Later today, Oracle CEO Larry Ellison and Google CEO Larry Page, both likely accompanied by other top-level executives, will attend a court-ordered mediation session at the United States District Court for the Northern District of California. If they don't reach an agreement, today, the court may order them to additional meetings between tomorrow and the end of the month.
The court can, and did, order them to attend. It can't force them to agree on any particular terms. Either party can say no because it's entitled to its day in court. If they don't agree at this stage, they could still do so later. This is the best chance for a settlement so far, but not the last.
Both parties officially welcomed the presiding judge's proposal for settlement talks. That doesn't mean anything. They know the judge wants them to settle, and snubbing him is not a sensible option. Even if neither party truly believed that today's talks are likely to lead to a result, each party would want to appear constructive and reasonable, hoping that the court will attribute to the other party any failure to work it out. It's all about goodwill.
In the build-up to these talks, I analyzed the state of affairs of Google's 20 affirmative defenses (they can still raise them all at a possible trial but things haven't gone well for them so far) and what the terms and deal structure of a settlement could look like, saying that even a bid over pi billion dollars may not be enough.
In this final post of a three-part series, I'd like to share my thoughts on how likely they are to work it out:
The judge played this smart in many ways and has ensured that both parties have some remaining risks in this so that either one may choose certainty. A jury trial always comes with many imponderabilities. However, that was also the case between Oracle and SAP, and nevertheless the case went to trial.
As I'll explain further below, I believe the judge made one -- only one -- mistake here by ordering settlement talks without a firm trial date. The absence of a firm trial date is a major risk factor to this effort in and of itself, and even more so in combination with an important circumstance: the regulatory reviews (in various jurisdictions) of Google's proposed acquisition of Motorola.
Google's chances of getting out of this lawsuit unscathed are rather slim. Oracle probably wouldn't win a trial on all counts, but it's highly likely to win on at least some of them, and quite probably the counts on which Oracle would prevail would be powerful enough. However, Google has to think about the implications of any settlement of this particular litigation for its overall business.
I believe there's a reasonable chance that Larry Page will show the strong leadership he's demonstrated since taking over the helm and make Larry Ellison an offer too good to refuse: a ton of money in exchange for a perpetual license. The side effect of encouraging other patent holders to expect similar pay-offs is inevitable but not a reason not to do what needs to be done.
Such an offer could come with a tacit or between-the-lines threat of possibly asserting patents against Oracle, especially some of the 2,000+ patents recently acquired from IBM. However, Oracle is rather likely to have decisive leverage over Google in the current, fairly advanced case before any late countersuit would bear fruit. Therefore, Google might prefer to go down Escalation Avenue only if the Android/Java case gets stayed, in which case some new litigation would have a chance to catch up.
If Oracle is adamant about a per-unit royalty or Google doesn't offer nearly enough to make a one-off payment palatable, Google will have to change its Android business model to accede to Oracle's demands on a commercially viable basis. That's the biggest problem. Without that problem, they might have settled this a while ago.
Changing Android's licensing model from "free of charge for OEMs" (even if not really open) to a commercial licensing model with which Google could pass on its license fees to OEMs may have to happen sooner or later, just because of all the patent holders out there who are collecting royalties. But in the current situation, such a change would be particularly difficult for Google to do. Google is currently trying to secure regulatory clearance for its proposed acquisition of Motorola Mobility. To the extent that regulators raise concerns about how other Android companies can compete with Google post-acquisition, Google will certainly argue with a continued commitment to Android's "openness". Having to close it right now in order to meet Oracle's demands would undermine Google's arguments in favor of the Motorola deal. It would also add to the concerns of Android device makers.
Theoretically, Google could also try to solve the Oracle problem by switching to a GPL-based codebase that would incorporate code published by Oracle/Sun under the GPL. However, the GPL's "copyleft" -- the requirement to publish all derivative works under that license -- would then apply without any debate. Copyleft may already apply to far greater parts of Android than Google admits, in light of its use of Linux API files that is even more questionable in light of the judge's denial of Google's summary judgment motion against Oracle's API-related copyright infringement claims. But Google will at least try to argue -- for as long as possible -- that Android is Apache-licensed, not GPL-licensed.
Since Oracle is in a fairly strong position in this case and quite likely to win an injunction, which would give it the ultimate leverage to dictate the terms of a settlement including protection against any lawsuits Google might bring later, the judge apparently thought it prudent to create significant additional uncertainty for Oracle with respect to the trial date. Currently, a trial is scheduled for Halloween, but that schedule depends on an unrelated criminal case (unrelated except for scheduling issues on the court's part) and, more imporantly, the judge repeatedly indicated that he might stay the case pending reexamination of Oracle's patents by the patent office unless Oracle reduces the number of claims to be put before the jury to a "triable number".
Again, I understand that the judge felt this additional uncertainty was needed so that Oracle wouldn't feel too strong to settle. I still think this was wrong, and today's mediation talks would be more likely to succeed with a firm, near-term trial date on the horizon. Here's why I think so:
As I explained in my previous post on settlement terms, if Oracle insists on a per-unit royalty, which would be the opposite of unreasonable (in fact, it would be the way most major deals work in the wireless industry), it won't even help if Oracle cuts its demands in half, or even if it goes down to a fifth of its original demands, as long as Google wants to maintain its free-of-license-fee business model. They can't just meet in the middle. In the middle, Google would still have to switch to a commercial license. Even at a level that's a fraction of "the middle".
Consequently, if Oracle insists on a per-unit royalty, it doesn't help to create some uncertainty for Oracle. In fact, based on the median time to trial of the Northern District of California for patent cases, this case is ahead of schedule by a year and Oracle has many options to save time (such as ITC complaints against Android OEMs). Oracle will take the time it needs, unless a deal is too good to refuse. So in a scenario in which Oracle insists on a per-unit royalty, the only chance for a settlement is Google's surrender, not some gradual or even substantial concession on Oracle's part.
From Google's point of view, as long as there isn't a firm trial date, it's very tempting to refuse to settle. If the case gets delayed due to the court's resource constraints, that could be just enough to get the Motorola deal approved by regulators. If the case gets stayed pending reexaminations, the merger question will for sure be decided before this lawsuit goes to trial, and Google could additionally try to start countersuing Oracle. But since Google couldn't get the copyright part of the lawsuit dismissed at the summary judgment stage, a complete stay is not too likely anymore -- maybe a partial stay concerning some of the patents.
If these settlement talks fail, I believe the combination of the factors I described will likely be the reason, but for now let's hope that they can work this out and avoid a trial. If they don't work it out now, there'll be another chance in the build-up to a definitive trial date. Should these talks fail, the judge should confirm the Halloween trial date on a definitive basis and then the parties may still be able to work it out -- maybe just the night before the trial begins. Those things happen.
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: