Monday, May 12, 2014

Refresher Q&A on Oracle v. Google after appellate ruling: this copyright case is NOT about $1 billion

This is my fourth post since Friday's Federal Circuit ruling that held 7,000 lines of concededly highly-creative Java API declaring code copyrightable. I reacted immediately to the decision and quoted the most important holdings; on Saturday I explained that this ruling was a spectacular comeback but not a surprise or novelty since the structure, sequence and organization (SSO) of computer programs has been copyrightable for almost a quarter-century; and on Google+ I showed that this blog, unlike any other online resource, was spot-on on five of the six most important smartphone IP decisions in the first half of 2014. (On this occasion, I would like to also give credit to Boston-based software and IP lawyer Lee Gesmer, who also discussed the possibility of Google losing this appeal ahead of the appellate hearing.)

Since some of the reporting reflects a fundamental misconception concerning what's at stake (by focusing on a billion-dollar damages claim that Oracle itself did not consider its first priority in the build-up to the 2012 trial) and given that I was one of a very few people not to lose track (out of sight, out of mind) of this case between the fundamentally flawed district court ruling and the appellate opinion, I thought this was the right time to provide a comprehensive, up-to-date Q&A. It is, however, too early to go into complete detail on the retrial. We'll cross the bridge when we get there. I'm happy to talk about it in very general terms already.

  1. Q: Do software developers and users in general, and Android app developers and users in particular, have to worry now that the pendulum has swung in Oracle's direction?
    A: No. No. No. The Federal Circuit has merely stated what has been the law for decades, and only the expressive aspects of Application Programming Interfaces (APIs) are copyrightable, not broader functional ideas. As for the Android-Java situation, we can rely on both companies' ability to resolve this situation in ways that make Android even stronger. Unlike Apple, Oracle is not waging a holy war against Android, but will be a big friend of Android once Google takes a license (the most likely outcome now). Android is the mobile Java platform now.

  2. Q: Is Oracle's copyright stronger than Apple's patents?
    A: Apple wants leverage over Google. Oracle now has it. But that is so only because Oracle's case, unlike Apple's patent assertions against Android, is truly about copying, not assertions of IP against independently-developed program code. In more than four years of Android litigation, Apple has not been able to show even one line of copied program code. Google never saw a need to negotiate a license with Apple before launching Android, but it was negotiating with Sun Microsystems about Java years ago, before it elected to infringe.

  3. Q: Could Google solve the problem by making another programming language, such as Go or Dart, the primary language for Android app development?
    A: It would have to change the API as opposed to replacing the programming language. The keywords and syntax of the Java programming language are free for everyone to use. This case is about the Application Programming Interfaces (APIs).

  4. Q: What are the next steps and how soon will the retrial take place?
    A: Google has already said it's considering its options. Given that Google will almost certainly request a rehearing (and potentially ask the Supreme Court to hear this case), a retrial is unlikely to happen this year, but the most likely year for this is 2015. A settlement may occur well ahead of a retrial, of course.

  5. Q: If and when the retrial takes place, who's going to prevail?
    A: In 2012, the jury was hung on "fair use" (it couldn't agree). After the Federal Circuit ruling, Oracle will be in far better shape than last time for procedural as well as substantive reasons.

  6. Q: Will the retrial be, in part, a billion-dollar damages trial?
    A: Oracle is seeking damages, but that's not the most important part, contrary to widespread misbelief. If Oracle was awarded a negligible monetary amount but an injunction, it would be in a better position to achieve its objectives (a license deal with Google) than if a jury awarded billions of dollars but the court denied an injunction.

  7. Q: What are the parties' business objectives in this dispute?
    A. Oracle once stated its intent to "bring Android back into the Java fold" by making Google comply with the Java rules the rest of the industry has accepted. Google seeks to defend the status quo: the use of Java material in Android without a license.

  8. Q: Does this litigation involve Google's rights to use certain Java code under an open source license?
    A: No. Google could have incorporated Java into Android on open source terms, but chose to eschew the GNU General Public License (GPL) because of its "copyleft" feature.

  9. Q: How do the parties' business models and intellectual property strategies with respect to software platforms compare?
    A: Google's approach to Android has amazing parallels to Oracle's approach to Java. The parties disagree only on how to deal with someone else's IP in a platform.

