Saturday, March 21, 2015

Revival of Oracle patent claims is another reason for Google to consider Java license agreement

If one applies "If you can't beat them, join them" to intellectual property rights (IPRs), the formula reads like this: "If you can't beat them and still need them, license them."

When Google's Android team decided to use Oracle's (at the time, Sun Microsystems') Java technology without a license, it knew pretty well that this could spell trouble. Andy Rubin himself wrote in an internal email that "[doing] Java anyway" would make it necessary to "defend [that] decision, perhaps making enemies along the way." I wouldn't consider parties to every legal dispute enemies, and in my view Oracle and Google could and should be great partners. But adversaries they are. Until they settle, that is.

More than 4 1/2 years ago, Oracle (after acquiring Sun Microsystems) sued Google for alleged patent and copyright infringement. Initially, the public perception (including mine) was that the seven patents in the case were more important than the copyright part. The initial complaint stated a set of specific patents and we were all able to look at them, while the copyright claim was initially unspecified (Oracle even had to amend its complaint for that reason). But as the dispute progressed, Oracle's lawyers placed an ever clearer emphasis on the copyright part of the case. I have no idea whether this had also been the priority from the get-go, but even though it might have been misperceived as a "throw in the kitchen sink" claim, I'm quite sure Oracle knew all along that the copyright claim all by itself had the potential to convince Google of the benefits of a license agreement.

In parallel to the district court case, Google challenged all of Oracle's patents-in-suit through reexamination requests filed with the United States Patent and Trademark Office. Depending on the rules applicable to each patent, Google requested either an inter partes (where the requester remains formally involved throughout the process) or an ex parte (where the requester triggers the proceedings without becoming a party to them) reexamination. Then a variety of first Office actions (which are sometimes just meant to challenge the patentee without necessarily indicating that a patent is doomed) and of "final" (but, in reality, far from final) decisions came down. It looked like Oracle's Java patents were falling apart. The district judge was unwilling to take patent claims to trial as long as the USPTO deemed those claims invalid. Oracle took only two patents to trial (on which it didn't prevail in district court and which it didn't even care to pursue on appeal) and dropped five with prejudice in order to save time.

Even while the trial was ongoing, one of the dropped patents rose like Phoenix from the ashes, but the judge said this reversal came too late for the patent to be reasserted.

As I explained then, this still doesn't mean Google would never have to worry about that patent: even if one assumed for the sake of the argument that Judge Alsup succeeded in getting Oracle to drop certain patents without any exception as far as Google itself is concerned, Android device makers might still be liable for infringing that patent with their devices.

The same holds true for any other patents that Oracle could salvage, and one other patent is now on the road to recovery. Yesterday (Friday), the United States Court of Appeals for the Federal Circuit handed down a ruling (PDF) on Oracle's appeal of the USPTO's rejection (affirmed in November 2013 by a Patent Trial and Appeal Board, basically a USPTO-internal court) of various claims of U.S. Patent No. 6,910,205 on "interpreting functions utilizing a hybrid of virtual and native machine instructions." While claims 1 and 8 remain invalidated, the Federal Circuit has held that the USPTO (to be more precise, the PTAB) had misconstrued a key claim term (which relates to overwriting data) and therefore remands for a new decision with respect to the claims for which the interpretation of that term is outcome-determinative: claims 2–4, 15, 16, and 18–21.

Like the aforementioned "Phoenix" patent, this one has a 1997 priority date, so it won't expire before the year after next. Theoretically there would still be time (after the USPTO reaches a new decision) for Oracle to seek injunctive relief (such as an ITC import ban) over it--but in more practical terms, any assertion would likely be about damages. This is actually still more hypothetical than anything else. I guess the parties will continue to focus their litigation efforts on the copyright part of the case and just sort out the patent-related aspect as one of various items whenever they settle. In settlement negotiations all options of both litigants must obviously be considered.

For Oracle (represented by a Morrison & Foerster team led by Marc Hearron), yesterday's ruling is another appellate victory over Google that demonstrates its determination to defend its Java IP. No matter how many years it may take.

Many patents asserted in the major smartphone patent disputes have faced uphill battles when their validity was challenged by defendants. It happened to those Java patents as well. But the longer it takes, the clearer it is that some "rejections" of claims were wrong.

I plan to write about the copyright part of Oracle v. Google again soon. In January, the Supreme Court asked the Obama Administration for input. After the SCOTUS receives the Administration's brief, it will decide on whether to hear the case (if it doesn't, or if it does grant certiorari but later affirms the Federal Circuit, Oracle will have prevailed definitively on copyrightability and the case will go back to district court for "fair use" proceedings).

Just yesterday, the Wall Street Journal described Google as "one of the White House's closest corporate allies." The Journal went on to note that "Google was the second-largest corporate source of campaign donations to President Barack Obama's re-election effort" and that "Google executives have visited the White House scores of times since Mr. Obama has been in office, according to visitor logs." But Google is pretty much isolated in the U.S. economy (with the greatest respect for the likes of Oracle foe HP and Java free-rider Red Hat). The software industry is overwhelmingly on Oracle's side.

In order to support Google here, the Department of Justice would have to take outlier legal positions. As the Copyright Alliance's TerryHart wrote, "Despite legal gymnastics, Google can't land split in Oracle cert petition." Not only would the DoJ have to subscribe to daring legal theories to do Google a favor (which I'm absolutely not suggesting that it even intends to do) but it would also have to act against broader U.S. interests. By and large, the world's most valuable application programming interfaces (APIs) belong to U.S. high-tech companies or they belong to no one (because they're available on open source terms). Again, I'm going to talk about that soon.

If you'd like to be updated on the smartphone patent disputes and other intellectual property matters I cover, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents and Google+.

Share with other professionals via LinkedIn: