Thinking from first principles, and forgive me for going into hippie philosophical mode here but I can’t see how an API can have copyright that is owned.<p>Creating an API is of course a creative process. Getting the API right is, for me, the most creative part of programming. When well written, it describes to a human how the software truly works, with the implementation really being the computer version of what the function name is already telling you. A browser rendering engineer could be defined just as much by the model that the DOM API describes as it is defined by the actual source code. They both describe the outside and inside of the same thing, a system which, when it’s interface is shared, is shared between us all and not just under the sole ownership of the first person to describe it.<p>It’s hard to see an API therefore, particularly a published one, as a traditional piece of intellectual property that can be subject to copyright. One cannot copyright F=ma or e=mc^2. Once these object models about software are discovered they are like descriptions of the natural world and their formulae are open and shared for all to use.<p>Yes: how you build your specific machine that makes use of and conforms to the API — the actual source code for function implementations — can be your <i>intellectual property</i>, but the underlying description of the natural world and its API are discoveries that are part of
the commons, for all to interact with, use, and re-use.<p>The only difference between natural laws, in my analogy to physics, and software APIs is that there is only one physical world with a limited set of natural laws that describe it. With software engineering we create our own new universes everyday, but they are common universes for us all to share.