Oracle v. Google: We're not screwed yet

It may have fallen off the radar, but the dispute between Oracle and Google over Android and Java continues.

This week saw the delivery of the amicus curiae brief the Supreme Court requested from the U.S. Solicitor General. It looks like bad news for Google -- and in turn bad news for every programmer -- because it suggests APIs are copyrightable in the United States. Certainly every published first reaction reflects that interpretation.

So how bad is it? Merely including a header file could draw you in to the copyright maw of a software leviathan like Oracle or Microsoft, creating both an administrative nightmare for programmers and tool suppliers in the U.S. and a new competitive advantage for suppliers in Europe where the European Copyright Directive expressly excludes APIs from copyright. Oracle may only care about clawing back its investment in Sun and harming its bitter rival Google, but the collateral damage to the predominantly open source software industry could be massive.

Maybe we aren't screwed yet, though. Here are three reasons why.

1. SCOTUS may take the case anyway. The Supreme Court is not a tame animal and is completely free to accept the case as Google requests. All the same, it was SCOTUS who asked for the Solicitor General's opinion. That opinion is as unambiguous as it is uncompromising. So I'm not holding much hope on this one, and I even wonder if it would be good for developers, since there's a modest risk of the Supreme Court finding against Google and casting the next two straws to the wind.

2. It's only about massive APIs. The opinion is limited to substantial APIs on a scale that's rare in software. The Java class library hierarchy is something of a unicorn. It's hard to think of another programming API or library tree that is both enormous and proprietary. So it may not matter that unicorns are copyrightable if the rest of mammalia are spared.

3. The Solicitor General implies that Google is actually going to win but needs to pick a new strategy. The opinion is lengthy and seems to the casual reader to dismiss every point Google makes. But (among other places) it says on page 17:

   The general concerns that [Google] raises are substantial and important, but Section 102(b) is not the appropriate statutory provision to address them. Rather, legitimate concerns with interoperability and lock-in effects are far better addressed through the fair-use doctrine codified at Section 107.

Let's take that in two parts. Concerning what defense is being dismissed, my read of the Solicitor General's opinion is that it's saying the approach Google is using -- asserting that the API is not copyrightable -- is not the right defense and Google should not be able to use it. This implies a concern, valid or not, that making a distinction between APIs and their implementation would create a dangerous precedent that would chill the creative intent behind copyright law.

I think the Solicitor General's critique reflects the incorrect understanding of a non-programmer. An API is certainly written in code, but it's being used in a different way than the (indisputably copyrightable) use of code in an implementation. As Mike Godwin of the R Street Institute observes, asserting the first use of code is not copyrightable and the second is, is no more dangerous than asserting a threat to the copyrightability of prose by treating a recipe differently from the description of the actual preparation of a meal made according to the recipe.

But then look at the suggested defense. The Solicitor General seems to say Google has a great case and the correct defense is section 107 (fair use), specifically scènes à faire. A scène à faire defense would say that because Google is using Java, it has no option but to provide certain class libraries because they are inherent in the use of that programming language. In such a of situation, copyright is not operative. Godwin implies the Solicitor General may be predisposed against Google and that the negative opinion was rather expected. So if even he is positive about the chances of success of a fair use defense for an API, maybe the rest of us can breathe a little easier, too.

We still have a long wait ahead of us before there's any certainty. First we have to see if SCOTUS will accept the appeal. If it won't, we then have to wait for the original district court to consider Google's claim of fair use. We then have to wait for the appeals chain for that outcome.

Even then, we won't really know if an API is protected by copyright until another grasping corporation tries to use copyright law to rent-seek or chill competition. It's going to be a long road, so better grab any comfort that's within reach.