Mark Bowman is a JD candidate at Osgoode Hall Law School.
Google’s attempt to dismiss via summary judgment copyright infringement accusations from Oracle against its Java-based Android operating system have failed, with one exception. In a judgment filled with veiled disdain for Google’s motion, the United States District Judge William Alsup found that the main subject matter of the motion, the APIs, was not sufficiently defined by Google to pass the copyright challenges presented in Oracle’s original complaint, with the exception of names found within the APIs.
Oracle previously filed notice on September 12, 2010, for both patent and copyright infringement, and this latest motion from Google had aims to remove the copyright component from the complaint by showing that 12 source code files do not have copyright protection. Google also asserted that the names, scènes à faire in nature, used as a method of operation a degree of similarity with the original files and the fair use nature of the 37 API package specifications removed any copyright protection. Oracle has estimated damages from the patent and copyright infringement at $1.2 billion, and the two parties are actively negotiating a settlement, although reports are a settlement will not be reached in the immediate future.
APIs, or application programming interfaces, are extremely brief pieces of code that wrap around either a program or a library (a grouping of smaller pieces of code that does not run by itself but provides functionality to other programs) and act as a local specification or standard for any connecting program to use. The irony of this copyright infringement claim is that APIs are usually exposed to third parties to allow for software to grow beyond the original design and use, a technique that is found in almost every modern web-based service. Oracle has objected because Google has replaced code within 12 code files found on the other side of the Java APIs, something that Google probably has done to ensure positive and consistent Android customer experience on the various mobile platforms that it runs on.
Google argued that the 12 code files were a small part of a larger monolithic Java copyright and thus too meager to be considered copyright infringement (Java is exponentially larger in size than 12 files). The judge ruled against Google and found that each individual file would have its own copyright protection. Google also asserted that the API specifications themselves were not subject to copyright protection due to the general nature of the specifications as names, the scènes à faire nature of the technique, the fact that APIs are a ‘method of operation’, that APIs fall under the ‘virtual identify standard’, and that the use of the Java APIs was ‘fair use’.
Justice Alsup did side with Google on the first of these issues and found that the various names associated with the specifications (the names of the methods found within the API, the name of the API, and any other class name) were not subject to copyright. The judge however dismissed the other notions that the use of APIs within software was scène à faire (an element of created work that is either obligated or expected within the work, and thus not copyrightable). The Court also disagreed with the argument that, as a ‘method of operation’ within software development, an API is not copyrightable (as defined by 17 U.S.C. 102(b)). The judge further rejected Google’s arguments that in order to be deemed copyrighted work the material had to be virtually identifiable and not substantially similar (Android’s versions of the APIs would be similar but not identical), and the Java API packages cannot be deemed either fair use or not (as defined in 17 U.S.C. 107) due to existing disputed questions of material fact.
The response by the District Court to Google’s request still leaves much to be decided by a jury if a settlement cannot be reached before trial. Oracle’s claim that ‘Android fragmented the Java platform and locked Java out of the smartphone market’ is a bit unexpected as Oracle does not have any existing presence within the mobile market, nor has it announced any initiatives to make its presence known; the theory is that Oracle’s motivation behind this suit could be either to do a favour for Apple (Android’s main competitor in the mobile market), or to protect against future copyright infringement claims against their own products (including Java). It is also of interest that both Java and Android are open-source software covered by the GNU General Public License, which could be sufficient to suggest fair use if this complaint reaches trial.
Canadian copyright law contains a similar provision covering ‘using any method or principle of manufacture or construction’ (64.1(1)(d) of the Copyright Act, 1985), however this has not been tested against software APIs. BlackBerry, the third largest platform in the smartphone market, might provide the opportunity as they also offer Java as a development environment for their phones and, depending on how it is been implemented, could potentially face the same complaint from Oracle in Canadian Courts.