of software developers, or others) is fundamentally constrained by the rights you are willing to and capable of granting them.
If you cannot prove that your project uses only properly licensed datasets, or that you are the actual author of the code (and thus have the right to redistribute it), many businesses will avoid you for fear of starting an expensive legal battle in the future.
In the world of science, if you cannot grant access to the code or data you used to reach your conclusions, others will have greater difficulty reproducing and reasoning about your results and comparing against or improving on your work.
In these notes, I'll try to provide a basic introduction to the ideas involved, and where I can I will try to provide links to interesting legal interpretations and their effects all over the world.
The basic idea behind most licenses is one of the following three:
Here "consumer" means anyone who you transfer your work to in compiled form.
The self-propogating characteristics lead some to name "libre" licensing "viral;" the social welfare focus of the license terms lead some to call it "copyleft." Similarly, "open" licenses are sometimes called "permissive."
Between closed and open licensing strategies are licensing terms which restrict modification but permit redistribution (and sometimes make available the source); this is commonly used for artwork.
Examples of proprietary licenses are found in most commercial software; you receive a binary executable, and are not given access to the source code. You are generally not allowed to redistribute this executable.
Examples of "open" software are the underlying Android userspace (used to make the various flavours of Android, e.g. Google Android), the various BSD kernels (used to make the kernels of MacOS), Mozilla Firefox, and the dynamic programming language Ruby.
Examples of "libre" (sometimes called "free", as in свободно and not in the sense of бесплатно) software are Linux, all tools under the GNU project, and Python.
While open and libre licensing found their beginnings in the free software movement, their ideas have started to propogate outward into a "free culture" movement. The technical language used in the software licenses needed adaptation for other purposes.
The main kinds of works we will focus on licensing of are:
A scientific project might have all of these:
Each of these will likely require a different license, no matter the objectives of the author.