  10. Q: What ever happened to Oracle's patent assertions against Google?
    A: The district judge obligated Oracle to withdraw most of its patents. Two of the seven asserted patents were taken to trial and not found infringed. Oracle did not pursue those claims on appeal. Theoretically, patent assertions by Oracle against the Android ecosystem are still possible.

  11. Q: Is there a difference between U.S. and European law on software copyrightability?
    A: Copyright laws differ between jurisdictions, but commentators often overstate the scope of the opinion by the Court of Justice of the EU in SAS Institute v. World Programming, as I may explain in detail some other time.

1. Q: Do software developers and users in general, and Android app developers and users in particular, have to worry now that the pendulum has swung in Oracle's direction?
A: No, no, no. The Federal Circuit has merely stated what has been the law for decades, and only the expressive aspects of Application Programming Interfaces (APIs) are copyrightable, not broader functional ideas. As for the Android-Java situation, we can rely on both companies' ability to resolve this situation in ways that make Android even stronger. Unlike Apple, Oracle is not waging a holy war against Android, but will be a big friend of Android once Google takes a license (the most likely outcome now). Android is the mobile Java platform now.

Don't be fooled by the fear, uncertainty and doubt (FUD) strategies of those who are outraged that the Federal Circuit declined their invitation to misread long-standing statutory and Ninth Circuit law. They are now drumming up support for Google's very likely petition for a full-court review and Google's potential petition to the Supreme Court for writ of certiorari. They employed FUD tactics before (during the district court trial and after Oracle appealed), misleading some people to believe that anyone using an API to write apps for a platform would have to worry, which was never the issue. They still haven't come up with a better plan than FUD.

I'm an Android app developer and user myself and will now give you several reasons for which I believe Android will only become stronger if Google partners with Oracle. This doesn't mean to say that we app developers won't have to make some changes to our code if Google makes Android more Java-compatible as a result of all of this. I obviously can't guarantee you that this won't ever be necessary. I'm just saying that Android will continue to do well, and possibly even better, if Oracle becomes a strategic partner of Google's.

First, the Federal Circuit ruling is business as usual. In Johnson Controls, the Ninth Circuit decided back in 1989 that the structure, sequence and organization (SSO) of computer programs is copyrightable.

The sky isn't falling in the eyes of the software industry at large. Most major software companies, some directly and others through BSA | The Software Alliance, supported Oracle through amicus curiae briefs. Those companies have valuable software copyright holdings, but they also implemented other companies' Application Programming Interfaces (APIs) on various occasions. They are sophisticated enough to know that this cuts both ways, and they are comfortable with reasonable protection.

Google and its supporters argued that there was an industry expectation that APIs were not protected by copyright. Not only did the Oracle-supporting amicus briefs contradict that claim but the actual behavior of key players in the cloud computing space did so as well.

Second, some of the FUD (and the resulting panic) is based on inconsistent definitions of the term "API". Neither had Oracle submitted nor has the Federal Circuit ruled that "APIs" in the broadest sense of the term are protected by copyright. APIs have an abstract level, at which we're talking about broad ideas and concepts that can be protected only with patents, if at all. But they also have an expressive level: a creative SSO. Oracle's appeal was not about "APIs" in the broadest sense, but about a concededly highly-creative body of 7,000 lines of declaring code. It doesn't mean that minor, trivial subsets of that body, such as a definition of a function that determines the greater of two values, are copyrightable. Nor does it mean that Oracle's copyright in declaring code would prevent anyone from creating a Write-Once-Run-Anywhere (WORA) platform.

Third (and this is way more important than the third place suggests), if all you have is a hammer, everything looks like a nail, but there's a couple more tools available in this context. While something that isn't copyrightable is obviously free to use, non-copyrightability is not the only way to ensure free access. Fair use is also a powerful concept. I support Google on this one in connection with Google Books, but not its hijacking of Java (and I believe Oracle will prevail on this one on retrial, as I explain further below). In some other cases (for example, if someone just wanted to write programs for an existing platform), fair use would apply and result in free-of-charge access to APIs. That's why fair use is the better vehicle here: it depends on what someone does later with an API, while copyrightability must be determined based on what the original author created (before anyone could know what would be done later).

