We've all seen the fur flying on the issue of copyright infringement damages with respect to rangeCheck and the decompiled files. It's worth stepping back and putting this into perspective, and it's also worth considering the arguments advanced.
First, the perspective. The jury found rangeCheck infringed. The jury also found the decompiled files not infringed, but Judge Alsup, on Oracle's motion for judgment as a matter of law, has overruled the jury on that one. So both rangeCheck and the decompiled files are infringing. For our purposes, let's set aside everything else (e.g., the SSO) as being unresolved, either because the jury was unable to reach a determination (the jury WAS in agreement that the SSO was infringed but not in agreement on Google's fair use defense) or because the Court has yet to rule on an issue of law (e.g., whether an implementation of a specification constitutes a derivative work of that specification). Those unresolved copyright infringement issues are more important than the issues that have been resolved.
With respect to rangeCheck and the decompiled files, Oracle led Google and the Court to believe that it would only seek statutory damages. That is the $150,000 figure you see thrown around. But Oracle did not stipulate to statutory damages, and consequently, it cannot be bound by the statement of its counsel.
Having received a verdict that rangeCheck and the decompiled files were infringed, Oracle announced that it would, in fact, seek not only statutory damages but also damages in the form of infringer's profits arising from that infringement. There were two reactions to Oracle's position. The Court was astonished given the seeming trivial nature of the infringement, and as Google pointed out, Oracle was wrong in stating it could seek both statutory and actual damages - Oracle must choose. However, Oracle does not have to make that choice until final judgment is rendered, meaning it may know what the alternative damage amounts are, i.e., Oracle can go for actual damages in the form of infringer's profits, but if the jury disagrees or agrees but only awards a modest sum, Oracle could still elect statutory damages. So the $150,000 amount is the baseline, not the ceiling.
The other position that Oracle has advanced is that it has satisfied its burden, i.e., that there was infringement and that there is a causal relationship between that infringement and Google's profits from Android. In turn, according to Oracle, the burden is now on Google to demonstrate that the profits subject to Oracle's claim should be reduced. [You can read Oracle's arguments in 1106.]
Google argues Oracle is wrong on two counts:
- The law doesn't say that the copyright holder can recover both infringer's profits and statutory damages. It says that a copyright holder can recover actual damages AND infringer's profits OR statutory damages. Google is absolutely correct on this point. See 17 USC Sec. 504(c)(1). Google is wrong to assert that Oracle must make the election (or already has made a binding election) of infringer's profits over statutory damages at this time. That election may be made any time prior to final judgment.
- Oracle must offer more proof than it has in establishing a causal relationship between the infringement and Google's profits. Oracle argues rangeCheck is in Android and, therefore, all of Google's Android revenues are subject to Oracle's claim. Most of the arguments that Google makes in its response to document 1106 (see 1114) are really arguments about why the award should be limited, not that Oracle hasn't established a causal connection between rangeCheck and Android, and therefore between rangeCheck and the Android revenues. In Google's motion for summary judgment on the issue (1125) Google points to a number of cases, including Polar Bear for the premise that Oracle must do more than throw up Google's gross revenues. True enough, Polar Bear says that, but you must take that holding in context. The Polar Bear facts would equate to Oracle simply pointing to Google's gross revenues from all sources, not just Android. The other cases cited by the parties draw this same line. I suspect the Court will find Oracle has, in fact, met its obligation (i.e., rangeCheck infringed, rangeCheck is (was) a part of Android, and here are Google's gross revenues from Android, both direct and indirect) under Polar Bear. It is far less clear that Oracle has satisfied this obligation with respect to the decompiled files since those decompiled test files were never in Android. It would seem Oracle would need to show that Android benefited from those decompiled test files.
I expect the Court to deny Google's motion for summary judgment with respect to Oracle's claim for infringer's profits and for this matter to go to the jury. At that point Google can advance all of its arguments about why those profits should be limited, and a number of those arguments are persuasive:
- The small amount of code involved when compared to all of Android;
- The fact that rangeCheck has not been present in all versions of Android (specifically, it wasn't in the originally released version nor is it in the present version);
- The qualitative insignificance of rangeCheck;
- The lack of performance benefits from the presence of rangeCheck in Android; and
- The cost of development of rangeCheck when compared to the cost of development of all of Android.
However, I just don't see the Court denying Oracle its shot, and Google will bear the burden of proving why the damages should be limited.