Friday, May 6, 2011

Oracle strongly opposes judge's proposal for narrowing down claims against Google to 3, instead proposes 21 plus postponement of related decision

Yesterday many media reported on a tentative scheduling order of the judge on the Oracle vs. Google case, according to which Oracle would have to ultimately narrow down the number of patent claims asserted against Google from 132 to 3 in order to arrive at a "triable number" of claims, but most of the reports didn't make it sufficiently clear that it was just a proposal, not a procedural decision per se.

Today, Oracle and Google filed their comments on the tentative order. Oracle strongly opposes the judge's proposal and wants to assert three claims per each of the asserted seven patents, i.e., a total of 21 claims. "Just as important, the ultimate decision on the number of claims should be made closer to trial", Oracle's lawyers write -- and I totally agree with them on this, for reasons I'll explain further below. Don't get this wrong: I'm against software patents, including every single one of the seven patents Oracle is asserting against Google. But when it comes to matters of justice and due process, personal preferences and political positions don't matter.

In my opinion, Oracle makes a very compelling legal case for a larger number of patent claims to go on trial. It's always hard to predict the outcome even of subsets of a wider issue, but I believe that most likely we will see a compromise that will allow Oracle to continue to assert at least one claim for at least six of the seven patents-in-suit, and maybe an average of one and a half or two claims per patent.

In my opinion, Oracle's case for a larger number of claims is legally strong because the judge's initial proposal was too aggressive and does not appear to be consistent with applicable case law by the Court of Appeals for the Federal Circuit, which allows impositions such as the ones proposed in Oracle vs. Google but only if the asserted claims are duplicative. I don't see that three claims is the correct number considering the very distinct nature of at least six of the seven patents-in-suit. The judge's proposal does not balance case management and the interest of justice as well as Oracle's counterproposal. Throwing out duplicative claims before trial is acceptable; throwing out unique and distinct ones would unfairly disadvantage Oracle.

It's possible that the judge basically just tried to take an extreme position for "bargaining" purposes. If that was his strategy, he's already achieved a lot in terms of Oracle signaling a willingness to go down from 132 patent claims to 21 (subject to reevaluation at a later stage of the process).

I will now sum up the gist of Google's critique of the judge's proposal, and then I'll discuss Oracle's position and explain in more detail why I think it's legally and factually right.

Google's positions

  • It comes as no surprise that Google "agrees with the Court’s three-step process [for reducing the number of claims] in its entirety." Google would obviously benefit from a drastic reduction of the number of claims asserted by Oracle.

  • What Google mostly focuses on in its filing is the judge's request for comments on the possibility of delaying trial until after reexamination of (at least a couple of) Oracle's patents by the USPTO. That would stall the case for years, and Google strongly advocates this.

  • In the final part of its filing, Google notes that the judge has not yet proposed a schedule for narrowing Oracle's copyright infringement claims, and Google's position is that this can wait because Google still states (which may however be wishful thinking) that those claims can be thrown out on summary judgment. The judge previously disagreed with Google on that, but discovery has progressed.

  • I noticed that Greenberg Traurig attorney Heather Meeker joined Google's team. She's known for her work for the Mozilla Foundation. But she's also notorious for fervently advocating the patentability of software as a more suitable intellectual property right for this purpose than copyright, as I mentioned in this recent post. If Google opposed software patents, there's no way they would work together on this particular case with a lawyer who loves software patents...

Oracle's positions

  • Oracle admits that "[t]he extent to which -- several years from now -- the final results in the reexamination [of some or all of its asserted patents by the USPTO] could moot the need for a trial here on those patents depends, of course, on the outcome in reexamination."

    But Oracle points out that five of the patents are subject to ex parte reexamination, which according to "current PTO statistics" cited by Oracle takes 31.8 months on average. Google's representations were not inconsistent with this, and Google wrote that inter partes reexaminations (which have been initiated concerning two of the patents) take about six months longer on average. Those are two different types of reexamination procedures; both are initiated by parties filing requests, but in ex parte reexaminations, they don't participate going forward, which they do only in inter partes reexaminations.

    The bottom line is that waiting for the outcome of all reexaminations would delay the process by about three years, which Oracle would obviously like to avoid.

    However, Oracle does not restate in detail its opposition to a stay. Oracle's position is known, and Oracle would probably comment on this in more detail if the judge showed a stronger inclination toward such a stay than just asking the parties for comment (which I interpreted as thinly-veiled threat that the judge might take such a decision should Oracle refuse to accept a drastic reduction of the number of asserted claims).

  • Oracle's focus is on its opposition to the "tentative schedule" for narrowing claims. Their lawyers state that the court's proposed schedule "will severely prejudice Oracle", and they explain three reasons for this plus issue a warning to the court in terms of what this would mean for the further process:

    • Oracle wants to assert all of its patents in order to maximize its damage award. Oracle points out that "[t]he intellectual property that Google has infringed is technically and economically core to Android; Android is a multi-billion dollar proposition of core strategic significance to Google; and Android fundamentally undermines Oracle's Java technology." Oracle absolutely rejects Google's argument that "Oracle's patent claims do not read on a significant enough portion of Android to warrant a substantial recovery." Oracle still wants to prove that "given Android's enormous present and future contribution of revenues to Google (through advertising and other commercial benefits attributable to Android), Oracle is entitled to a substantial recovery."

      Oracle voices concern that if it can only assert a subset of its patents-in-suit, "Google may contend that Android could have been equally successful without any one, or two, or even three of the claimed inventions and hence that Oracle is entitled only to very limited patent damages."

