The judge presiding over the Google/Oracle suit has ruled that Google didn't infringe copyright by using the Java APIs, though he didn't rule on whether APIs themselves can be copyrighted.

  • http://www.facebook.com/profile.php?id=100000280198568 Felipe Budinich

    As an Atheist, all I can say is THANK GOD

  • s2redux

    ArsTechnica and PJ seem to think Alsup did rule that APIs are not subject to copyright. From the order: “So long as the specific code used to implement a method is different, anyone is free under the Copyright Act to write his or her own code to carry out exactly the same function or specification of any methods used in the Java API.”

    Any armchair lawyers out there care to translate this ruling into IANAL-speak?

  • http://www.beresourceful.net/ Rusty

    Basically the API is interpreted by the court as a combination of components that copyright does, and does not apply to. The portions that fall under copyright include the instructions on how a procedure (function, class, etc) takes a given input and generates the required output. All of that part of the API, both Oracle and Google agree has been re-written by Google, so copyright is not a concern there. The part that does not fall under copyright is the part that is needed for compatibility across implementations so far as compatibility is relevant. 

    In this case since Google specified that for Android, these 37 Java packages were needed to support the java programs that Google felt were applicable to the Android platform, a program that only calls upon procedures within those 37 packages will run on Android. Call on one of the other 166 packages in Java, and the program will fail. 

    Pointing at the user interface decisions related to Lotus 1-2-3 and Borland’s spreadsheet, the judge noted that the basic command structure (in this case the java language definition) of the API did not qualify for copyright protection.

    One other thing he noted is that if Sun had put in a patent on the functionality of Java API, that would have provided protection against Google creating it’s own implementation of Java and the API, but that was not done. In fact there are several statements in the decision that Sun’s initial position was that anyone was free to write java code, and implement their own Java runtime, but that to receive Java certification as fully supporting Java, they had to get Sun/Oracle approval passing the certification needed. And in this case Google is not interested in shipping a full Java implementation, so such certification is not needed, and Google does not advertise Android as being fully Java compatible.

  • http://shadowfirebird.tumblr.com shadowfirebird

    It seems to me that there is an obvious test here that the law does not apply.

    Copyright is meant to defend works that someone has created.   Creation implies effort (as does “works” i think).  So any so-called work that would be trivial to duplicate — like the pattern of an API call — cannot be protected.

    Mona lisa?  Protected.  Picture of a stick man that anyone could copy in five seconds? Nope.

    And so on.

    • alxr

      I’m not sure it’s as obvious as you necessarily make out. APIs are simple to copy, yes, but their design is non-trivial, at least for good APIs. And your test for copyright eligibility seems flaws – I could retype a page of text from a book in a few minutes; are books ineligible for copyright?

      • Nonentity

        It would be rather trivial to duplicate the Mona Lisa with modern equipment as well, to varying levels of verisimilitude.  A photograph would take seconds.

        Perhaps if “recreate” was intended instead of “reproduce”, but I’m sure there are artists who could create their own version from scratch in little time as well after seeing it once.

        • http://shadowfirebird.tumblr.com shadowfirebird

           I did indeed mean “recreate”.

          Artists could recreate what from scratch in little time?  The Mona Lisa?  Not sure that that’s true.

          • Nonentity

            After seeing the original?  I’m not talking about exact forgeries that would fool an art expert here, but with a little leeway for differences, absolutely.  Allow them a photograph of the original and the result will be even better.

            It also depends on what one means by “little time”, but there are places that make money as basically sweatshops turning out reproduction paintings as fast as possible.  Sure, it’s going to take longer than a stick figure, but it won’t take anywhere near as long as the original (especially if you count the years of revision the original went through). Comparatively, I’d say something like 96 hours to reproduce it with ballpoint pens ( http://www.youtube.com/watch?v=NFiib4QRsuM ) is very little time, and I have no doubt that paint would be much faster.

            Besides… the results of trying to apply your rule to something like Warhol’s output would probably tie up the legal system for decades, simply in trying to determine what point in the whole process would be the correct point to determine the ease of replication.

    • Shinkuhadoken

      It probably wasn’t the intention of your example, but it’s important to note the Mona Lisa can’t be protected by copyright because it’s too old. It’s permanently in the public domain, now.

      That aside, applying copyright to code is a tricky subject. You cannot copyright mathematics, and at a fundamental level, that’s precisely what must be avoided with the basic logical structures of code (which is basically applied math).

      It would be like copyrighting the word “the” and expecting everyone who uses that word to pay you royalties for each use. Or copyrighting sentence structures and demanding royalties from everyone who follows their nouns with a verb.

      At some point, however, copying word for word someone else’s writing as your own is plagiarizing it. But we mustn’t confuse the basic structures of code for a completed work.