Oracle vs Google

Here at Android Authority, we’ve been watching closely the unfolding Oracle vs. Google legal drama. Each week of the trial has brought us valuable insight in the inner workings of Android (and Google as a whole). We’ve learned about the early childhood of Android, about how much money Google made from Android, and, in general, about how our favorite operating system came to be.

But the Oracle vs. Google affair is worth scrutinizing for more than just the wealth of inside information that is trickling out of Judge William H. Alsup’s court. The trial has major implications for the future of Android, but also unpredictable consequences for the entire tech world.

Google infringed Oracle’s copyright on Java APIs, says jury

[Updated] As our reader Adrian points out, the jury was instructed to assume that APIs are copyrighted and make a decision based on that assumption. Judge Alsup will decide if Oracle actually holds the copyright over the Java APIs.

In the first part of the trial (which was broken up into three sections – copyright, patents, and damages), the jury was asked to decide whether Google infringed on Oracle’s copyrights for the Java APIs. APIs (application programming interfaces) are omnipresent and vital to the modern technological landscape. They are the glue that keep your favorite platform together, be it Android, iOS, or Windows. A ruling that is favorable to Oracle would create a dangerous legal precedent and potentially turn the tech scene into a volatile and unpredictable conflict theater.

Oracle has already won, at least partially, in this matter. On Monday, the jury decided that Google infringed on the structure, sequence, and organization (SSO) of the Java APIs. In other words, the jury considered that Google purposely mimicked the way the Java APIs are built and structured, when it created Dalvik, its own Java-based virtual machine.

However, the twelve jurors couldn’t agree whether Google’s infringement was protected by the fair use principle or not. So, the trial moved ahead to the patent phase.

Oracle did try to have Judge Alsup make a “judgment as a matter of law” on the fair use issue, but Alsup rejected the motion. On the other side, Google countered with a motion of mistrial, which, if accepted, would cause the API copyright issue to be retried with another jury.

A final ruling on the fair use question is still to be issued.

Patent phase

With the jury hitting a deadlock concerning the fair use issue, the trial moved on to the patent phase. As you might know, Oracle initially claimed that Google’s Android infringed on seven of its Java related patents. But the USPTO and Judge Alsup denied Oracle’s request for trial on five of these patents, leaving only patent RE38,104, which describes a “method and apparatus for resolving data references in generated code” and patent 6,061,520, which describes a “method and system for performing static initialization”.

Both Google and Oracle brought to the stand their experts and witnesses, including Andy Rubin in the Android camp and former Sun VP Brian Suthpin on Oracle’s side.

Although this second phase of the trial should be dedicated to patents, we saw a number of moves related to both the copyright phase and the damages phase. In a somehow surprising decision, Judge Alsup ruled that Google did in fact infringed on Oracle’s copyrights, by decompiling and copying the code in eight Java files. Alsup effectively overturned the decision that the jury made last week. The jurors agreed that Google was only guilty of copying the now infamous nine lines of code in the file, and not the eight files in question.

What lies ahead?

Alsup’s surprising decision on the eight lifted files is a victory for Oracle, but it’s a very small one. In the damages phase, Oracle is unlikely to get any major compensation from Google. Even Judge Alsup said that it would be the “height of ridiculousness” for Larry Ellison’s company to ask 9-figure damages for Google’s infringement of the nine lines of code and the eight Java files.

Both parties seem eager to end the trial as soon as possible, but probably for different reasons. Google hopes that its mistrial motion will be accepted, while Oracle might look forward to an appeal, now that its hopes for this trial have fizzled. Ellison and co. have little chance to get a share of Google’s Android profits, like they have asked for, and statutory damages (a kind of standard amount paid in such cases) would be insignificant .

At this point, the trial is set to go ahead in the damages phase, where the jury is expected to decide what kind of compensation, if any, Oracle should receive for Google’s infringement over its Java APIs and patents. First, though, we should see a decision about the two patents, which will probably come on Monday.

Overall, it seems that the scales of  justice are tipping in Google’s favor, or at least, that’s what I make of this complicated, at points obscure legal spat. For Oracle, the potential winnings from their legal action have plummeted from the lofty six billions floated around back in 2010, to a sum that might not even cover its legal expenses.

Of course, it all depends on the jury and the judge, and it’s possible that this trial is just one battle in an attrition war that could last for years. We’ll learn more next week, when the two parties are set to spar over how much money Oracle is entitled from the Android camp.

Bogdan Petrovan
Bogdan is the European Managing Editor of Android Authority. He loves tech, travel, and fantasy. He wishes he had more time for two of those things. Bogdan's phone is a Nexus 6P.
  • Adrian

    Uh, the Jury didn’t _decide_ APIs were copyrightable. They were _told to assume_ they were by the judge and make a ruling based on that. They decided Google was infringing on copyright (under the assumption APIs are copyrightable) but were unable to decide if Google were allowed to do so under fair-use.

  • Mike

    A couple of notes. The jury was instructed to consider APIs capable of being copyrighted. The judge still has to decide as a matter of law whether the API specification (SSO) is capable of protection under copyright. If he does find it capable of protection then Google might be in for trouble. If he does not then the damages and infringement are minimal.

    The judge is not weighing in on fair use however, tht will up to a jury in case of a mistrial.

    • AppleFUD

      You are correct about the judge still needing to make the decision — let’s hope he follows the EU on this one and says they are not valid for copyright protection.

      However, from what I get from your comment you are implying that if the judge does say that they are valid for copyright that that will increase the damages. No, it won’t. They are currently working under the assumption that they are able to copyright APIs and the damages aren’t going to be much even if they are. However, the ramification if they are valid for copyright is wide and very very bad for the tech industry. Nonetheless, if the judge says they are not valid for copyright then that part of the trial is a moot point and Oracle gets nothing for it.

  • maurid

    In case Google loses, what impact does this have on Android? From a regular Android user point of view, I mean…

    • Probably not much of an impact, I’d say. It’s obvious that Oracle just wants a check, so even if it wins, the impact on the end user will be minimal. Some have speculated that this affair would force Google to drop Java and rebuild Android with another language, but that’s highly unlikely.

  • Nandan A

    Any programming language has three inseparable parts:

    1. the core language syntax
    2. built-in / standard library ( used with its standard API )
    3. tools (compiler, debugger, etc)

    Thus, it is impossible to separate the standard library ( public API ) from the language.

    For instance, can any one write a Java program to display “Hello, World” on the screen without using Java’s standard (class) library ( System class ) ?

    No way.

    Similarly, C also has its own standard library ( accessed using its own standard API ). For example, printf() is an API function of the language’s standard C library to print a string on screen.

    Thus, without built-in library, a programming language has no meaning on its own, and the library is accessed using a public API specific to that language ( printf() in C and System class in Java ).

    Thus, when we say a programming language is FREE ( not tied to a copyright ), it also means that its standard library and tools are also FREE because they are inherent parts of the language.

    However, FREE in the sense the specifications are FREE and not any particular implementations.

    Thus, it is OK if you implement your own Java implementation ( tools and library with same standard API of the language ) for any specific target / platform WITHOUT copying anybody’s implementation (source code) of the same.

    Think if C’s standard API is copyrighted, then how can we have so many C implementations for numerous OS and processors, today ?

    C implemnetations ( tools + library ) exist for hundreds of processor and microcontroller architectures today, and all implement the same standard C library with same standard API.

    • I completely agree and thanks for the detailed explanation. Unfortunately, if the judge decides that API *are* separate from the language itself and *can* be copyrighted, it could create a dangerous legal precedent that would affect everyone. Oracle would beg to differ, of course…