Oracle Wins Revival of Billion-Dollar Case Against Google (bloomberg.com) 332
Google could owe Oracle billions of dollars after an appeals court said it didn't have the right to use the Oracle-owned Java programming code in its Android operating system on mobile devices. From a report: Google's use of Java shortcuts to develop Android went too far and was a violation of Oracle's copyrights, the U.S. Court of Appeals for the Federal Circuit ruled. The case was remanded to a federal court in California to determine how much the Alphabet unit should pay.
The dispute is over pre-written directions known as application program interfaces, or APIs, which can work across different types of devices and provide the instructions for things like connecting to the internet or accessing certain types of files. By using the APIs, programmers don't have to write new code from scratch to implement every function in their software or change it for every type of device. The case has divided Silicon Valley for years, testing the boundaries between the rights of those who develop interface code and those who rely on it to develop software programs.
The dispute is over pre-written directions known as application program interfaces, or APIs, which can work across different types of devices and provide the instructions for things like connecting to the internet or accessing certain types of files. By using the APIs, programmers don't have to write new code from scratch to implement every function in their software or change it for every type of device. The case has divided Silicon Valley for years, testing the boundaries between the rights of those who develop interface code and those who rely on it to develop software programs.
I gotta believe this is hurting Oracle (Score:2)
I know people who are actively *NOT* buying Oracle because of stupid lawsuits.
But maybe Larry doesn't care anymore, he has enough money to play, and can't spend it all before he dies.
Weird.
Java Lava [Re:I gotta believe this is hurting Ora (Score:5, Insightful)
They certainly are working hard to kill Java. Besides the lawsuits, Oracle also gives Java lousy support and upgrade options. Few companies can make Microsoft seem the less evil alternative (C-sharp/.NET), and Oracle is one of them.
Re:Java Lava [Re:I gotta believe this is hurting O (Score:4, Insightful)
Re: (Score:2)
Re: (Score:2)
Re:Java Lava [Re:I gotta believe this is hurting O (Score:5, Informative)
Few companies can make Microsoft seem the less evil alternative (C-sharp/.NET), and Oracle is one of them.
Microsoft has open sourced the core of .NET under an MIT license, plus a patent grant. It's not hard to seem more evil than that...
Re: Java Lava [Re:I gotta believe this is hurting (Score:5, Insightful)
Re: (Score:2)
Re: (Score:2)
The thing is if you are selling software development software, people are going to push the limit of that tool, to do things you do not expect. Because if they do only what you expect to product to build, then chances are the program is already written.
Oracle if being a sticker on how one uses their tools, and too apt to sue them if their product became more successful, then the net of your tool, then chances are people are going to shy away from it.
Java isn't the only player, and other companies are much
Re: (Score:3)
Re: (Score:3)
Re: (Score:3)
Or in other words, they are willing to do extreme damage to others for a comparatively moderate gain to them. The original definition of "evil".
Re:I gotta believe this is hurting Oracle (Score:4, Interesting)
No, the world won't. The amount of infrastructure built on top of Java is pretty staggering, and much of the world will continue to use Java.
Re:I gotta believe this is hurting Oracle (Score:5, Informative)
And the amount of client-side Java code being replaced by HTML5 is staggering. Java is considered obsolete, and is being replaced just about everywhere it exists.
I work in the cable industry. The amount of set-top box code that's being refactored away from Java is in, and of itself, mind blowing.
Re:I gotta believe this is hurting Oracle (Score:4, Interesting)
Re:I gotta believe this is hurting Oracle (Score:4, Funny)
Cable industry is obsolete
Re: (Score:3, Insightful)
And if Oracle wins, they'll sue your ass because you're simply copying the Java API and replicating it in HTML.
In the case of Android, Oracle should win. Early Android essentially was Java. It's more flagrant than Compaq copying IBM's BIOS. (Yes, they claimed they made a functional equivalent in a "clean room" way where there was no chance of actual copying and won in court. Everyone knows that's total bullshit.)
I just wish they could both lose. Java continues to be awful and Oracle continues to sit on
Re: (Score:2, Interesting)
Oracle has its panties in a bunch because Google has deep pockets.
What about OpenJDK?
That's NOT created/supported by Oracle... yet you don't see Oracle suing the pants off of whomever shares does maintain OpenJDK. IANAL, but if an OSS implementation of the Java Spec (e.g. API) can exist in the form of OpenJDK, then there shouldn't be an issue with another group making their own implementation of Java.
Microsoft isn't suing the pants off of the WINE developers, nor are they targeting ReactOS.
Re:I gotta believe this is hurting Oracle (Score:5, Informative)
Re: (Score:2)
server side Java and embedded Java is an enormous market
Oracle is doing everything they can to put an end to that.
Re: (Score:2)
Modern server side Java is pretty fast, relatively usable, reasonably expressive with vast library and build tooling support.
And is rapidly losing ground to Go, which meets or beats Java in all those categories, is considerably less cumbersome, and does not have a litigation cloud over its head.
Re: (Score:3, Insightful)
And not much written in java is particularly long lived anyway.
Looks at CVS repo with commits with Java code as old as 1996 that is still running in production...
I work on helping third-parties integration with our APIs, and a lot of our customers are using Java Struts early-2000s code. You greatly underestimate how quickly companies replace working code.
Re: (Score:3)
Re: (Score:2)
People ever had goodwill towards Oracle?
News to me. Sun, sure; but they've left the building.
Re:I gotta believe this is hurting Oracle (Score:5, Insightful)
The danger is in the precedent. If they recognize function headers as protected by copyright law then writing any drop in library will become a licencing nightmare. Who owns "int round(float)" anyway?
This is way bigger than Java and Android.
Re:I gotta believe this is hurting Oracle (Score:5, Insightful)
A win for Oracle here will means utter chaos in the computer industry. Copyright is automatic and it lasts under current law for about 175 years. Based on this every API ever developed on a computer would suddenly become under copyright law. Think about that. SQL would be under copyright. The C run-time library would be under copyright. Game emulators, Samba, Wine, there are thousands of entries in the list. And every one of those is going to blow up into a legal fight.
int round(float) would likely end up being owned by AT&T who would then have the legal right to charge royalties for its use.
Patents are already doing enough damage to the computer industry in America. If you want to totally destroy the American tech industry, API copyright is the way to do it.
Re: (Score:2, Interesting)
For one, I would have just gone with a .NET implementation, because CIL and CLR are open standards. Yep.
Also, as per the Bill of Rights I wrote, people have a right to their own cultural heritage. I intend to pass a copyright act which limits copyright terms to 7 years, plus 7 more on renewal at $1,000 (inflation-adjusted), then an annual renewal at the 14 year mark each priced at double the prior. At 25 years, that renewal costs $1 billion.
To file for the first copyright renewal, you must supply us
Re: (Score:2)
That was about compatible electrical connections -- like the PCI bus with have now, but way older. It did not involve software.
Not just APIs, but compatible ANYTHING (Score:3)
What Google made was a Java-compatible runtime.
Oracle claims that making it compatible, using the same interface, is unlawful.
Consider anything that uses any kind of network protocol, serial, protocol, etc. If Oracle wins, the precedent is that it's unlawful to make an accessory compatible with any device which uses a non-standard protocol, because you'd be copying interface, just as Google copied the interface to Java.
Re: I gotta believe this is hurting Oracle (Score:5, Insightful)
The only part of Oracle Java that Google copied is the function signatures. It has been unequivocally proven over the years that nothing else was copied. And of course Google copied the function signatures (ie the API) sincethere is no way to make software compatible without copying the function signatures.
Since the only possible avenue to cloning is by copying the function signatures, Alsop and the jury rules this fair use.
Oracle's current case hinges on making it a copyright violation to clone the function signatures. And the problem with doing that is if they win suddenly every cloned API in the history of computing is going to turn into a copyright infringement. And the lawyers of the world will declare a week of celebration before they all become billionaires.
Re: I gotta believe this is hurting Oracle (Score:5, Funny)
Good to know that we'll be able to continue using Linux, so long as the round function is changed to compute the square root instead.
Oracle says "what it does", Google says "how" (Score:2)
> Wait, what? Int round(float) isn't what is in question here, it's what the function round does and the underlying code to execute it.
Half right.
Oracle's claim is that Google isn't allowed to have round(float) round it's input. Google says the copyright protects the *expression* of the idea, the implementation. The actual copyright statute says the idea is not protected, the expression of the idea is. Oracle says it's unlawful for Google to have a function called round() which rounds it's input, becau
Re: (Score:2)
But Google didn't implement a JVM. They created a completely new virtual machine called Dalvik (later replaced by ART in 4.4+). Dalvik didn't even use the same underlying operating principles. The JVM is stack based whereas Dalvik was register based. They did provide a tool that would convert JVM bytecode to Dalvik bytecode but Google did not recreate or put a native JVM into Android.
This whole lawsuit is about Google copying the Java library declarations (so yes, the int round(float) parts). Google re
Re: I gotta believe this is hurting Oracle (Score:3, Informative)
You realise that itâ(TM)s not illegal or infringement to clean-room reimplement a language, right?
The only thing Oracle has, if no significant copyright code was stolen, is to argue that APIs are copyright-protectable.
That is why this is all about APIs. And why the outcome, and every company arguing for it via amicus brief, is a threat.
Itâ(TM)s now a proxy fight over the future of the entire service-based internet.
Fucking Christ (Score:5, Insightful)
FFS... we need a special court for tech cases.
Re: (Score:2)
Re: (Score:3, Funny)
Isn’t there a court that handles nothing but tech cases? Somewhere in East Texas?
Re: (Score:2)
FFS... we need a special court for tech cases.
It seems that way, but I'm not sure why it should be. Suing a company for using your API's, is like suing a customer for calling your company's AVR system and pressing buttons on their phone to retrieve billing info and make payments. I think I could make that pretty obvious to most people, even a judge. Maybe Google's lawyers are making things too complicated?
Re:Fucking Christ (Score:5, Informative)
FFS... we need a special court for tech cases.
It seems that way, but I'm not sure why it should be. ... I think I could make that pretty obvious to most people, even a judge.
Unfortunately it often isn't. The initial judge learned to program so he could understand the case. He admitted his learning was rudimentary, but enough to understand what was going on. The appeals judges did not and that has been a point of contention. The only one in the legal system who learned to program was the trial court judge, and he threw out most of the arguments as a result. The appeals courts and the solicitor general have made their decisions without that, and Google's appeals say the knowledge is imperative.
Consider back in 2011, Justices Elena Kagen (age 51 at the time) and Stephen Breyer (age 73 at the time) had game consoles brought in and they learned to play multiple violent games so they could fully understand the case. They followed instructions on how to play their way through to the scenes that the plaintiffs found most objectionable, the ones where they say people were lighting children on fire and smashing their heads to get points. [washingtonpost.com]
If judges don't know enough to make a properly informed decision and they cannot gain that knowledge from case law and subject experts, it is their duty to gain that knowledge. There have been judges who took up hiking in dangerous situations, visited a wide range of locations from pig farms to remote deadly roadways, and studied arcane subjects in order to make a fair judgement.
Note that the first time around the question was split into three parts: copyright, patent, and damages. The judge learned how to program. In trials there are question of law and questions of fact; the judge decides questions of law and the jury decides questions of fact. The judge ruled from his experience learning to code that many of Oracle's claims were invalid on their face, and didn't allow them to go to the jury trial. The declaration was that "it does not matter that the declaration or or method header lines are identical," even though Oracle decided they were. The non-programmer appeals court judges reversed the decision. The Solicitor General (who also doesn't know how to program) recommended against a SCOTUS ruling at the time, letting it go back to the lower courts for further defense.
It headed back to the lower court for this second round. The judge and the jury decided there should be no damages because of the fair use exception in copyright law. The jury included people who knew how to program, and they decided it was perfectly legal due to fair use. Now the appeals judges (who still don't know anything about programming) is countermanding the jury's findings, which rarely happens.
Google should absolutely appeal. The next level is the full en banc review. They asked earlier for members of the three judge panel to learn to code, but they apparently did not convince them. Their lawyers should (and I'm sure they are) using that fact in their appeals. While the judges are comfortable enough with books and movies to apply copyright law in those cases, they can argue that unlike the trial court judge who learned to program to answer the questions, the appeals judges did not have a sufficient understanding to make these determinations, thus even though they understood the wording of law they still committed an error of judgement by failing to fully understand the content of the case.
Re: (Score:2)
Not Over (Score:5, Insightful)
Too much money and power in play.
This could take decades.
Dividing silicon valley my ass (Score:5, Informative)
Re: (Score:2)
On the one side you have the engineers that rely on APIs to Get Stuff Done, on the assumption that that's why the APIs exist. On the other side you have the 1% parasites who realize "oops, somebody took our work and made billions off it. Bring in the lawyers!"
An API is just a set of function declarations to call the code, you can't avoid making them because that's how it works but it doesn't have to mean you want everyone to use them. To make a car analogy, imagine you have an proprietary car with proprietary wheels. The width, length, thread and location of the wheel bolts are the interface. If you're building a new car, do they have to be the exact same? No, but if you want interoperability they do. So a lot of people here would say hey, I want to be able to m
Re: (Score:3)
Making compatible competitor implementations is more difficult if the interface itself cannot be reused. It's my opinion that allowing such reuse fosters more competition. The cheapest systems come about when multiple competitors produce products around central standards. Walled gardens are usually more expensive and less flexible. The loss of some reimbursement for interface design is a worthwhile price to pay for implementation competition.
Re: (Score:2)
The loss of some reimbursement for interface design is a worthwhile price to pay for implementation competition.
Particularly true in the United States, where the legal basis for copyright existing at all is Article 1, Section 1 of the US Constitution:
Copyright law that clearly retards progress isn't just a bad idea, it's unconstitutional.
Re: (Score:2)
Allowing homeless people to use your bathroom while you're at work, or allowing non-car people to borrow your car for some chore or fun ride is also efficient (similar to "more competition"). But who's gonna allow this? No one.
I would allow this in a heart beat, as long as it costs me nothing. And by cost I don't mean potential income I might get by renting out my toilet or car.
For my car: if it is available when where I need it, maintenance is performed, any damaged is fixed immediately with no inconvenience to me. It is just impractical to do so, but if self driving cars it may become possible.
For the toilet: as long it was cleaned, I can guarantee nothing in my house was damaged or stolen, and me and my family are no less safe
Re:Dividing silicon valley my ass (Score:5, Informative)
Sun claimed for years and years that Java was an open platform, and large parts of the Java platform were developed and contributed freely by third parties. That's why Google didn't design the APIs themselves.
On top of that, Sun itself took other people's APIs and built proprietary products around them numerous times, so the company that created Java thought it was OK.
Re:Dividing silicon valley my ass (Score:4, Informative)
Yes, we really did show that Sun copied everyone else's APIs in the lower court case. Hopefully this will come up when the supreme court hears this one.
This is in general bad for Free Software and anything else that depends on interoperability. It does mean it's easier to enforce licenses like the GPL, though. No more assuming that dynamic linking protects you. I've never believed it has, but a lot of people did. APIs pass across linking.
Re: (Score:2)
So what if they are similar to invented words? If someone invents a word does it mean no one can use that word without paying the inventor royalties? I have used words most of my life, not once have I paid someone for using one, even though I am sure every one was made up by someone.
You may point to trademarks but they are different I can say Coke as much as I like, as long as I don't call my product Coke, which is basically deception.
Using an API vs Reimplementing an API? (Score:4, Insightful)
The dispute is over pre-written directions known as application program interfaces, or APIs, which can work across different types of devices and provide the instructions for things like connecting to the internet or accessing certain types of files. By using the APIs, programmers don't have to write new code from scratch to implement every function in their software or change it for every type of device.
Have I completely misunderstood this case or has Bloomberg?
Re: (Score:3)
Re: (Score:3)
And of course there's no option on the website for people to point out factual inaccuracies such as this.
How to tell a reputable news source from an unreputable one.
So... (Score:2)
So... is 'Larry the Scumbag considered the lesser evil, yet??
Re: (Score:2)
Well, I for one will be cheering when he leaves this world. That will be the one time he made a positive contribution. Not that he is the only one for whom that is true.
EU Got one thing right... (Score:2, Informative)
Of all the bad laws being thought up in the EU, at least they got the API copyright decision correct.
APIs are not copyright-able in the EU according to their highest court.
https://arstechnica.com/tech-policy/2012/05/eus-top-court-apis-cant-be-copyrighted-would-monopolise-ideas
Re: (Score:2)
Just shows how utterly demented powerful parts of the legal profession are in the US. No connection to reality anymore at all and apparently also incapable of listening to actual experts. Like some politicians (the orange one for example) I could think of.
Re: (Score:2)
Re: (Score:2)
> ORCL Mkt cap 189.87B
> GOOG Mkt cap 726.29B
Nah, I think Google is actually pretty happy they didn't buy Sun.
Re: (Score:2)
Clearly the only difference between the companies is whether they bought Sun. I mean, they both produce databases, and query databases as their primary business. Now, Oracle does it as a middleware, and Google does databases of tracking, ads and searches (in that order) which it then provides as a service, but it's the same thing. Computery-stuff is all interchangable. That's why I have web designers write all my C code for embedded devices.
Re: (Score:2)
Google and some others too. It was utterly demented to allow Oracle to get control of Sun.
Comment removed (Score:3)
A more accurate report summary (Score:2)
The dispute is over pre-written directions known as application program interfaces, or APIs, which can work across different types of devices and provide the instructions for things like connecting to the internet or accessing certain types of files. By using the APIs, programmers don't have to write new code from scratch to implement every function in their software or change it for every type of device.
Let me take a stab at using my journalism skills to improve the clarity and accuracy of this summary:
At the heart of the issue is a technical lingo known as API, which is how programmers tell computers what to do. Since computers can only do as they're told, and only understand ones and zeroes, programmers must use a special robot language to talk to computers, which gets translated from the programmer's magical jargon to a series of ones and zeroes. Unfortunately, a series of ones and zeroes looks very similar to other series of ones and zeroes, so it can be difficult for programmers to keep track of what they're telling the computer.
That's where APIs help; they let the programmers speak more normally to computers, using a pseudo-robot language like "while iterating a data structure perform calculations on the user's social data network", which gets translated into a series of ones and zeroes like "1000101".
Doesn't matter anymore. (Score:2, Insightful)
J2ME, SavaJe, etc.. (Score:5, Interesting)
Sun and later Oracle were so fixated on the nonfunctional J2ME stack that they didn't see Google coming. Sun had the skeleton of what later became Android with the SavaJe acquisition but didn't have the funds or desire to make it happen. They (and later Oracle) failed to adequately invest in mobile and now Google has. Larry is simply pissed that someone other than himself is making money on something he neglected to capitalize on. To Lawsuit Larry, that's a crime - to anyone else, it's called business.
Re: (Score:2)
Lets just be glad there's no #metoo Leisure Suit Larry
Re: (Score:2)
To Lawsuit Larry
Remember that Ellison got his start working for the CIA. Businesses that do a lot of government contract work start believing that they have been granted an exclusive license from a king giving them sole rights to a market or product. And in some ways this is true. I wouldn't be surprised that if things start not going Larry/Oracle's way, some of the TLAs will pull a few strings to get things put right. Or, you know, that surveillance database might stop working.
Oracle's private sector customers (or anyone
Java is popular still? (Score:2)
15 years ago, I was handed a project to interface with a system where the only documentation was the Java reference implementation. The company we were dealing with did everything in Java; there was no support from them if you strayed from the safe lands of Java.
When they started doing "proper" documentation, Java was first and foremost in it. There was slight mention of other programming languages.
Now, you cannot find Java mentioned in any of their documentation. You can submit support requests for PHP, .N
WTF? (Score:2)
If that was Google's defense then they should have lost in court. Java was developed for mobile devices FIRST! That was Java's intended use. The desktop market took over when the developers realized the potential the JVM provided.
Thug on thug violence (Score:2)
I'm conflicted about who to root for in this thug on thug match up.
If Oracle wins (Score:3)
Then R.I.P. Java. It is too dangerous to develop new software if APIs suddenly acquire copyright protection because the licensing terms effectively change and could impact your business. At this point to run a business responsibly you should negotiate up front what your license terms for tools, compilers and APIs should be. Going with platforms, tools and languages that are open licensed is going to be necessary for many businesses that don't have the resources to wrestle with licenses, lawyers and sales staff with complex licensing payment terms.
Re: (Score:2)
Re: (Score:2)
Is it? If I write a plug-in then did I duplicate the implementation of an API? Is the direction of call why you're trying to distinguish here? I think you may be trying to differentiate two things that are really two sides of the same coin.
Re: (Score:2)
Re: (Score:2)
If you "implements" or "extends" is that allowed? We would have assumed yes, but everything is thrown into question. What is "core" what can be extended, what interfaces are your permitted to use in your own implementation? Can I write an image file loader for Java? there is already a core class for that, what if mine implements the same interface, did I violate the license agreement? did I infringe copyright?
I'm not a lawyer so I don't know, but in light of this I suggest that no matter how small your busi
I thought this was decided ! (Score:2)
geez -- feels like a tennis match. This case is still going on?!?!
Fair use, not fair, Out of bounds. Lob. Love.
call me when it's over.
An API is like an electronic schematic (Score:2)
It describes to a person who can understand it how, at least theoretically, something works or is supposed to work, but does not itself offer any useful function or purpose beyond educating a person with a sufficient understanding of relevant source material on how whatever is described by it might be built. It's actually just a description of how the thing would work *IF* you actually built it.
It's worth noting, also, that electronic schematics are not ordinarily even considered to be copyrightable by
You would lose Wine (Score:5, Insightful)
A ruling in Oracle's favor would shut down the ability to use Wine to run Windows-exclusive applications on your Mac because Microsoft could assert copyright in the Windows API against the Wine developers. Is that an outcome that you find helpful to the economy?
Re:But would you, or is that really bad (Score:5, Insightful)
A ruling in Oracle's favor would shut down the ability to use Wine to run Windows-exclusive applications on your Mac
First of all, I question if this is really true. It's a different case because it presents a way for executables to make system calls that work, it's not the same as an API specific to a programming language. It is very similar, I grant you that.
99.9% of WINE is the re-implementation of Microsoft libraries, just like Google reimplemented the Java libraries. There is no difference. You could just as well call the JVM an operating system, which in several ways it is.
Also, WINE is not making any money, whereas Google has made a ton of money from Android.
Most of the WINE developers are employed by CodeWeavers, which sells a commercialised version of WINE.
Double also, Microsoft benefits from people being able to run Windows applications in more places, because it encourages more Windows development.
Microsoft is trying to transforming itself into an advertising platform. Everyone that runs Windows applications without sending in as much usage data as possible to Microsoft is a loss for Microsoft.
Lastly though, I have to say think it would be awful to lose WINE. But the question is, would it be wrong?
The "commercial/exploitation rights" part of copyright are a tool to encourage innovation and competition (at least in jurisdictions that distinguish between the moral and the exploitation rights; in the US, the whole shebang is enshrined as intended to advance the useful arts etc). I don't see how extending them to cover APIs would help that.
Re:Why is this wrong? (Score:5, Insightful)
It's true that without the large number of Java programmers around, Android would not have had nearly as fast a rate of adoption as far as application creation went.
So? API's aren't copyright-able. Linux wouldn't exist if they were. This was settled ages ago. It's only Oracle's ability to fund the caliber of lawyers necessary to engineer an appeal that is keeping this alive. It's really just a reflection of our corrupt pay to win judicial system.
Re:Why is this wrong? (Score:5, Insightful)
You have to nothing to receive a copyright. It is granted to you automatically and under current law the copyright it is yours for 175 years. For the entire 60 year history of computers everyone has believed that APIs can not be copyrighted.
What happens if Oracle manages to change this?
The computer industry is going to collapse into utter legal chaos that it will likely never recover from. That's because if you make Oracle's API have a copyright then that ruling is going to apply to every API in existence since they are all still under copyright. Think about it -- consider how many APIs have been reimplemented over and over. Ownership of SQL is going to revert back to IBM allowing IBM to demand royalties from everyone using SQL. The C run-time API (like printf) is going to revert back to AT&T as owning it. What about Posix? What about game emulators? This will result in total legal chaos in the computer industry.
Re: (Score:2)
What if the positive ruling is such that somehow it only applies specifically to Java API and what Google has done with it? If so that may not be a bad outcome at all. I wouldn't mind seeing some of the Google's power, or even a good chunk of it, being transferred over to Oracle.
Re:Why is this wrong? (Score:4, Interesting)
IANAL but I think this decision is pretty binary - Alsop originally ruled that APIs could not be copyright. Oracle forced him to change that ruling to API could be copyright but cloning them is fair use. But now Oracle is trying to reverse the fair use ruling by the jury.
If that happens all APIs will be under copyright and the only way to clone them will be with a license. And since those copyrights are going to get constructed retroactively since the dawn of computers the only possible outcome is total legal chaos.
Re: (Score:3)
It's all about copyright, correct? Copyright only applies to creative expressions. Purely functional expressions are not protected. APIs only provide context for calling functions and passing parameters. They, themselves, do not perform any function but merely describe how the function should be called.
The original ruling was mistaken. Google's use was not Fair Use. What was copied wasn't subject to copyright to begin with. Google used Java APIs but wrote the code behind the APIs. Oracle invented a
Re: (Score:2)
In the first court case Alsop ruled correctly the APIs were purely functional expressions and could not be copyrighted. Oracle appealed this to a bunch of Luddite judges on the Appellate court who overruled Alsop and declared that API could be copyright. In the second court cast Alsop/Google were forced to accept the ruling that APIs were copyright. The new (and correct) decision by a Federal jury was that if APIs are copyright, then it is Fair Use to clone them. Oracle appealed this to another set of Lud
Re:Why is this wrong? (Score:4, Interesting)
But that value of an API is created not by the company that created the APIs but by the companies that invest their time and money in using the APIs.
In the case of Sun/Oracle, there are three additional factors to consider: (1) Sun achieved rapid and widespread adoption of Java's APIs by falsely claiming that they would make those APIs an open standard; that is, Java APIs would not have succeeded in the market if Sun had been clear from the beginning that they would not allow compatible third party implementations; (2) large parts of the Java APIs were not even developed by Sun or were copied from other APIs; (3) Sun/Oracle has a long history of building their products on other people's APIs.
Re: (Score:3)
Why should Google have to pay Sun/Oracle anything? That's like saying that authors need to pay Merriam/Webster/Oxford for using words.
The main problem with this case is that Google have never adequately explained what an API is. If you read the original testimony/filings and decision, the definition of what an API is is laughably vague, and as a result none of the lawyers know what they are fighting about. The correct definition of an API in copyright terms is a "blank form", that requires that you fill
Re:Why is this wrong? (Score:5, Insightful)
"Why is that the case?"
First of all, look at the can of worms it opens.
Time SetTime(hr, mm, ss, ms)
That's an API. For creating some sort of 'time' construct, initialized by using a method called SetTime, and passing it the hour, minute, second, and milliseconds.
class Object {
boolean equals(Object o)
}
That's an API, you have an object, and it has an equals method, that compares itself to another object and returns a boolean based on their equivalence.
The minute you rule that an API is protected by copyright, someone gets to claim ownership of that API. And every other language, library, or tool that came afterwards is in violation.
" most languages these days are only powerful because of extensive libraries and API's"
If APIs are protected by copyright, then not only is an exact duplicate protected, then a 'derivative work' is also protected.
Most APIs for similar things are quite similar. Whether I'm using Java or Python or C# or C++ the various standard and even 3rd party APIs for for a List container, or for a Button widget are very similar, often identical. The lists all have an 'add' method, and a 'contains' method. The button has a position. size, label and property and a clicked event handler. Even when they aren't identical, surely they are clearly related -- and i could easily argue that your button api that came after mine is a derivative work ... look how similar they are.
Why it's almost like they are similar on purpose! Someone owes me a billion dollars!
Further; identical APIs are particularly useful as drop in replacements to correct buggy or crufty or non-performant or legacy solutions. We have openGL wrappers for directX, directX wrappers for OpenGL, we have 3DFX glide wrappers that allow old 3dfx games to run hardware accelerated on modern hardware. All these 'wrappers' exist to allow software written to one API to be used without modification on a 'new backend'. If companies can own and assert copyright on the API, then API wrappers cannot exist.
If you write some software against some 3rd party library, and then years later want to change the backend out, you can rebuild the software to a new API, or you can write a wrapper class for the old code to call that translates it to the new backend. The new backend has its own API. But the wrapper class is effectively a re-implementation of the original API.
It's classic software design pattern: "the Adapter pattern"
https://en.wikipedia.org/wiki/... [wikipedia.org]
Aside from Adapters, lots of people have written initially against 'standard library' or 3rd party library functionality; and then later gone back and dropped in custom built optimized-for-their-needs solutions to those libraries. That's usually done by building against the original/standard/3rd party API and then re-implementing the API you need with your own custom stuff later on, as you need it, if you need it.
Then we have projects like WINE which implements Windows API, FreeDOS which implements the DOS API. We have all manner of 'plugin architectures' that function in like ways, but that constitutes an API. We have chip emulators that emulate the hardware API. We have hypervisors that emulate certain hardware API. We have virtual devices that emulate the physical device APIs. We have all kinds of software bits to run cpu instructions in software on chips that lack hardware support -- that's API.
The Browser DOM is an API, and the web would be very different if browser vendors could have simply asserted that other browser vendors couldn't include their DOM extensions.
Software engineering and software development has spent the last several decades under the assumption that this was all ok, because it WAS ok. It's ludicrous to change the game now. This doesn't just affect Google vs Oracle and the implementation on Android.
Changing the rules on what you can do with an API rewrites the software industry.
Re: (Score:2)
People familiar with the extent of Java API's would be able to develop applications far more quickly for Android than if they had just used Java the language but come up with a whole new standard library.
That is actually what they did, but you have to implement all of java.lang (e.g. String) and obviously it makes not much sense to provide your own APIs for java.util.*, java.io.* etc.
Basically you have to implement everything that starts with java.*
Re: (Score:2)
Re: (Score:2)
The API is an interface specification. It is "mathematics". As such it has neither copyright protection nor patent protection. This is akin to claiming that a "thing with a handle and you can cut other things with it" is a protected concept. It is not. Far too general and obvious.
On the other hand, if Oracle wins this then Java is basically dead. That would be a good thing.
Re:For once, I am on Google's side (Score:5, Informative)
Re: (Score:3)
Then whistleblow. Call the company that Oracle defrauded and let them file criminal charges against oracle and your old boss. Then make some popcorn.
Re: (Score:2)
Re: (Score:2)
OpenJDK came along later
Re: (Score:2)
Actually, OpenJDK is using the Oracle API as well, AFAIK.
Re: (Score:2)
One Rich Asshole Called Larry Ellison
Re: (Score:2)
Re:All I can say to this is... (Score:4, Funny)
Larry has his own private Hawaiian island to retire to... if I were him, I would retire already!
If Larry were you, he'd probably say that if he were Larry he'd retire.
Re: (Score:2)
Re: (Score:2)
If that is your most important issue with Python, you should probably not be writing code...