In Italian "api" means "bees" (plural). API in computer science means "Application Programming Interfaces", which are bits of code in a computer program that expose functions and calls to other computer programs so that they can interact. For instance, a platform (say, Gnome or MS Windows) can expose a function to call upon a printing dialog, and all application running on it only needs to invoke that function and an interaction can happen, so the application can print. The internals of the same API can be reimplemented many times without changing the outward facing interfaces, so that the applications written against them remain workable ‒ actually this happens on a regular basis.
A Federal Judge for the Northern District of California has ruled that the source code that declare a method to invoke the same function as in the API (thus representing the outward facing part) is not subject to copyright. This follows a quasi identical ruling of the European Court of Justice in the SAS case [Case C‑406/10].
I salute this pair of decisions as a cornerstone of the copyright law applied to software.
Too many "copyright maximalists" wish to say that copyright in a software program is like copyright in a literary work, thus any time you copy the same language, you infringe the copyright of the former author. Others times they say that since you cannot reproduce the same structure and plot and names of a book, doing the same with software is as well as copyright infringement (this was what SAS case was all about). Copyright in software is not the same copyright as the one granted to literary works. Software is protected as if it was a literary work, but in order to interpret and apply the copyright rules to software, one shall not disregard the utilitarian nature of software, and the need to achieve interoperability so that ‒ to use the words of the Advocate General in SAS, Mr. Bot ‒ ideas underlying software are not monopolized.
The two subjects use the same concepts, the same words, some of the same rules, but the different nature makes it impossible to just apply the same without making all necessary translations. Fortunately, there are two judges across the Atlantic that understand this and made a very wise decision. The decision from Judge Aslup in the States is even more to be appreciated, because ‒ unlike the ECJ ‒ he did not have a clear guidance on interoperability as the Software Directive (Council Directive 91/250/EEC of 14 May 1991 on the legal protection of computer programs) gave to his European Counterpart.
As always, good and extensive recounts of a very complicate case can be found in Groklaw, as well as all the documents of the case.