Wednesday, January 4, 2012

Oracle-Google jury will be informed of status of (at least) three patent reexaminations

A couple of hours after a final pre-trial order that envisions Oracle's lawsuit against Google to go trial "on or after March 19, 2012", Judge Alsup ruled on nine of the ten (five per party) motions in limine. He'll address the remaining one, a Google motion on damages, in a separate order.

Three months ago I already reported on some of those motions in limine.

At the end of today's order on nine motions in limine, the judge makes it clear that any denial of a motion to exclude a particular kind of evidence "does not mean that the evidence at issue in the motion is admitted into evidence — it must still be moved into evidence, subject to other possible objections, at trial". Conversely, "a grant of a motion in limine does not exclude the evidence under any and all circumstances; the beneficiary of a grant may open the door to the disputed evidence, for example".

On balance, those decisions are undoubtedly much more helpful to Oracle than to Google, but there's some light and shadow -- and a great deal of remaining uncertainty -- for either party. Two of Google's three most important motions in limine were denied entirely; two less important ones were granted to a certain degree. Only one Oracle motion was denied entirely, while four of them were granted to at least some extent.

Here's a summary of the decisions, in descending order of importance based on my subjective view:

  1. Oracle sought to exclude any evidence relating to any ongoing reexaminations of patents-in-suit by the USPTO until the reexamination process, including appeals, is closed.

    Judge Alsup draws the line right between first and final Office actions. He notes that "the initiation of reexamination and the customary first office action prove little". Almost six months ago I also said that first Office actions must not be overrated. But Oracle also wanted to exclude final Office actions, arguing that those can still be appealed to the Board of Patent Appeals and Interferences (BPAI) and on to the Federal Circuit (and potentially, if a sufficiently important question comes up, all the way to the Supreme Court). But Judge Alsup finds that once "the examiners have gone to the end of their process", he doesn't want to "conceal this important information from the jury". He attaches particular importance to the fact that those reexaminations considered prior art that was not known to the patent examiner during the original examination process. Judge Alsup doesn't want the jury to be misled on the presumption of validity of USPTO-granted patents.

    This particular decision is clearly more helpful to Google than to Oracle. While Oracle may point out to the jury that all asserted claims of the '520 patent were upheld, Google can show that all asserted claims from the '720 and '476 patents have been rejected thus far.

    The parties can also "request permission to present to the jury some specific item of information from the reexamination of the other patents". Given how those other reexaminations have gone so far, this is also much more of an opportunity for Google than for Oracle. However, I doubt that Google can get too much mileage out of this, given that the judge is reasonably skeptical of first Office actions.

  2. Google has failed for the sixth time (and will probably fail for a seventh time with a petition to the Federal Circuit) to withhold the Lindholm email from the jury.

    I reported on Google's first five failures to exclude this piece of evidence and commented on its long-shot appeal to the CAFC. For the (in my view, very likely) event that the CAFC denies Google's petition, it brought a motion in limine in the district court case to exclude this piece of evidence as misleading (while all other attempts to exclude it where based on claims that it's privileged and confidential).

    Google presented to Judge Alsup a declaration by Tim Lindholm, the author of that important email, denying that he was in the position to analyze whether Android infringe on Oracle's Java-related intellectual property asserted in this particular litigation. But Judge Alsup wasn't swayed and denied the motion. His order cites Mr. Lindholm's background. After joining Google in 2005, he "immediately worked on Android as a 'generalist and interpreter of the engineering/business/legal ecosystem'", helped "negotiate a license for Java", and in Judge Alsup's view a reasonable jury could find that he was "quite knowledgeable about Java and Android technology as separate platforms and any potential crossover between the two platforms". Not only does the judge consider the Lindholm email relevant to the issue of infringement but also the one of damages: "It goes to show that Google had no viable alternatives to Java. It also goes to willfulness because the email was sent after Oracle accused Google of infringement. Since Mr. Lindholm had a deep background in Java and Android technology, the email goes to show that there was an objectively high likelihood that Google's actions constituted infringement of a valid patent."

    Assuming that Google's petition to the Federal Circuit fails (as I expect), this means the Lindholm email will give Oracle powerful leverage. It wasn't certain but it's now clear that Judge Alsup is willing to admit the Lindholm email as evidence not only in connection with damages and willfulness (trial phase three) but also in connection with infringement. For a jury, some of the technical issues will be hard to understand, but the Lindholm email is pretty straightforward and strongly suggests to the jury that there is some infringement somewhere.

    This here is very positive for Oracle. Depending on how the jury reacts to the Lindholm email, this could far outweigh the negative impact of the jury being told about two patent reexaminations that haven't gone well for Oracle so far.

  3. Another Lindholm-related decision grants, up to a certain point, an Oracle motion to exclude any evidence and argument contrary to statements in the Lindholm email. Judge Alsup still thinks it's reasonable for Google to argue "that it had viable alternatives to the patents-in-suit or Java, that it did not need a license for Java generally and for each patent-in-suit, and that not all statements in the Lindholm email were true". But since Mr. Lindholm refused to speak out on certain questions during his deposition "on grounds of attorney-client privilege and work-product doctrine" (a position that is consistent with Google's attack on the admissibility of the Lindholm email as evidence), he can't testify on those questions to the jury. Those questions relate to

    • "what technical alternatives to Java he investigated";

    • "who thought the alternatives 'all sucked'";

    • "what he meant by technical alternatives"

    • "what license terms he had in mind"; and

    • "whether any statements of fact or opinion he made in the email were false".

    Judge Alsup says that it's in "the interest of fairness" not to let him testify on these matters later. Basically, this decision penalizes Google for presumably having instructed Mr. Lindholm to be the opposite of forthcoming in his deposition. It's no secret that Judge Alsup doesn't buy the "attorney-client privilege" and "attorney-client work product" theories at all.

  4. Judge Alsup also granted, to a certain degree, an Oracle motion to preclude Google from arguing that it relied on legal advice in its decision to develop and release Android. Any reference to this will have to be approved by the court. This further increases Oracle's chances of the jury finding Google to have infringed Oracle's intellectual property willfully because it makes it harder, or even impossible, for Google to claim that it relied on legal advice according to which there weren't any infringement issues.

  5. Google failed with a motion to exclude an Oracle analysis of performance benefits Android allegedly owes to the functionality covered by Oracle's asserted patents. Google had complained that the relevant benchmarks were created by Oracle employees. Judge Alsup nevertheless believes that "[t]he tests used benchmarks that were widely accepted by the industry, including Google, as reliable proxies for real-world performance" and suggests Google should have "design[ed] its own performance tests" in order to have "competing results" on the table.

    Apart from the two patents that the jury may consider invalid based on (appealable) reexamination results, I guess those performance tests will bear very significant weight with the jury. They indicate that Google really had a strong technical motivation to implement features Oracle owns, and suggest that the alleged infringement contributed in fairly important ways to Android's success.

  6. Google succeeded in severely restricting Oracle's ability to present evidence obtained from Motorola Mobility: "The jury will be told that the only Motorola device accused of infringement is the Motorola Droid and that other Motorola devices cannot be considered for infringement or damages. Oracle cannot argue that other Motorola devices infringed or that damages should be awarded for other Motorola devices."

    The judge closed a potential loophole for Oracle to bring new infringement contentions after the related deadline.

    This is of limited value to Google. It has no relevance to the most important issue (a possible permanent injunction) and somewhat limited relevance to the question of damages.

  7. The judge denied an Oracle motion to exclude evidence or argument that third-party OEMs changed infringing components of Android.

    While formally denying Oracle's motion and stating that "Oracle must prove direct infringement by OEMs as a predicate for proving indirect infringement by Google" (which is a burden that Oracle would rather avoid), he reminds Google of its claims to be "ignorant on the subject of direct OEM infringement" according to its own discovery responses. Just like in the case of Mr. Lindholm's refusal to answer certain questions earlier on, Google as a whole is now precluded from suddenly claiming to have knowledge it previously denied, except that "Google is not barred, however, from presenting third-party percipient witnesses or third-party documents obtained via trial subpoena from OEMs on the same topic" and it still has the right to "cross examin[e] Oracle's own witnesses that OEMs did use unmodified code".

  8. Google will be allowed to present some arguments concerning Oracle's past actions with third-party application programming interfaces (APIs). Judge Alsup doesn't want Google to go too far back in time and talk about Oracle's use of third-party APIs "long before 2006". Therefore, Eric Schmidt's September 1994 testimony in Congress (he was a Sun executive at the time, and by acquiring Sun, Oracle sort of inherited Sun's API-related practices) isn't admissible as evidence in this litigation. The exclusion also relates to the "American Committee for Interoperable Systems" (an interoperability initiative), "Sun's pre-2006 distribution of Linux APIs" and "Oracle's pre-2006 distribution of the Oracle Database that contained APIs originally developed by IBM". But, fortunately for Google, "testimony and evidence regarding Oracle's and Sun's policies and practices after January 1, 2006, is not excluded even if those same policies and practices began before 2006".

    A rational, non-impressionable jury will not attach too much importance to Oracle's and Sun's own practices concerning third-party APIs. It will understand that each of those APIs has its own technical, commercial and legal characteristics, and consider that even if Oracle and/or Sun did something, they didn't change U.S. copyright law through whatever they might have done. Only a jury that is suspicious of Oracle's intentions could be swayed by this kind of evidence.

  9. Google succeeded in part with a motion to exclude portions of a report discussing commercial success ("Goldberg report"): "Before Dr. Goldberg testifies, Oracle must introduce before the jury sufficient evidence from which it could reasonably conclude that the disputed nexus [between the patents at issue and the commercial success of certain products] exists". Commercial success is a secondary argument that can serve as an indication of an invention not having been obvious at its time.

    I don't know whether Oracle will be able to satisfy the "nexus" criterion, but at any rate, we're talking about a kind of patent validity argument that is not really the most important one.

These decisions pave the way for an interesting trial. The decisions all appear well-reasoned to me, with one exception: I disagree with the judge on the issue of OEMs modifying code. Given that Google's Android licensing terms (which have to be respected by those who license the little green robot trademark and are allowed to distribute the Android Market client and other closed-source apps) impose restrictions, I believe a court should require Google to prove that an officially-licensed devices doesn't infringe, provided that the accused program code is part of the code segments that cannot be modified by OEMs without violating their license agreement.

The problem I have with the decision taken in this case is that Google gets to have its cake and eat it: it gets formal control over a technology, but it also gets to shirk responsibility for infringement. It gets the benefit of the doubt that truly free and open source software would legitimately enjoy but it doesn't deserve it. I don't know if Oracle will appeal this particular decision, but even if it doesn't, this issue will quite probably come up again in future Android-related lawsuits involving Google. I don't think Judge Alsup's conclusion on this one is fair and reasonable, considering how Google's business model works.

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: