Friday, December 6, 2013

Detailed analysis of Federal Circuit hearing in Oracle v. Google: copyrightability is certain

This week's -- actually, this year's -- most important development on the smartphone IP front was the reversal of fortunes in the Oracle v. Google Android-Java copyright litigation. Independent reporters who attended the hearing agreed that Oracle was now on the winning track and last year's district court ruling was going to be reversed.

The United States Court of Appeals for the Federal Circuit published the official recording of the hearing very late on Wednesday. I wanted to take my time to listen to it more than once in order to fully understand what positions the judges indicated -- and I had to attend a full-day German patent validity trial yesterday. Now I've had the opportunity to listen and to reflect. And if you saw any of the unjustified criticism I was exposed to after last year's erroneous non-copyrightability decision, you'll believe me that I'm glad that reason and logic have prevailed over confusion and irrationality.

Fortunately for creatives and honest innovators, but unfortunately for Google and its supporters, this matter is being adjudicated now by the CAFC, not the EFF or CCIA (the latter engaged in blatant spin-doctoring in its report on the Wednesday hearing, published on a "project" website).

I'm now going to focus on substance rather than procedures. For the procedural context and implications of this appeal, let me refer you to my pre-hearing Q&A.

Copyrightability will be decided in Oracle's favor: Google's reasoning, adopted by Judge Alsup, has collapsed completely

All three judges on the panel took positions that indicate a reversal of Judge Alsup's clearly erroneous non-copyrightability ruling. There wasn't the slightest indication of even just a partial disagreement between the judges.

At the beginning of the hearing, one judge (Circuit Judge Taranto, as I learned from one of the reports) interrupted Oracle's lead counsel, Joshua Rosenkranz, to ask several questions about the infringing activity. In particular, he wanted to understand how Google's app developers use the hijacked Java APIs and how and where they obtain them. He said that these were "almost non-legal questions about how this works". None of those questions suggested doubts about the copyrightability of the asserted declaring code.

In my commentary on Google's first brief on appeal (which was its answering brief to Oracle's appeal and, at the same time, its opening brief for its own, purely tactical cross-appeal) I listed three ways in which the court would have to disagree in order for the finding of non-copyrightability to be overturned:

  1. Google raised interoperability arguments in the copyrightability context, but the circuit judges made clear that they are going to relegate compatibility and similar issues to fair use:

    "I confess upfront that I am not an expert in copyright law. We don't deal with it quite that often. So help me understand. I would have thought fair use was a relevant issue on infringement. [...] I would have thought fair use relates to the question of whether a particular use infringed an existing copyright because fair use is sort of an exception to the copyright coverage rather than [...] the question of copyrightability. As I read the trial judge's opinion I'm confused about how he viewed fair use because it seems to me he was merging it with the question of copyrightability. [...] Is that a significant confusion or is it just my confusion?" (Judge Prager, at 23:34)

    "I specifically agree with you that the district court imported fair use principles into the upfront portion of the analysis, which was wrong." (Judge O'Malley at 26:00)

    "So why shouldn't we relegate those discussions -- I'm not saying they're not relevant factors -- why shouldn't they be considered in the context of fair use rather than crammed into the copyrightability [analysis]" (Judge O'Malley, repeated at 46:36, initially stated in shorter form at 46:04)

  2. Google argued that the use of the Java API structure was "dictated", but the correct point in time at which to determine copyrightability is when an author creates something (i.e., whether the original author was restricted in their choice).

    "Could you not have [...] written your own declaring code. You could have done that. [...] That wouldn't have been easy for other Java users, but that doesn't mean that it wouldn't have had the same exact function?" (Judge O'Malley, at 35:05)

    "So how could that possibly be a method of operation if, as the district court found, at the time [...] there were effectively infinite numbers of ways in which it [Sun] could have written that expression" (Judge Prager, at 39:09)

    "What is the expectation standard? I mean [...] that's Judge Taranto's point. You're supposed to make the determination of copyrightability at the time that the creative act occurs, correct? Does copyrightability get lost because it becomes popular? I mean is [...] no longer copyrightable [...] not copyrightable anymore because people like it?" (Judge O'Malley at 53:42)

  3. Judge Alsup and Google relied heavily on Sega v. Accolade, Sony v. Connectix and Lotus v. Borland in the copyrightability context though these are fair use cases. The Federal Circuit made clear, as I already quoted two reporters in my first reaction to the hearing, that Sega and Sony are inapposite in the copyrightability context. If at all, Lotus may support Google to some extent, but it was a controversial ruling that was affirmed only by an equally-divided Supreme Court -- and even Google's counsel conceded a difference in terms of quality and quantity between the material at issue in Lotus vs. the Android-Java case.

    (in response to citations to Sega and Sony) "For fair use, obviously [...] you have to look at when the use occurs. But for copyrightability and determining whether something is expression or whether a claim of expression extends to something..." (Judge Taranto at 54:40)

    "[When we get to fair use], you can cite Sega and Sony all you want." (Judge O'Malley at 56:08)

    I'm not 100% sure based on the recording, but it could be that Judge O'Malley even told Google's counsel at some point to simply "stop citing Sega and Sony".

The moral of the story: you can fool some people for all time, all people for some time, but not all people for all time

Google's counsel, Robert van Nest, did a fantastic job before the trial court in terms of confusing Judge Alsup and confusing the jury (except for its smart foreman). He had no other choice than trying to muddy the water because his client had a weak case from the beginning. It worked once, but not twice.

It's a fundamentally different challenge to confuse one judge (and his advisers) than to confuse an entire panel of three high-level, high-power judges. One person can be driven by emotion, and can be misled because there's no peer around to help out. But a panel analyzing a matter collectively is far more likely to identify and see through smokescreens. And at an appeals court, there's no jury of laypeople around either.

In the previous section I discussed the three key reasons for which Judge Alsup's holding of non-copyrightability will be reversed. Judge Alsup had adopted Google's position, though it was fundamentally flawed. Then Google, obviously, determined that its best shot on appeal was to defend Judge Alsup all the way in that regard -- except for its tactical cross-appeal. It placed a bet on deference. It didn't have a Plan B: an alternative legal basis on which the appeals court could have disagreed with Judge Alsup but still ruled in favor of Google. So Google was stuck with the whole idea that Sony and Sega established an interoperability exception from copyrightability. And that's simply not in the law. It's not in the statute. It's not in the case law unless one misreads Sony and Sega.

Google failed to make a consistent and logical proposal for line drawing between copyrightable and non-copyrightable computer program code

In my post on Oracle's reply brief (Oracle to appeals court: Google concedes away the entire case under established copyright principles), I discussed the importance of line drawing and talked about how Google and its supporters who submitted amicus curiae briefs failed to really make a clear and workable proposal for distinguishing between copyrightable and non-copyrightable computer program code. This shortcoming of Google's (and its supporters') position was on full display at the Wednesday hearing.

Several of the judges' questions related to whether a particular theory would be overreaching in terms of denying copyrightability to all computer programs, or at least to key parts of all computer programs -- or, conversely, would render the carve-out of methods of operation meaningless. The exception for methods of operation is key, and Oracle had a reasonable answer that appeared to make sense to the court, while Google contradicted itself during the hearing.

The court noted that "almost all computer code must have functional purpose", and one judge asked whether it isn't "sort of difficult to try to find computer code that doesn't have some functionality to it". Oracle agreed, and it clarified that it only claims copyright in its own specific way of naming each of the 362 routines in the Java security API and the way to organize them ("unique and original organization"), but that it cannot use copyright to monopolize the idea of organizing security functions of a computer and prevent others, such as Apple or Microsoft, from doing this. Oracle is not doing what the plaintiff in Baker v. Selden did: he wrote a book on an accounting method and later claimed a monopoly, based on copyright, over that method. But Oracle does own the copyright in its own specific selection of function names and their organization -- and that's what Google "stole", or "plagiarized", as Oracle told the court.

Google did not have a convincing answer to the line-drawing question. In the context of Oracle's main appeal, Google's counsel stressed repeatedly that Google wrote 15 million lines of implementing code and said they're "not claiming here that it's not copyrightable" -- but Google's cross-appeal relates to implementing code that Google claims isn't copyrightable either.

The Federal Circuit is clearly unwilling to decide this case in a way that would call into question the copyrightability of computer program code just because it has, besides its expressive quality, a functional purpose. And interoperability/compatibility is a functional purpose, for example.

Retrial on fair use is possible, but may not be necessary

Reporters felt that the Federal Circuit was hesitant to decide fair use as a matter of law in Oracle's favor. In that event, a reversal of Judge Alsup's non-copyrightability finding would require a redetermination of fair use by the district court, i.e., a retrial, unless the district court determined at this stage that it could decide this question as a matter of law.

Having listened to the recording more than once, I think there's a decent chance that fair use will be decided by the Federal Circuit, though it does seem a bit more likely that fair use will be resolved in district court on remand.

At the beginning of the fair use discussion, the judges appeared somewhat skeptical of a judgment as a matter of law (JMOL) on fair use. They didn't disagree with Oracle's arguments against Google's fair use defense in any way. But they felt that there could be questions of fact left to be determined by a jury. One judge said that it would be easier for the court to decide based on a verdict -- but there was a full trial, as Oracle's counsel noted.

The way I understand the recording, Oracle's counsel did make some headway with respect to a request for JMOL on fair use (or, in the alternative, at least some clear instructions to the district court, especially with respect to transformative use and market impact). Then, toward the end of the fair use discussion, it again seemed like a remand was more likely than a JMOL -- but a JMOL is now still more likely than it appeared to have been prior to the hearing.

When the ruling comes down, which will take a few months, the decision on fair use will be the part to look at because copyrightability, barring a major surprise, will certainly be decided in Oracle's favor and presumably unanimously. On fair use a remand is possible, and then the question is whether the Federal Circuit opinion will make a finding against fair use more likely. On fair use there could also be a dissenting opinion by one judge.

Most commentators got things wrong after the district court ruling

It's difficult to argue with success, and after Google obtained a favorable district court ruling, it looked like Oracle had lost the case. After the Wednesday hearing it's clear that Oracle is on the winning track again, at least with respect to copyrightability, the most strategic issue here. And I believe it should also prevail on fair use, one way or the other.

In a September 2012 ruling, Judge Alsup wrote that "no commentary has in any way influenced the Court's orders and ruling herein save and except for any treatise or article expressly cited in an order or ruling". It's unfortunate that he and his clerks weren't influenced by this blog. If they had been, the Federal Circuit would now agree with them.

The Federal Circuit opinion will probably contain some harsh criticism of the shortcomings of Judge Alsup's non-copyrightability ruling, but I encourage commentators not to go from one extreme to another. After last year's trial, some people were impressed with his heavy-handed case management and a number of vocal activists (a minority, but as I said, vocal) were overjoyous because of what he determined. There was an echo chamber, and what some people said and wrote about Judge Alsup was almost an apotheosis -- which the judge himself presumably never even wanted. I disagreed with him, but respectfully. That's why no matter what the Federal Circuit opinion will say, I will continue to respect him a great deal. This was a difficult matter, and Google misled him, manipulated him, confused him, and used him for its purposes. Same with the jury, by the way. Except for its foreman.

The law is not an exact science. I have held pro-copyright beliefs for a long time. This is not a question of right or wrong in a mathematical sense.

One journalist described Oracle's case as a "dead lawsuit walking". It's alive and well after the Wednesday hearing, however.

One blog that is not active at the moment repeatedly told people that the outcome of the Oracle case was fatal for my credibility (as did one Google employee, a programmer who previously worked for Sun, on his personal blog). Quite the opposite is the case. Knowledgeable, reasonable people always knew that there's a difference between factual reporting and accurate analysis on the one hand and accurate predictions on the other hand -- and personal opinions aren't necessarily predictions, by the way. Nor does someone who reports on a multi-billion dollar damages claim and accurately notes that this is the risk to the defendant necessarily predict that this would be the final outcome.

Just to be clear: a hearing is not a ruling. But the Federal Circuit expressed clear positions, some of which I quoted in this post. And that fact alone -- even if there was a totally unforeseeable about-face between now and the actual decision -- proves that Oracle and those who, like me, shared its position on copyright had a reasonable basis for their beliefs.

Also, a programmer whom I really admire regardless of occasional nonsensical comments on intellectual property issues, Linus Torvalds, wrote the following Google+ post last year:

"Prediction: instead of Oracle coming out and admitting they were morons about their idiotic suit against Android, they'll come out posturing and talk about how they'll be vindicated, and pay lawyers to take it to the next level of idiocy.

Sometimes I really wish I wasn't always right. It's a curse, I tell you."

No need to worry about that one, Linus. Listen to the recording of the Wednesday hearing and you'll see that you weren't right. Granted, you were right that Oracle was going to appeal, and that was an easy one that doesn't count. But you were wrong in terms of "morons" and "idiocy". Contrary to your prediction, Oracle did the right thing.

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: