Why Oracle will win its Java copyright case – and why you'll be glad when it does
Open source needs strong copyright; weak copyright only helps bullies
Comment Oracle will ultimately prevail in its Java copyright lawsuit against Google. And if you're a free software developer or supporter, you should be cheering them all the way to the wire.
Blogger John Gruber last week observed that virtually no one is rooting for Big Red. This is really quite astonishing. That's because Google has been deploying the "cuttlefish strategy," which it uses when it's threatened – spraying ink into the water so no one can really see what's going on. It's really, really good at this, and it has never used the strategy so well as here.
Google threw out so many diversions and red herrings that free and open-source software (FOSS) supporters were even cheering for a verdict that kicks away the legal basis for open source and free software. In the eyes of the "civil society" NGOs (non-governmental organizations) and compliant academics (many of whom are funded by Google), and backed by a chorus of bloggers and tech journalists who prefer a simplified, cartoon view of the world, the story was indeed simple. In their eyes, the good guys won, and that's all there was to the case.
So why is the jury's broad application of fair use in reality bad news for open source? How did Google win last week? And why will Oracle ultimately prevail? Let's take these three questions in reverse order. And strap in for the ride: The Register is not responsible for any disorientation or cognitive dissonance experienced over the next two pages.
Oracle will ultimately prevail over Google for a very simple reason: Google is guilty. Google copied 11,000 lines of someone else's copyrighted code without a license to do so. It could have chosen some other code to copy; or it could have obtained a license; or it could have not copied anything and created every single line of Android code from scratch. All three were options that Google didn't take. It's really as simple as that.
So on to the next question. How is this verdict bad for open software, when almost everything you've read insists that you reach the opposite conclusion?
In a nutshell, free and open source software depends on simple, strong copyright law. Access to justice is also a factor: if you're a small developer, you should be able to go to court to defend your license with the presumption in your favor. Contract integrity is also pretty important, but that's a given; in a judicial system in which contracts mean nothing, there is little justice for anyone, in any context. A defining characteristic of gangster states, banana republics and totalitarian regimes is that contracts mean nothing in court.
Copyleft itself is founded on strong copyright, as Richard M Stallman has pointed out, when telling off the Pirate Party about the dangers of weakening intellectual property (IP). Challenges to the GNU GPL have been infrequent and unsuccessful, not because a judge gets misty-eyed about GNU Emacs or GDB, or breaks into a spontaneous rendition of The Free Software Song – it's because of strong copyright. When a challenge to the GPL is rejected, it's because there's been a blatant infringement. On that basis, the license can be enforced. Weak copyright turns a skirmish into a legal crapshoot.
In Stallman's future utopia, there would be no copyright to speak of at all. While we're imagining futures, there's a whole fascinating new topic of discussion about the merits of software not being under copyright; that we create brand new (sui generis) IP rights more suitable for software, something (say) that's somewhere between a design right and copyright. But for now, software developers have to operate in the real world. Stallman and Eben Moglen wrote a very effective tool for the real world: the GPL. But it only works as long as copyright is strong and straightforward. Without copyright protecting your code, defending your license would be vastly more expensive and uncertain.
And that grim prospect was given life by a homesick San Francisco jury last week, whose verdict augurs a mushy middle ground where property rights exist in name only, and can't be defended. If this became the norm, a software license would mean nothing. It would be a bully's charter.
But just as there was no reason for FOSS supporters to celebrate last week's verdict, there's no reason for FOSS supporters to be overly alarmed, either. Do not panic. Oracle's lawyers are correct in asserting that it's a terrible verdict for FOSS – less so when they suggest it sets a precedent. No legal precedent has been set, for every assertion of fair use needs to be examined each time. (That's why I choose words like "prospect" and "augur" carefully; the appeals court will have little choice but to toss it out.)
It was a very peculiar set of circumstances that led to the jury finding for Google, which brings us to the third question: how exactly did Google manage to win?
A 'confused' judge was covering his ass
The circumstances of the verdict are strange and narrow. We have a somewhat red-faced Judge William Alsup, seeking self-vindication after his earlier reasoning was subsequently dismantled [PDF] by an appeals court. Judge Alsup had presided over the original trial, which reached a hung verdict in 2012. He'd instructed the jury to exempt APIs from copyright. At the retrial that ended last week, Alsup steered the jury to make a very narrow verdict. They had to answer only one question: whether a mitigation of fair use could be claimed.
In effect, Alsup was using the jury to mark his earlier homework, but advising them that there was only one acceptable answer: that Alsup was right.
In this context, all the jury's verdict means is: "We Really Don't Care." We don't care that Google didn't have a license, they were saying, and we don't care how much it copied. In Oracle's appeal hearing – which is most likely to happen next year – three experienced judges will be obliged to address the full picture, and like all good courts, they're sympathetic to property rights. The same judges were sympathetic to Oracle in 2014, when they tossed out Google's argument that the APIs enjoy some special exception from copyright.
That's what the verdict was about. So why was there so much confusion?
Copyrightability of APIs is FUD. Just ask SCO how well that went...
In 2012, Oracle appealed the original decision, seeking clarification on whether APIs were copyrightable. The appeals court concluded that without a statutory limitation or exemption just for APIs, Oracle's software interfaces had to fall under copyright. There couldn't be any other reasonable conclusion.
A great deal of FUD has been generated by Google (and its supporters, like the activist Professor Pamela Samuelson), over the issue of the "copyrightability" of APIs. In actual fact, the "copyrightability of APIs" hasn't changed since the 1976 Copyright Act came into force. And the law hasn't changed since SCO tried to use APIs in its Linux litigation - and lost.
The SCO case demonstrates that simply because APIs have always been copyrightable, it doesn't follow that APIs can be successfully deployed as the basis of an infringement claim. Think about it. Copyright requires the arrangement of a creative work or expression to be original, and pointing to an odd piece of code here or there doesn't cut it. Words are "copyrightable". But that doesn't mean we can "copyright" a word like "cloudobile", or "Itanic", and file for infringment. It isn't enough. The same goes for APIs. Google's copying of Java was so extensive, at 11,000 lines, that it knew it needed a license. Rubin said so himself.
The appeal hearings into Alsup's handiwork shed much light on the matter, with Judge Kathleen O'Malley scolding Google for muddying the waters. Fair use doesn't belong in upfront discussions about whether or not something is subject to copyright. Nor does the ability to copyright a piece of work change because the work becomes popular. Faced with Google's cuttlefish strategy, O'Malley said, Alsup had "confused" matters.
Well, don't rule out just how much some people want to be confused about copyright, and in fact willingly seek out confusion, because it lends support to their dystopian or conspiratorial political position. It's that persecution complex we keep reminding you about – persecution complexes fall apart when the sufferer realizes they're not actually being persecuted. A guilty verdict for Google wouldn't have ushered in a new IP regime where interfaces are copyrightable, because the 2013 hearings established that interfaces were already copyrightable. Nor is there a chilling effect on past or present work.
Google's cuttlefish strategy requires free software supporters to be fearful and irrational. That's not compulsory, by the way. And if you're at all serious about defending open-source software, it isn't really a sensible state of mind. ®
Consultant Florian Mueller's coverage at FOSS Patents has been far more sober and accurate than the mainstream pros. For example: the New York Times warned of a dark era of copyrightable APIs, but based this dystopian prediction on just one expert source. That turned out to be one of Google's favorite professors, Pamela Samuelson. Samuelson participated in the astroturf shill group the Authors Alliance designed to defend the ad giant in the Google Books case.
Although Mueller has consulted for companies including Oracle and Microsoft, he says he hasn't consulted for Oracle since last year. He simply reads what the law is, and isn't, and thus finds himself in a strangely unpopular place. But we live in a conformist age, where the Enlightenment practice of thinking about something from first principles, and even thinking for yourself, nullius in verba, has been replaced with: "Gee. What's the correct opinion to hold, here?"