I'm not finished yet with the extremely important third point. Even if an implementer lost on fair use, it would not be the end of the story: if an antitrust argument of a dominant market position can be made, there is the possibility of a compulsory license on fair, reasonable and non-discriminatory (FRAND) terms. In fact, FRAND is mentioned in the Federal Circuit ruling. Without even a need to invoke antitrust law, Google could have received a license for a clean-room Java implementation on FRAND terms early on if it had decided to be fully compatible and to accept the other terms of such a FRAND license.

Fourth, the first three points were only about the law, but most of the time there won't be a legal controversy (not even outside the courts) in the first place because creators of APIs typically enable and encourage broad adoption since it's in their business interests. It's also in an API creator's business interest to enforce only reasonably. No one wants to scare their current ecosystem or behave in ways that make it hard to create and grow future ecosystems.

If Oracle keeps winning, Oracle and Google should ultimately reach a deal based on rational decisions on both sides. I can't predict their decisions, but these are extremely well-run companies. Oracle obviously has to seek impactful remedies because otherwise Google won't take a license. But it's not on a mission to destroy Android or anything like that. If Apple's patents were as strong as Steve Jobs mistakenly believed, it would definitely want to leverage them to regain market share from Android -- and Steve Jobs himself wanted "thermonuclear war" (according to his official biography) or a "holy war" on Google (according to an internal email). Apple has repeatedly taken the position in litigation that every Android device sold causes major harm to its business. That is not at all the case with Oracle once it has a deal in place with Google. Android is the only major mobile platform centered around Java, unlike iOS and Windows. Oracle apparently wants Google to become a good citizen of the Java ecosystem so as to reunite the Java community, and it wants to be compensated. However, I'm sure Oracle will take great care that Android emerge stronger, not weaker, from such a deal. I don't think Oracle will just maximize its piece of the cake: it will want Android to continue to grow.

2. Q: Is Oracle's copyright stronger than Apple's patents?
A: Apple wants leverage over Google. Oracle now has it. But that is so only because Oracle's case, unlike Apple's patent assertions against Android, is truly about copying, not assertions of IP against independently-developed program code. In more than four years of Android litigation, Apple has not been able to show even one line of copied program code. Google never saw a need to negotiate a license with Apple before launching Android, but it was negotiating with Sun Microsystems about Java years ago, before it elected to infringe.

As a strategic weapon, patents can be much more powerful than copyright. Copyright is narrow in scope. Patents are often not as broad as their owners claim (once there is a serious challenge to their validity, and claim construction by a court of law), but many software patents are indeed far broader than software copyright.

In the specific case of Oracle v. Google, it's not breadth that gives Oracle leverage (and the case has nothing to do with breadth of copyright, contrary to what some suggest). It's simply that Google has built Android in a way that the asserted Java API declaring code is a key building block. It elected to do so on a basis that the Federal Circuit found to constitute infringement (barring the affirmative defense of fair use, which is subject to a retrial). Now there's a million or so apps that make use of APIs Google (but not the Federal Circuit) thought were for the taking.

Look at it this way: the Great Wall of China is a huge building, but its existence is not a threat to your house. That's what software copyright is usually like: it's like real estate that belongs to someone else but that you don't have to worry about. Here, those Java APIs are like one wall of your house. If someone now argues that he owns that wall, then you're in trouble, not because of breadth but because you decided to use it.

Unlike in most software patent cases, the issue here is that the infringement occurred intentionally. Google knew that it was "perhaps making enemies along the way".

Of course, this infringement wouldn't really be an infringement -- or none that matters -- if Google had been right on copyrightability. But the Federal Circuit says it was not right, and after the appellate proceedings Google also faces a considerably steeper challenge with its "fair use" defense. The fact that Google took Java (as opposed to incidentally infringing a broad patent) gives Oracle a higher moral ground than, for example, Apple can claim. I also think Apple has a legitimate basis for enforcement, generally speaking. But Apple says "copying" when it means that others were inspired by the iPhone and independently developed creative imitations of certain features Apple was first to provide.

When Oracle complains about Google's use of Java, we're talking about actual, outright copying. In more than four years of Android litigation, Apple has not discovered a single line of program code stolen from iOS (it never claimed that there was, and due to differences in programming languages, the stolen code would have to be "translated" to other languages, but copyright also covers direct translations). Apple's infringement arguments were all based on abstract descriptions of functionality, not verbatim copying.

Google never negotiated with Apple or any other right holder I could think of before developing technologies for Android that later gave rise to IP infringement complaints. But it was negotiating with Sun Microsystems (which developed Java and was acquired by Oracle in 2010). And then decided to take its chances. That would have been acceptable if it had been in its right to do so. The district court cleared its behavior, and the Federal Circuit has reversed the district court ruling in part and remanded the rest.

In its disputes with Android device makers, Apple likes to argue (and this is not wrong in principle) that it shouldn't have to compete with products that infringe its own IP. Still, those competing products were developed independently. Apple is competing with what may sometimes involve implementations of its own ideas, but then those ideas are not necessarily patentable to the extent Apple believes (because of prior art). In Oracle's case, Java is simply unable to compete with Android, and Android includes what may be the most creative part of the Java codebase (as opposed to just another virtual machine-based platform).

I like Google a lot and find myself in agreement with it on many issues, but the Java thing must be rectified.

3. Q: Could Google solve the problem by making another programming language, such as Go or Dart, the primary language for Android app development?
A: It would have to change the API as opposed to replacing the programming language. The keywords and syntax of the Java programming language are free for everyone to use. This case is about the Application Programming Interfaces (APIs).

Just like the term "API" has divergent definitions, different understandings of what "Java" is (in the context of this dispute) also give rise to misunderstandings. The brevity of headlines and tweets compromises precision (that's also true of my headlines and tweets, of course).

If a tweet or headline says the appeals court allows Oracle to pursue claims over Java or held that Google infringed Java, it has nothing to do with the keywords and syntax of the programming language per se. Theoretically it would be possible to use all Java keywords and the Java syntax without implementing any API. One could start from scratch and write new APIs.

Google doesn't need to replace the Java keywords and syntax. If programming languages like Go or Dart used the same APIs, it would run into the same issues. As I said further above, copyright also covers translations of copyrightable material to other languages.

4. Q: What are the next steps and how soon will the retrial take place?
A: Google has already said it's considering its options. Given that Google will almost certainly request a rehearing (and potentially ask the Supreme Court to hear this case), a retrial is unlikely to happen this year, but the most likely year for this is 2015. A settlement may occur well ahead of a retrial, of course.

There are three things Google can do to pursue the cause of non-copyrightability. It can ask for a rehearing by the same panel of judges, an en banc rehearing (full-court review), or ask the Supreme Court to hear the matter (this is called a petition for writ of certiorari).

Petitions for a rehearing by the Federal Circuit often are requests for a panel rehearing and an en banc at the same time, though only one would be granted at most. Such petitions rarely succeed. In this case, the panel was unanimous. A rehearing by the same panel is a long shot to say the least. For an en banc the panel's unanimity also makes things harder, though not impossible. But it wouldn't be easy to convince the Federal Circuit, which doesn't deal with copyright cases too often, that this matter needs the attention of the entire appeals court. Even if a rehearing of some kind occurred, I doubt Google would get a better outcome.

Asking the Supreme Court for "cert" is something that Google could do, if only to further delay and create uncertainty while negotiating a settlement with Oracle. The Supreme Court does sometimes disagree with the Federal Circuit, but there really isn't any Supreme Court precedent that Google can claim the Federal Circuit decision was inconsistent with.

A company like Google that has to deal with many legal issues, including a number of very important ones, has to choose wisely the ones it wants to ask the Supreme Court to look at. Otherwise it would have to petition for cert every other month. I am absolutely convinced that Google would make a much smarter choice if it decided to play the Supreme Court card in connection with a couple of issues from the "Posner case" (Apple v. Motorola): injunctive relief over standard-essential patents, injunctive relief over non-SEPs, functional claiming, maybe also non-SEP damages. Considering (i) the importance of the issues to Google and its ecosystem (Samsung and other device makers), (ii) the fundamental conflict with Apple (and some other patent holders sharing Apple's related interests who aren't Google's best friends either), (iii) the fact that the Federal Circuit panel was divided in a way that very much invites a further appeal, and (iv) the fact that the Supreme Court may be more interested in, and more willing to support, Judge Posner's take on how to fix patent law (on what would definitely have nationwide impact) than Judge Alsup's wholesale adoption of Google's outlier non-copyrightability theory under Ninth Circuit law, I think the "Posner case" is a hands-down superior choice for a cert petition over the Oracle case, which Google could and should settle instead. Apple will never be part of the Android ecosystem (for as much as I like Steve Wozniak's suggestion that Apple, too, should build Android devices). Oracle will be a strategic partner. So Google should fight Apple and befriend Oracle.

Without a rehearing and a cert petition, I guess the Federal Circuit would issue a mandate to the district court in a matter of months, and the district court would then have to pick up the thread from 2012 and prepare a retrial. All in all, I guess the retrial is most likely to happen in 2015. But a cert petition could delay it (late 2015 or early 2016), and if the Supreme Court actually decided to hear the case, then 2016 becomes the most likely year for the retrial.

5. Q: If and when the retrial takes place, who's going to prevail?
A: In 2012, the jury was hung on "fair use" (it couldn't agree). After the Federal Circuit ruling, Oracle will be in far better shape than last time for procedural as well as substantive reasons.

Given that the retrial won't take place too quickly (though I guess Oracle will do all that it can to speed things up after the mandate issues), it's too early to discuss each and every detail of the "fair use" analysis in light of the appellate opinion. I've taken a clear position before (during the 2012 trial and between the appellate hearing and the appellate opinion): this is not a case of fair use. Hijacking can never be fair use, especially not transformative. I support Google on fair use in the Google Books context (where I was initially skeptical but after taking a closer look came down on its side). But not in connection with Java, where Android displaced mobile Java.

I do realize that many people now want to look at the "fair use" question again, given that it's the sole remaining defense for Google on remand. I've received requests on Twitter, for example.

I will probably do a whole series of posts on the fair use factors between now and the retrial. For the purposes of this Q&A, I am, however, prepared to explain in general terms why I think Oracle is in much better shape now.

I was much more accurate in my coverage of the 2012 trial than some others, who only appeared to have been accurate because the district judge erroneously agreed with them. But I admit that I was absolutely wrong on one thing: I thought the hung jury was due to a minority defending Google. Based on interviews with jurors after the trial, it turned out that it was actually just the jury foreman (who may have been smarter in this context than the rest of the jury) who declined to find in Google's favor. I have a high hit rate but it's not 100% and never will be.

But I'm optimistic I will be right next time and the fair use defense will fail. The Federal Circuit ruling has substantive and procedural implications, all of which will create a better factual and psychological basis for Oracle's efforts to overcome the fair use defense.

In substantive terms, the Federal Circuit made clear that Google stretches the envelope of "transformative" use, and suggested that any interoperability argument in the fair use context may have to be viewed on a package-by-package basis (37 packages are at issue). I will talk more about substantive aspects relevant to the fair use factors closer to the trial.

Google would have liked a retrial (which it ideally wanted to avoid in the first place) to be about infringement and fair use at the same time. But the Federal Circuit has directed the district court to reinstate the infringement findings and conduct further proceedings only with respect to fair use. This means the infringement finding (barring a successful fair use defense) is going to be law of the case. It's a safe assumption that there will be a fight over jury instructions, but those are not going to be great for Google in psychological terms because the law of the case is unfavorable now.

Google has also definitively lost its equitable defenses (it didn't appeal that loss). This means that on remand, Oracle will be able to seek exclusion of certain testimony and argument that was relevant to equitable estoppel and licensing-related defenses but will, arguably, be irrelevant to the narrow question of fair use. The Jonathan Schwartz testimony is one such example. It was really about equitable defenses and willfulness, but chances are that it influenced the jury in connection with fair use nonetheless. A stricter focus on the four fair use factors will benefit Oracle.

Both parties will optimize their fair use-related argument to the jury, and will likely get more time on fair use per se than last time. Knowing that the last jury, even though presumably due to some confusion and to jury instructions that were prejudicial to Oracle, nearly sided with Google, Oracle will presumably put a lot of effort into the fair use retrial.

With a view to further proceedings, whichever party loses the fair use retrial may appeal that one. On appeal they could raise issues with jury instructions, for example. But if Oracle prevails and then seeks an injunction, pressure on Google to take a license will grow.

6. Q: Will the retrial be, in part, a billion-dollar damages trial?
A: Oracle is seeking damages, but that's not the most important part, contrary to widespread misbelief. If Oracle was awarded a negligible monetary amount but an injunction, it would be in a better position to achieve its objectives (a license deal with Google) than if a jury awarded billions of dollars but the court denied an injunction.