      In Oracle's opinion, it would be unfair for them to receive a lower damage award than they'd be entitled to if the court made the effort to decide on all seven patents. Theoretically, any single patent could result in the substantial award Oracle seeks (I guess they are aiming for something north of a billion dollars, similar to what they received from SAP in a different context), but "under the Court’s proposed limits it will likely be more difficult to persuade a jury that a substantial portion of Android's value and the harm to Oracle is due to infringement."

    • Oracle wants to hold Google liable for the full extent of its alleged infringement, and if it can't assert all of its patents, Oracle is afraid that Google would be provided "with a greater likelihood of prevailing on liability."

      This is obvious from a statistical point of view. Each asserted patent is a bite at the apple. Under the judge's proposal, Oracle could at most assert three different patents. If Google had (only for the sake of the argument) a 70% chance to fend off one patent, its likelihood to fend off three would be 34%, but the probability of fending off all seven would be only 8%.

      The legal argument Oracle puts forward is that "[t]he seven [patents-in-suit] each claim materially different inventions." There are two patents which Oracle admits "share specifications", but even those are claimed by Oracle to "claims substantially different inventions". And none of the other five patents shares (according to Oracle) specifications, filing dates, claims to priority or inventors with any other asserted patent.

      This is all about whether or not Oracle's claims are duplicative. I'll explain the significance of that further below. It is an important point, and the judge's proposed order did not provide any reasoning at all as to why Oracle's claims had the kind of duplicity that would justify the drastic razor-style reduction he proposed, which would require Oracle to drop more than half of its asserted patents without any chance for subsequent assertion in a new trial in connection with current versions of Android (at best, Oracle could claim that new versions represent new infringement issues).

    • Oracle's third argument against the proposed reduction is that this would "deny Oracle due process." Oracle's argument here is mostly about a recent decision by the Court of Appeals for the Federal Circuit in a case in which the judge also ordered a narrowing of the asserted claims. I'll talk about that case further below.

      In this context, Oracle complains that the court would only limit the number of Google's prior art defenses (to eight) but not impose any limit on Google's "other invalidity defenses" (such as obviousness, statutory subject matter etc.). Oracle believes that this would be unfair because Google could "maintain a highly elastic defense strategy and present multiple invalidity defenses against Oracle's excessively narrowed claim set target."

      I agree that there's some asymmetry in this, and I think the court may very well conclude that the number of defenses raised by Google will have to be limited in general, not just in connection with prior art.

    • After the three reasons for which Oracle sees itself potentially prejudiced by the court's proposed schedule, Oracle also issues a warning to the court that the proposed limits "will place considerable pressure on summary judgment." Oracle says that if it can only take three patent claims to trial, it will have to file more summary judgment motions because "[o]nly through that vehicle will Oracle have even a chance to protect its case on liability and damages for all seven patents." Oracle speculates that Google might "face similar pressures to present its invalidity defenses on summary judgment as well."

      Oracle warns the court that "[t]hese will be technical motions with technical evidence in the form of expert
      declarations and deposition testimony. They will be as brief as the content allows. But they will require time to resolve."

      Oracle clearly worded that part of its statement so as not to appear to threaten with anything that would make the judge's life harder. Oracle presents this as an inevitable consequence of the court's proposal and advocates a balance between issues that are decided on summary judgment and issues that go to trial. Oracle says that "[they] raise this issue now to prevent surprising the Court regarding the scope of these motions."

      While Oracle presents this in a logical and reasonable form, this still looks a little bit like countering the judge's implicit threat of possibly staying the entire case pending reexamination -- those are separate issues, but the approach is the same: outlining undesirable things that might happen in the event that no amicable agreement is reached.

Federal Circuit case law: In re Katz Interactive Call Processing Patent Litigation

A couple of months ago -- on February 18, 2011 -- the Court of Appeals for the Federal Circuit (to which the Oracle vs. Google could also be appealed if one of the parties disagrees with whatever decision the US District Court for the Northern District of California takes) handed its decision in the matter mentioned above.

The federal judge in charge of Oracle vs. Google based his tentative order on that case, but the way I see it, Oracle's position on how to properly apply that case law to the case at hand is legally correct and the tentative order proposed by the judge overshoots the mark by far.

In the Katz case, which was explained very well by law.com, the CAFC found that the lower court (in that case, the US District Court for the Central District of California) did not appear to unfairly prejudice the patent holder by "the allocation of burdens in the claim selection procedure [...] by creating a significant risk that Katz would be erroneously deprived of property rights in unselected claims."

I also heard from a Texas-based patent attorney today that the kind of reduction proposed in Oracle vs. Google is "not unusual". But the question is whether the judge's decision that Oracle should assert only three patent claims is legally the right number, or arbitrary.

I believe the judge's tentative order would not apply the Federal Circuit decision on Katz correctly. This is the key sentence in the ruling that plays in Oracle's favor:

"Based on [the lower court's] initial determination that the asserted patents contained many duplicative claims, it was both efficient and fair to require Katz to identify those unasserted claims that, in Katz's view, raised separate legal issues from those raised by the asserted claims."

In Oracle vs. Google, there's no initial determination on duplicity in place. The judge simply proposes to order a reduction to three claims. If the judge cannot make a case for duplicity, the decision could probably be successfully appealed by Oracle. I assume that Judge Alsup will now want to determine a number of claims that will not even give rise to an appeal, and that would be defensible in the event that Oracle appeals.

Oracle's non-duplicity argument is strong, and to me it's the primary legal reason for which we will most probably see a new court order that will increase the number. The judge may furthermore be willing to agree with Oracle that the exact number should be determined at a later stage (when the court as well as the parties will have more knowledge of the relevant facts than at this stage). That would also make sense from a due process point of view, and in light of the fact that Oracle clearly is willing to go down a whole lot from the original number of 132 asserted claims.

I will report on the final scheduling order when it's issued. Theoretically that could still happen today, but more likely next week.

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.

Share with other professionals via LinkedIn: