This multi-year legal battle seem like insanity to me.
What does Oracle, who practically prints money, expect to get out of this other than exercising a fleet of expensive lawyers? At best the case was always iffy and even if they "win" it's unclear what that really means for them given the inevitable subsequent legal maneuvers. Even more puzzling is how are they making the argument that Google has, in any way, harmed their business?
It's just mind boggling. A slip-and-fall insurance scam makes more rational sense to me.
It's business admin 101, really. The unfair advantage IBM/Oracle/SAP et al have, is that they are entrenched.
It is infinitely cheaper to develop things in-house or with smaller software shops, But then you won't be able to interface with your factory's material providers' system, or other Banks, etc because they only talk some protocol that is copyrighted/trade secret of those firms.
They are not fighting google. Google just happened to enter in the cross fire and have deep pockets to fire back! What oracle (and others like IBM behind the scenes, via DC Federal Circuit "lobbying") are fighting back, is the freedom of smaller software shops to provide their clients with fully compatible offerings to their money cows.
This is the only time someone conscious of freedom and privacy should be rooting for Google.
" is the freedom of smaller software shops to provide their clients with fully compatible offerings to their money cows."
Thanks for this summary! I had no idea this was the underlying reason. I feel silly now because I thought it was just some weird obsession of someone in Oracle's C-suite.
Oracle doesn’t actually print money. They are cash-rich but are struggling to convert long-suffering customers from on-prem products to their cloud equivalents. Nobody trusts them; you can sort-of deal with a crocodile that shows up once a year to collect, but when you go SaaS you are basically walking by the crocodile all the time, hoping it won’t bite you with higher prices at a day’s notice. And their support is not very good either, which again is more critical in saas scenarios than on-prem.
In this situation, a significant new revenue stream ( an Android tax) is definitely worth fighting for. And Ellison loves lawyers; he even sued to compete in the America’s Cup...
Wasn't that in reaction to the previous America's cup challenge where the Australians introduced a winged keel that had previously been disallowed from 3-meter boat competitions? IIRC, both were allowed because the charter gave the challenger the power to set constraints.
> In this situation, a significant new revenue stream ( an Android tax) is definitely worth fighting for.
Except that there is no hope of that ever happening. Even if Oracle won in a big way, then Google would eat the one-time cost of deprecating the Java API on Android, similar to what Apple did between Objective C and Swift. There is no sustainable revenue stream there.
They have already done that, in practice - Google keeps coming up with other ways of writing Android apps and alternative runtimes, and encouraging developers to adopt them. It would have been silly not to, the minute Oracle sued, as insurance. But it’s a mammoth task and nowhere near completion. Moving such a humongous ecosystem is always hard; if you cut too abruptly you risk opening the market to competing platforms. This goes even more for a platform that has customers both above (developers) and below (manufacturers) the stack. Apple didn’t even change anything of substance internally, and still a lot of (most?) apps are not built with Swift. It will take years to replace Dalvik; and in those years, the crocodile would be happy to collect.
You can't do it overnight, but you can do it over an abbreviated time frame. Especially once you're sure you need to.
Then it could be announced that the Java API will be removed in a specific future version (so update your applications now), the Google Play cut of applications using it will now be 50% instead of 30% for those without it, some of the OEMs will immediately start selling phones that don't have it and those phones are less expensive so people will buy them and developers will have to use the new replacement to reach those app customers, etc.
Forcing a transition to happen faster is not fun, but that isn't the same as not possible.
Still the same stdlib API, at least the way Google has introduced Kotlin to Android. If they had any ambition to use Kotlin as their vehicle for abandoning the Java API I would expect at least some visible efforts of opening Flutter to languages other than Dart. But in typical Google fashion they seem to deliberately avoid forcing their teams down in the trenches into some unified grand strategy, giving them the freedom to build the best version of their product that they can. Given Google's deep pockets and how badly coordination overhead would scale to their project dimensions, this hands-off approach seems reasonable enough.
Not defending Oracle here, but the fact that the decision in Google's favor has been overturned twice suggests Oracle's not 100% crazy to pursue this. And the upside is mandatory licensing fees for every Android device.
Or that Oracle is politically connected to the court of appeals for the federal circuit, hence why a patent appeals court is claiming jusdiction over a copyright matter, and making it's own copyright case law in contradiction with the normal copyright appeals court case law to begin with.
The supreme court can send it back to a lower court with instructions, they can also decide on the case but note that their decision would not cover particular arguments if they were made.
It is zero-sum iff the benefit to Oracle is exactly equal to the deadweight loss imposed by the tax.
I tried to calculate this in the simplest econ 101 way, and got a surprising result. I would appreciate someone else commenting...
Assume that the schedule of supply and the schedule of demand are straight lines. We have three groups: (1) purchasers of phones; (2) vendors of phones; (3) Oracle.
In the absence of a tax imposed by Oracle, an equilibrium quantity Q_e of phones trade at an equilibrium price P_e. [1] This generates gains from trade, producer and consumer surplus, equal to the area circumscribed by the schedule of supply, the schedule of demand, and the line Q = 0.
When Oracle imposes a fee on phone sales of value F, the price and quantity-traded of phones adjust. The new quantity traded Q_f is lower than Q_e, and the new price P_f is higher than P_e. The gains from trade are now the trapezoid circumscribed by the line Q = 0, the line Q = Q_f, the supply schedule, and the demand schedule. The missing gains, the deadweight loss, are equal to the triangle circumscribed by the line Q = Q_f, the supply schedule, and the demand schedule.
Oracle receives F×Q_f in fees, a benefit equal to F×Q_f.
The deadweight loss is a triangle with base F and height (Q_e - Q_f). It is therefore equal to (F×Q_e - F×Q_f)/2.
Balancing the gains to Oracle against the deadweight loss, we see that they are equal when Q_f is one third of Q_e, or in other words when the fee is so massive that it cuts the quantity traded by 66%. If Q_f is larger than that, the benefits to Oracle exceed the loss imposed on purchasers and vendors.
This result would seem to imply that we could improve the efficiency of the marketplace by allowing random third parties to impose large taxes on pretty much any good. This is hard to believe. Where did I go wrong?
[1] In reality, there are a variety of models available at a variety of price points. This doesn't affect the problem I have with my result, though it is probably relevant to the phone market in specific.
You inverted the sign. The zero sum thing is the prospective transfer to Oracle. If you pay $100 and Oracle receives $100, net across everyone is zero.
Then whatever lack of commerce caused by the higher prices occurs is a deadweight loss with negative sum. You start from zero and go down.
To get a positive sum you need to be creating something with value greater than its creation cost.
So, before the fee, there's a certain amount of producer surplus, a certain amount of consumer surplus, and zero Oracle surplus.
After the fee, there's less producer surplus, less consumer surplus, and more Oracle surplus. As a consumer, I don't see which part of the price of the phone goes to the producer (generating metaphysical good) and which part goes to Oracle (generating metaphysical evil). I just see that the phone is worth more than its price by an amount equal to my consumer surplus. My surplus is measured in dollars just like Oracle's surplus.
But it seems that after Oracle imposes the fee, it generates so much Oracle surplus that it could cover the deadweight loss out of its own pocket and leave everyone better off. This is surprising.
I think your comment is mostly correct, but I'm going to try to put it in terms that make more sense to me.
The answer is that the loss of producer/consumer surplus is being calculated incorrectly. Deadweight loss is the triangle described, but the loss of consumer surplus is more than just the consumer-surplus part of the deadweight loss. It also includes the rectangle separating the price the consumer would pay in the absence of the fee, P_e, from the price the consumer does pay, P_f. My mistake isn't that I inverted a sign, it's that I compared the wrong quantities -- I included a benefit to Oracle without accounting for the cost to consumers and producers who are still willing to trade under the new regime. (Where the deadweight loss only includes consumers and producers who were willing to trade under the old regime, but not under the new one.)
(bananamulch and 34624523 both said the same thing I'm saying here.)
If the fee is per phone it adjusts the supply schedule by moving it towards a lower quantity as a higher consumer price is required to compensate the producers at the same amount.
The schedule doesn't change. The quantity traded does, but I mentioned that:
> When Oracle imposes a fee on phone sales of value F, the price and quantity-traded of phones adjust. The new quantity traded Q_f is lower than Q_e, and the new price P_f is higher than P_e.
The deadweight loss is the net loss to the combined system (producers, consumers, and oracle leech). So if the total surplus of the system was X, and the deadweight loss was (F×Q_e - F×Q_f)/2, the total surplus of the system with oracle is (X - ((F×Q_e - F×Q_f)/2)). This is less than X.
The producers/consumers combined would lose both the oracle tax and the deadweight loss (lack of opportunities that otherwise would have occured).
The loss to consumers and producers combined is both the oracle tax and the deadweight loss. So the lack of efficiency is measured by the deadweight loss, and the optimal amount of deadweight loss is 0.
But then again, what if phones have a negative externality from a society where no one talks to eachother? The oracle tax might be good for society.
It's certainly a downside for society, because there are people who otherwise would have purchased android phones that can't afford them because of the oracle tax.
The only ones winning this are the lawyers, salivating at the idea of this blob of mess, playing on words beyond recognition and way above any common sense.
I saw a lawyer's office earlier. Big logo said "We fight for fair". Sure, per minute.
I have used lawyers to defend my contract rights against people that try to weasel out of paying up for some services. How about we don't go black and white on lawyers (go case by case instead), to appreciate the complexity of human interactions and the corresponding complexity of determining who's right?
My lawyer is fantastic. On the few occasions I’ve used his services, he’s advocated passionately for me, and given me great advice. He cares personally about his clients and I don’t regret a single dollar I’ve spent on his invoices.
It’s strange how law and finance are both heavily demonized. Why is that? Both professions are productive, yet for some reason, it appears that most will only focus on the negative aspects of these two professions.
It’s the speculative part of finance that is loathed. The one that don’t care about fundamentals, product quality, humanity and the earth.
The law is in general a good thing, but you have this whole business going that is from a fundamental point of view just meaningless. This business is about patent claims, and often all it manages to do is stifle innovation.
You could argue the productivity of both these specialities.
I used to work in finance. People don't like finance because 99% of the time if there's a relationship involved we wouldn't have to compete on price.
It would be obvious if someone's shopping around a loan, to the extent that it was very easy to get a broker to send over a competitor's offer sheet so you could compare.
People hate finance because at the core it's sales and you make your keep by being predatory since everyone does it. If you don't, you sink.
One negative example for lawyers would be dragging a case about pensions out for over a decade so that over time people would die and reduce the cost to the company.
There are advantages to controlling a dominant platform. Even if Oracle doesn't get much direct revenue from Java they can use it to block strategic moves by competitors.
While I think this lawsuit is terrible and I hope Google wins the legal case, from a moral standpoint they kind of acted like assholes back in 2005 when they decided to fragment the Java ecosystem rather than negotiating a licensing deal with Sun. So now we have real Java in multiple JDK / JRE implementations plus this weird pseudo-Java offshoot in Android. The language used by Android looks a lot like Java but it hasn't kept up with improvements in the core language and standard library, so now developers can't reliably reuse code across those platforms. That harms the software industry as a whole.
There are plenty of commercial 3rd party companies selling alternative JVM implementations, none of them has ever had any issues with either Sun or Oracle.
They were foresighted enough to see mobile as a key platform, and critical for long-term advertising... but not foresighted enough to try to work with Sun, to make a subset of java giving reasonable performance/efficiency on mobile back then.
Probably because Sun went along with it, so why bother?
In Italian there is an expression “putting a good face over a bad game”. Sun was in shambles and had no muscle to force Google to do the right thing. The CEO was already so embattled that he couldn’t admit he had mismanaged the relationship. So they just smiled and tried to spin it as a great thing for the Java ecosystem at large; which, at the time, it kinda was. After all, it basically gave Java programmers a mobile story that wasn’t the dying J2ME, and even a new client-side story that could replace the quickly-disappearing desktop toolkits.
Other companies at least have some notion that they should be helping society or something like that, and occasionally they do. Oracle has a long track record of being evil.
There was a Freakonomics episode about just that. Studies show that most companies show of altruism is just social signaling and that people are willing to work for less if they think their company is doing good.
It's just a way to say "hey, see what happens if we get angry ?" so that smaller players will always try to comply to their ludricrous future claims. You can see that even if they will spend millions, knowing to loose, against a giant like google, for a reason that ridicoulous, a tiny business won't stand a chance.
Because you can't just take other software and change the license on it.
The industry's, justified, hate for Oracle is blinding them to the larger underlying concern.
Many people forget that Oracle technically won the legality of the last case, hence why we're here. They just weren't awarded damages. Even the judge didn't buy Google's arguments about open source or compatibility concerns.
They only got that ruling by appealing a copyright question to the patent appeals court, who for some reason decided to just go off into the weeds and set their own precedent in open conflict with the appeals court these questions would normally go to.
That's why all the rulings keep getting overruled. The lower court judge is sane (he's actually awesome, you should check out Alsup) and keeps ruling in accordance with established case law, and the appeals court (who shouldn't even be answering copyright questions) keeps going 'nah, all that case law is stupid'.
Hence why this should go to the supreme court who'll most likely tell the court of appeals for the federal circuit that they can't just go make their own law.
It's not true that they won the legality of the case through an appeal. The original ruling sided with oracle on technical and legal matters. They just weren't awarded damages.
By "the original ruling" are you referring to https://www.eff.org/files/alsup_api_ruling.pdf , which says, "on the specific facts of this case, the particular elements replicated by Google were free for all to use under the Copyright Act. Therefore, Oracle’s claim based on Google’s copying of the 37 API packages, including their structure, sequence and organization is DISMISSED. To the extent stated herein, Google’s Rule 50 motions regarding copyrightability are GRANTED"?
Can you point to what part of this ruling sided with Oracle on either technical or legal matters? (With the exception of rangeCheck.) Or can you link to what ruling you mean by "the original ruling"?
Google did not take software (other than rangeCheck), they took interfaces.
That said, it is common wisdom around F/OSS folks that you can't take other people's interfaces and change the license on it, e.g., I can't write and distribute a proprietary Linux kernel module using EXPORT_SYMBOL_GPL functions. I haven't seen a good analysis of whether this case is relevant to that—would a Google victory mean that VMware's use of vmklinux is legal (well, at least in the US), that ZFSonLinux is in the clear, that Nvidia can write proprietary drivers using dma-buf, and that CLISP should have been able to use readline without relicensing itself under the GPL?
> I can't write and distribute a proprietary Linux kernel module using EXPORT_SYMBOL_GPL functions.
That is because by using EXPORT_SYMBOL_GPL functions you are invoking code that is GPL'ed. If you made a clean room implmentation of a kernel that exposed the same interfaces as Linux, you could licence it anyway you want.
By calling any kernel API you are invoking code that is GPL'd. It is not like EXPORT_SYMBOL functions are BSD-licensed (and even if they were, they themselves call GPL-licensed code). In fact even making a normal system call from userspace invokes GPL'd code.
So I think there are only two interpretations. One is that EXPORT_SYMBOL_GPL attempts to claim a copyright on the interface, and not the code. The other is that redistributing a GPL-incompatible object that invokes GPL'd code is always infringement, and (possibly) the special exception for userspace in Linux's copyright statement, https://github.com/torvalds/linux/blob/master/LICENSES/excep... , is the only thing that makes normal userspace processes non-infringing, and that regular EXPORT_SYMBOL is basically also an explicit exception (which may or may not even be legally meaningful - during the dma-buf discussion, various copyright holders expressed that they did not interpret EXPORT_SYMBOL as an exception).
>By calling any kernel API you are invoking code that is GPL'd.
What does this have to do with anything? If you think this is a relevant statement you are severely confused about the current subject matter and what is being argued in court.
Please explain how the above statement is relevant to the current case.
That sentence is a reply to the comment above it, which states, "That is because by using EXPORT_SYMBOL_GPL functions you are invoking code that is GPL'ed."
My claim is precisely that this is not a relevant statement.
On a legal level, no, that's not a mistaken belief. Prior to this case, that was in fact the legal consensus. Given this case, it's only a mistaken belief if the Supreme Court decides for Oracle (or declines to take the case).
Yes, that is exactly my contention. An interface is no more software than a recipe is food, than a credit card statement is money.
From a technical perspective, interfaces in, say, C are mostly compiled out. There is no place in a (stripped) binary where you can point to a structure definition. There is a place you can point to names of functions that are yet to be dynamically linked, but only their names, that place is different from the program code (text section) itself.
None of this is to say interfaces are not an important part of software engineering. Interfaces, APIs, internal zero-cost abstractions, tests, comments, design documents, threat models, bug triage, commit messages, code reviews, operational techniques, runbooks, monitoring and alerting systems, end-user documentation, etc. are all important work products in the discipline of software engineering. They're also all (IMO) something different from the software itself.
> Please explain to me the fundamental difference between an interface and what you consider software.
"Software" is code. An interface is the expectations of code.
In particular, if software X implements interface A, it is possible to write a compatible drop-in replacement of X using software Y, which does not incorporate any portion of software X, is it not?
GNU libc and Apple libc both implement the UNIX libc interface. If you claim that an interface is indistinguishable from software, is there a portion of the GNU libc software that also present in Apple libc? Where is it? Can you point to it? (The source code of both is public, so this should be easy for you to do.)
Google gave another example in the blog post: the interface of keyboard shortcuts. If I implement a text editor that uses Ctrl-X for cut, Ctrl-C for copy, and Ctrl-V for paste, have I copied the software of another text editor that uses that interface?
> An interface in all its various forums as we understand interfaces are placed in binary. From object implementations to ABI's and API's.
An API, an "application programming interface," is distinguished from an ABI, an "application binary interface," precisely because it does not mandate what appears in the binary.
If you think that there is in fact a way where you can point to a structure definition in a compiled C object file, please identify where it is in this object: https://godbolt.org/z/SVhsJk
I agree with you: the interface is represented in actual binary format. But Google didn't copy that either. Dalvik doesn't even use the same binary format as Oracle's JDK.
What does Oracle, who practically prints money, expect to get out of this other than exercising a fleet of expensive lawyers? At best the case was always iffy and even if they "win" it's unclear what that really means for them given the inevitable subsequent legal maneuvers. Even more puzzling is how are they making the argument that Google has, in any way, harmed their business?
It's just mind boggling. A slip-and-fall insurance scam makes more rational sense to me.