I've already mentioned further above that various media reports mischaracterize this as a case that is about whether or not Oracle will get $1 billion. In January 2012, Oracle placed the emphasis more clearly than before (when it was already clear to me) on its push for an injunction. It is also seeking damages, and yes, these could amount to billions of dollars including willfulness enhancements. And if no injunction issued (which I think is hard to imagine here given the irreparable harm that has been caused), there would have to be postjudgment royalties, i.e., ongoing payments, which could also be substantial.

The biggest problem with a lot of the reporting is that it covers the trees instead of the forest. I recently stressed in connection with Apple v. Samsung II that, until the time of that post, mass media reports simply hadn't focused on the real issue in that dispute: workarounds. Apple can win all sorts of rulings, but as long as Samsung's products still look the same (or pretty much the same) to consumers, it won't slow down, much less stop the erosion of its global market share. In Oracle's case, monetary relief is definitely inadequate compensation for what Google did to Java. This is an injunction case much more so than a damages case, and the strategic value of an injunction far outweighs any damages claim the court would ever allow Oracle to present to a jury, even with willfulness enhancements (triple damages).

7. Q: What are the parties' business objectives in this dispute?
A: Oracle once stated its intent to "bring Android back into the Java fold" by making Google comply with the Java rules the rest of the industry has accepted. Google seeks to defend the status quo: the use of Java material in Android without a license.

Between the district court judgment and the appellate hearing, the two companies' CEOs blamed each other for the unresolved situation. In May 2013, Google CEO Larry Page said that a positive relationship wasn't possible because "money is more important to them than any kind of collaboration". Oracle didn't respond to that allegation, but in August 2013, Oracle CEO Larry Ellison reminded Google of its "Don't Be Evil" meme and said that "what they did [using Java in Android without a license] was absolutely evil". Mr. Ellison also called out Mr. Page on his personal responsibility.

There are disputes in which the only question is how much one party owes the other. If Oracle v. Google was one of those disputes that are only about money, I guess there would have been a settlement a long time ago. But this is really about two platforms: Java, which Oracle acquired as part of its $7 billion acquisition of Sun Microsystems, and Android, which lets programmers write apps in Java that generally won't run on Oracle's original Java. In January 2012, Oracle told the district court that it wanted "to bring Android back into the Java fold" (i.e., ensure true compatibility between Android and Java) and end "Google's lawless conduct". That allegation was based on the fact that Google was actually negotiating a Java license with Sun (before it was acquired by Oracle) and decided to go ahead and ship Android without a license. An order by the district court quoted the following October 2005 email by Android founder Andy Rubin:

"If Sun doesn't want to work with us, we have two options: 1) Abandon our work and adopt MSFT CLR VM and C# language - or - 2) Do Java anyway and defend our decision, perhaps making enemies along the way"

Google elected the second option, and by now Oracle has lost most of its Java business opportunity on mobile devices as a result of that. Java-based mobile platforms such as Blackberry and Nokia's Series 60 phones were extremely popular. By now, Android owns more than 80% of the worldwide smartphone market.

There are no signs of Oracle intending to cause harm to Android. Oracle's focus is on what it believes is best for Java, not on what is worst for Android. But Google wants to make its own rules. It's not just about whether Google would be willing to pay royalties (which it's not prepared to). For Google it's also about unrestricted autonomy. It wants to do with Java (and to Java) whatever it pleases. This, to Oracle, is fragmentation, which already had Sun worried back in 2007, shortly after Android was launched. It's of greater concern to Oracle what Google does with and to Java than whether and what it pays for using it (though that is also important).

After the Federal Circuit ruling it's really a risky game for Google to play if it still refuses to settle.

8. Q: Does this litigation involve Google's rights to use certain Java code under an open source license?
A: No. Google could have incorporated Java into Android on open source terms, but chose to eschew the GNU General Public License (GPL) because of its "copyleft" feature.

When Oracle filed this lawsuit in August 2010, it may have looked at first sight like a patent attack by a closed source software company on a piece of open source software. In its answer to the complaint, Google raised some open source issues that were presumably directed at the court of public opinion rather than the court of law. From the beginning Google has been playing the open source card, but it never had a formal defense based on its rights under the GNU General Public License (GPL), the free and open source software (FOSS) license under which Sun Microsystems published Java. In other words, Google put up an open source smokescreen and told an open source tale, but if you use software on open source terms and someone tries to prevent you from doing so, you raise a defense that says, "I am licensed under [whatever open source license] and I'm perfectly entitled to do what I'm doing!"

If Google had ever claimed in this litigation that Android's incorporation of Java was authorized under the GPL, this would have been inconsistent with its claim that it didn't use any copyrightable material (you can only license something, on whatever terms, if it's protected) and, far more importantly, a blatant violation of Rule 11 (truthful pleading standard) resulting in sanctions for Google and its counsel. The GPL affords you four freedoms: it allows you to use software without paying for it (freedom 0), to modify its source code as you please (freedom 1), to redistribute copies (freedom 2), and to distribute your modified versions to others (freedom 3). But once you exercise freedom 3, you're subjected to the copyleft rule: you must make your modified version available under the GPL. As a result, Google would have had to publish Android under the GPL. But it did not. Android builds on top of Linux, which is GPL'd, but Android itself is, according to Google, Apache-licensed. The Apache Software License 2.0 is a non-copyleft license and, therefore, absolutely incompatible with the GPL. You're lucky if you can run GPL and Apache software on top of each other (which can also raise complicated copyright and copyleft issues), but you certainly can't take GPL'd software, such as Linux, and release it under the Apache license.

As I'll discuss in more detail in the next section, Google seeks to exercise as much control over Android as possible. If Android had to be released under the GPL (and only under the GPL, or at least a compatible copyleft license), Google would have a hard time keeping a growing number of core Android components like its Mail and Map clients or even the new on-screen keyboard (again, more about that in the next section) closed -- and Google's hardware partners would have to release their proprietary enhancements such as Samsung's Touchwiz and HTC Sense under the GPL, which would run counter to their objective of differentiation because their competitors could then use the same code.

So there is no open source license-based defense. The closest thing to it that Google raised related to only a limited amount of code that it obtained under the Apache license. But that is also irrelevant. If I take software that I don't own and put it under an open source license without authorization, this still doesn't deprive the actual right holder of anything and, especially, doesn't allow third parties to use that software without a license from the actual right holder. Those third parties may, however, point to their reliance on what appeared to be an open source license in connection with the question of willfulness. Google never made that argument in connection with the program code that's really at issue (many thousands of lines of Java API declaring code): it obtained that one directly from Oracle's (Sun's) codebase, not indirectly through other parties or third-party open source projects.

9. Q: How do the parties' business models and intellectual property strategies with respect to software platforms compare?
A: Google's approach to Android has amazing parallels to Oracle's approach to Java. The parties disagree only on how to deal with someone else's IP in a platform.

It would be mistaken to believe that Google's Android business model is more permissive or more generous than Oracle's approach to Java. Let's face it: these two companies wouldn't be as wildly successful as they are if they acted like charities. Oracle acquired and has further invested in Java to make money with it; Google acquired and has further invested in Android to make money with it, too.

There are some differences, but there are also striking parallels. At the structural level (not getting into detail on how things are run), the Open Handset Alliance (OHA) is to Android what the Java Community Process (JCP) is to Java. Both Android and Java are available under open source licenses (Apache and GPL) as well as proprietary licenses. There are neat things that can be done and indeed have been done on an open source basis -- but both companies know that those seeking to make money with Android or Java will, not in all but in most cases, prefer a proprietary license.

Both companies also have in common that they fight fragmentation of their respective platforms.

The claim that Android is "free" and "open" is no longer taken seriously by anyone in this industry. If you're interested in a thorough, well-written debunking of that claim, I recommend this Ars Technica article entitled "Google's iron grip on Android: Controlling open source by any means necessary" (also, note what it says right below this headline: "Android is open--except for all the good parts"). Not even the current Android on-screen keyboard is open source...

If things had worked out differently between these two companies and they had developed Android together (or if they had co-developed Java, but it goes back to before Google was even founded), there wouldn't be a fundamental disagreement over what's desirable or undesirable with respect to a jointly-owned platform. However, Oracle has a tradition of respecting other companies' intellectual property rights (I can't remember any major IP infringement action against Oracle itself, obviously excluding what some acquisition target may have done prior to a takeover), while Google faces IP infringement allegations by major companies and organizations all the time. Google is defending itself very well so far, and that's positive for Android. But I sometimes feel it really likes to stretch the IP envelope...

In short, they don't disagree much on what each company wants to achieve with respect to its own platform, but they have a problem because of what Google wants to do with and to Oracle's platform.

10. Q: What ever happened to Oracle's patent assertions against Google?
A: The district judge obligated Oracle to withdraw most of its patents. Two of the seven asserted patents were taken to trial and not found infringed. Oracle did not pursue those claims on appeal. Theoretically, patent assertions by Oracle against the Android ecosystem are still possible.

Oracle originally asserted seven patents against Google, and at the early stages of this litigation it looked like more of a patent than copyright case simply because each patent was one "count" of the complaint, while copyright as a whole was only one more count. But item counting is never the proper methodology for establishing the relevance of issues in a case.

Oracle's patents-in-suit came under reexamination pressure. I must admit that I've lost track of those reexamination proceedings, but first Office actions are of rather limited relevance and even "final" Office actions tentatively "rejecting" a patent aren't truly final (as the examples of certain Apple patents such as the '381 rubber-banding patent -- key claims ultimately confirmed -- and the "Steve Jobs patent" -- all claims ultimately confirmed -- show). But Judge Alsup leveraged the state of affairs in those reexamination proceedings against Oracle, urging it to narrow its case for the jury lest the case would be stayed pending reexamination (i.e., for years).

In January 2012, Oracle proposed that the patent part of the case be severed and stayed in favor of a near-term copyright trial. This was the first public filing by Oracle in this entire litigation that demonstrated an unequivocal focus on copyright infringement issues.

Judge Alsup didn't adopt this proposal because he didn't want to have more than one Oracle v. Google trial. Oracle then had to withdraw most of its patents -- even with prejudice, which, by comparison, another (patentee-friendlier) federal judge in the same district, Judge Lucy Koh, did not require Apple and Samsung to do (they only had to withdraw patents without prejudice, allowing them to reassert them later). Ultimately, two patents were shown to the jury, which didn't find them infringed, though Google's own source code comments and file names clearly weighed against a finding of non-infringement with respect to one of them. The judge didn't overrule the jury, and Oracle is not pursuing those patents on appeal (one of them has expired and the other one was of relatively low commercial value).

Oracle's withdrawal with prejudice of five patents and abandonment of two other patents on appeal relates to Google, not to other parties, and even with respect to Google only to the patents-in-suit. Should any of those patents be confirmed at the end of reexamination (including appeals), Oracle would not be able to sue Google over them, but it could (and I have no idea whether it would want to do that) still sue Android device makers (even before the end of reexamination). It could also sue Google over other patents. Without speculating on the likelihood of this happening, I believe the successful appeal of the non-copyrightability holding makes copyright, not patents, an obvious priority for the further proceedings. I just wanted to point out that Oracle's preclusion with respect to patent assertions against Android isn't unlimited. That is another reason for which Google should settle.

11. Q: Is there a difference between U.S. and European law on software copyrightability?
A: Copyright laws differ between jurisdictions, but commentators often overstate the scope of the opinion by the Court of Justice of the EU in SAS Institute v. World Programming, as I may explain in detail some other time.

The district court appeared to take some potential interest in the European SAS Institute case, but its ruling has been reversed. This is not the time and place to go into details on SAS Institute. Suffice it to say that what it said was blown out of proportion by many people. Maybe I'll talk about it some more on some other occasion.

But Oracle v. Google could have worldwide effects. While the Federal Circuit opinion in Oracle v. Google isn't formal precedent for other jurisdictions, it will be interesting to see how judges outside the U.S. view software copyrightability. In principle, it would make sense in other jurisdictions as well to focus in a copyrightability analysis on how creative the original author was as opposed to whether concededly highly-creative declaring code later serves an interoperability purpose. There are similar concepts as fair use in other jurisdictions. There are compulsory licensing options in all jurisdictions to remedy extreme cases with the help of antitrust law. The Federal Circuit opinion may inspire non-U.S. courts as well to decline invitations to adopt "if all you have is a hammer" type positions and to address interoperability on a case-by-case basis (based on exactly how someone makes use of interoperability-related code).

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: