Slashdot Mirror
Google Asks Supreme Court To Rule On When Code Can Be Copyrighted (theverge.com)
Google is asking the Supreme Court to make the final call in its infamous dispute with Oracle. "Today, the company announced it has filed a petition with the Court, asking the justices to determine the boundaries of copyright law in code," reports The Verge. From the report: The case dates back to 2010, when Oracle first accused Google of improperly using elements of Oracle's Java programming language to build Android. Oracle said that Google's use of Java application programing interfaces was a violation of copyright law. Google has responded that APIs are too fundamental to programming to be copyrighted. The case has led to two jury trials, and several rulings have doled out wins and losses to both companies over the course of eight years. Last year, a favorable Oracle decision set Google up to potentially lose billions of dollars.
Google asked for a Supreme Court hearing on the case in 2014, but the Court rejected the request at the time. The company says new issues are now at play, and is asking the Court to decide whether software interfaces can be copyrighted, and whether using them to build something new constitutes fair use under the law. In its new petition to the Supreme Court, Google says the case is not only important to copyright law, but has "sheer practical importance," as it centers around two touchstones of computing: Google's Android and Oracle's Java. The Court's intervention could alter the future of software, the company argues.
Google asked for a Supreme Court hearing on the case in 2014, but the Court rejected the request at the time. The company says new issues are now at play, and is asking the Court to decide whether software interfaces can be copyrighted, and whether using them to build something new constitutes fair use under the law. In its new petition to the Supreme Court, Google says the case is not only important to copyright law, but has "sheer practical importance," as it centers around two touchstones of computing: Google's Android and Oracle's Java. The Court's intervention could alter the future of software, the company argues.
Lets hope the special interests and patent trolls don't get the court to screw up!
;)
Just my 2 cents
It's somehow pleasing to see a huge copyright holder get shafted by the laws they helped promote. Hey Google, if this stuff went into the public domain in a timely manner like it's supposed to you wouldn't have your tit in a wringer right now.
You fucking take that shit about BASIC back right this second.
The U.S. legal system needs many, many improvements.
It's an interface. It should not be allowed to be copyrighted.
You progressives are remarkably bigoted.
I think the confusion here is because it's called an "application PROGRAMMING interface". Despite having the word "programming" in it, an API is not a program.
APIs are to programming as plumbing is to water. Or as electricity is to an electric socket. NOT the same thing at all.
It's as if we said "you can copyright electricity", and then somebody said well it's called an "electrical connector", therefore I can copyright it.
"Shoot, a fella could have a pretty good weekend in Vegas with all that stuff."
If Google loses, what happens to Unix?
I can imagine that the court would rule that merely copying the style of the API is not copyright. But copying the exact text of it, down to the names of the methods is another matter entirely.
You cannot copyright a genre of novels, or the types of characters within. But you can certainly copyright specific characters with specific names.
It flows exactly to *Nix. SCO might live again.
We would have to go through all of our applications and rename the methods. Or new legislation would need to be passed, which is unthinkable.
Interestingly, a big loser in that would be Oracle themselves. They should have bought SCO (or whoever now "owns" Unix) before starting this. Then they could have it all.
Wow, I can tell you really know your stuff. No one in their right mind would use Java. Any day now everybody will see the error of their ways and stop using it.
You so smart!
Al: the FSF was so insistent on the adoption of the GPL version 3
because the GPL version 2 is not operative against the grantor.
This deficiancy was, in their eyes, so fatal to the purposes that they
invisioned that they, as you have pointed out, ellected to employ
enhanced means of converting projects to version 3.
Eben Moglen's contention that the GPL 3 exists to internationalize the
GPL is a lie of omission.
It is true that that is a partial reason-to-be for the GPL version 3
but it is not the main nor most important reason for the insistence on
version 3 or "v2 or any later version".
The very heart of the GPLv3 is the addition of the "irrevocable by
grantor" clause and the "term of years" clause.
Additionally the contention by the FSF and, as I recall, Eben Moglen
that the reason the FSF only accepts code where the copyright is
transfered to the FSF - is so they may have standing to sue - is
another lie by omission.
Yes, standing to sue is vital, if you are going to sue.
But the real beating heart of the copyright assignment policy is to
prevent a programmer from rescinding license-to-use regarding his
freely given code.
Without copyright assignment, any contributor to the GNU project could
elect to rescind the gratis license grant at his pleasure, at any
time, as of right.
Without an "irrevocable by grantor" clause (as seen in the GPL version
3, but ommitted in GPL version 2 due to a mis-assumption on the part
of the drafter) there is not even an affirmitive defense of promissory
estopple.
We come to lie number three. (Neccecitated by:)
The incorrect assumption made by the
drafter of version 2 of the GPL and memorandized by Eben Moglen in
this very thread in an attempt to protect the error in drafting.
Lie number three: If terms regarding termination are proffered in a
copyright license they are the only means of termination.
This is true... If those terms are supported by bargained-for
consideration.
That is: the words Eben Moglen wrote were true; as regards to
commercial software, where the licensee has paied valuable
consideration for the terms offered.
The drafter of version 2 of the GPL was familiar with such commercial
licensing agreements. He failed in his due-dilligence if RMS did
indeed request an irrevocable-by-grantor license. The drafter stopped
his research at the typical copyright license stage, and did not
research furthur into what underlys the constuction there-of.
And thus we have the lie of ommission number 3.
(All, naturally, to protect Eben Moglen's former client and to
discourage litigation where there is, indeed, good and sufficent
cause for such litigation) (Which is all part of an attorney's
duties, I can assure you)
It's a good thing APIs weren't copyrightable in the olden days or Compaq would have been shut down on their firmware clone and there's be no PC industry and IBM would still be the bad guy and Apple would have died after their 1984 commercial and you'd be stuck with CGA video and 640k of RAM.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
Freely-available UNIX implementations have been available for DECADES . BSD. Linux. Hell, Solaris was open-sourced at one point.
And the owners of what you call "the UNIX kernel API" did NOTHING.
Doesn't matter. Estoppel doesn't apply. Novell didn't make any promises not to exert their copyright. They didn't know they could exert their copyright, until the Federal Circuit starting blathering around in not one but two idiot rulings. Depending on who you ask, they still can't. The Second, Fifth, and Tenth Circuits all say APIs aren't copyrightable at all, so the whole question of fair use is moot. The Third and Federal Circuits say they are copyrightable. The Federal Circuit went even farther, saying they're copyrightable and there's no way to make fair use of any API.
I read the 21 relevant pages from Google's writ, and I'd say it'll be granted. The second half of it hinges on fair use, and it's a weaker argument. The first half is all about the disagreements among the circuit courts, and that always attracts the attention of the Supreme Court.
Google's lawyers did attempt to address the earth-shattering ramifications if the Federal Circuit's ruling is allowed to stand, but not as much as I'd like. Hopefully they were saving the best parts of that for oral arguments. This Supreme Court has spent years asking, "Where's the harm?" and if you can't answer, you lose. If Google's lawyers successfully convey the harm, they have a good chance. They'd better describe the harm to all software everywhere though, not just the harm to Google software.
Google did point out that the Federal Circuit (full name United States Court of Appeals for the Federal Circuit) is disingenuous at best, malicious at worst, in their second overturning of the jury verdict that declared Google's use of Java APIs fair use. The Federal Circuit said that they assumed all findings of fact were completed by the jury (as they're required to do), then they contradicted both themselves and the jury and tossed out the verdict. I'm starting to wonder if the Federal Circuit botched their ruling on purpose in order to precipitate intervention by the Supreme Court.
It licenses the interface to develop compatible software and Google refused to license.
Licensing something to me that you don't own because you can't own it is fraud, not some ex post facto justification that a thing can be owned.
SCO spent years trying that shit, and ultimately lost, on every front. Their very last attempt was shut down in 2016, dismissed with prejudice, which means they're not allowed to ever make the argument again.
I'm not saying you're wrong in your conclusion that the Java API can't be copyright, but I'm not convinced by your argument.
Titles of popular songs:
Come Home
Stay
One
Smile
I Love You
Given that song titles are typically 1-3 words, of course the same word is used as a title for multiple songs. In many cases there are ten different songs to hit the billboard charts with the same title.
The Java API is thousands of words. (Grotesquely oversized, some would say). The only way you end up with the same thing that Java made is if you intentionally copy them. In fact, you'd have to copy them very carefully, it would be easy to mess up and have a method take arguments in a different order or something.
With song titles, if you pick any common English word, that's probably the title of an existing song. You almost can't avoid using the the same title. With Java, it's very hard to copy it exactly. It's also very easy to write an API that's just as good, or in fact better, that isn't identical to Java.
Song titles aren't copyrighted because they are one or two words, maybe three. Other people have already used the same title before and they will again purely by chance. The same is not true of Java.
Google needed to do lots more in the first trial to make a case for APIs being purely mechanical. Appeals are about errors in law, not in the case made by the party I agree with.
davecb@spamcop.net
You've got some santorum between your teeth.
I'd love to find a law that lets Google win this. I haven't seen a good one yet. Maybe Congress could make a law about interoperability, if Congress can ever do their job.
It's not one sentence. You can use one sentence from a book. You can't use thousands of sentences. And certainly not arranged in the same in the same chapters (objects).
Google has all the same stuff, thousands of words, arranged in the same structure. We're not talking about two or three words like a song title, or a sentence. The Java API is thousands of words - a complete book. I can't think of any law that allows copying thousands of words, though I wish I could.
OK. I take it back. Only because BASIC stomps the dogshit out of Java.
I read the 21 relevant pages from Google's writ, and I'd say it'll be granted. The second half of it hinges on fair use, and it's a weaker argument. The first half is all about the disagreements among the circuit courts, and that always attracts the attention of the Supreme Court.
Lawyers are trained to be convincing, and they are by tradition only giving one half of the case. Thus when you read one side's argument (or sit in court listening to one side's argument), it will usually sound convincing. To get a clear view, it is necessary to read both sides. Even better, read the judge's conclusion. But don't make your analysis based on only one side's lawyers. They are biased by design.
"First they came for the slanderers and i said nothing."
It's not one sentence. You can use one sentence from a book. You can't use thousands of sentences. And certainly not arranged in the same in the same chapters (objects).
It can be fair use. To claim fair use, you need to consider all of the four factors. Copying a large portion doesn't prevent a use from being fair use.
"Using most or all of a work does not bar a finding of fair use. It simply makes the third factor less favorable to the defendant." Wikipedia suggests copying entire TV shows (which is more than thousands of sentences) as an example that was ruled fair use.
If you have time I strongly recommend reading the court's finding. It is clear and well reasoned.
"First they came for the slanderers and i said nothing."
You don't understand their claims. At all.
You are correct, and Jesus fucking Christ, why should he have to understand the cluster fuck of a situation this is? Copyright law should be pretty simple.
This is what we get when we let a bunch of fucking lawyers start doing stupid shit like copyrighting code in the first place. Code is math and algorithms, period. You cannot copyright math. Additionally, why don't you have to release source code to secure a copyright? If you don't, then it is a trade secret that you are getting copyright protection for. The whole point of the system is creating something of value (not idiocy like one click purchase, that's the covered by the 'obvious to a professional' part of the deal), and then sharing the details with the public in exchange for a short term monopoly.
If we keep going down this road, more and more really stupid shit like this will happen. Ban ALL software patents and end this madness.
HA! I just wasted some of your bandwidth with a frivolous sig!
If interfaces can be copyrighted then WINE and ReactOS are dead unless they receive the blessing of Microsoft.
If WINE is dead, Steam on Linux is also dead.
If interfaces can be copyrighted, the message then is: if it's not an ISO standard so similar, don't touch it.
"Everybody's naked underneath" -- The Doctor
A better analogy would be if you wrote a book. Then I wrote a book with a different book title, but used all the same chapter titles that you did. Except the actual content of each chapter is different from yours.
Oracle is basically saying that writing a book with the same chapter titles is copyright infringement.
You're right the amount of copying isn't the only factor. The four factors are:
the purpose and character of your use (is it commercial, transformative?)
the nature of the copyrighted work
the amount and substantiality of the portion taken
the effect of the use upon the potential market for the original
Google fails the first. It's commercial. Original Java is a programming API, Google's copy is a programming API; there is no transformation into something else.
Google fails the second. There have been lots of programming languages. Google not only could have made their own language different from Java, but they HAVE since done so, while also publishing a copy of Java.
Google fails the third factor. They copied thousands of lines
Google fails the fourth factor. They wanted to license Java for Android and Oracle wanted to license it to them. Google basically decided to go ahead and use Java commercially, in the same way that Oracle intended to use it, and just not pay the license fee. Oracle was trying to sell Java for phones via other avenues as well.
You're correct, you don't understand. First, copyright and patent are two different things.
Second, what you can't patent is the laws of nature, including the laws of physics and the laws of mathematics. So you can't patent gravity - you CAN patent an elevator, which uses the laws of physics. You can't patent " x+y = y + x", you can patent machines that use gears and levers (which are multiplication). Java is not a law of nature. In fact in may be an abomination of nature. :)
Comment removed based on user account deletion
Google is mistating merger.
Walker v. Time Life Films, Inc., 784 F.2d 44 (2d Cir. 1986), is a great example of merger. It was a movie about cops in the South Bronx. Like another movie about cops in the South Bronx, "the scenery would include drunks, stripped cars, prostitutes, and rats." That's not a copyright violation, the court ruled, because ANY MOVIE about Bronx cops would have those sorts of things.
What's not protectable are elements that EVERY WORK in the genre tend to have. Space movies have laser weapons, and spaceships which inexplicably have wings.
Computer languages have functions like copying a string, which may be called copy, strcpy, strncpy, string.Copy, or just "=". Merger says can't complain that another language, much like yours, has string copy, because that's part and parcel of writing a programming language - they all have that, it's needed.
You said "In the Java language, there is only one way to declare a function". The key words there are "in the Java language". In other words, "if you're making an exact copy of Java ...". Google has no inherent right to make and sell a copy of Java.
Merger would apply if you could say "the only way to make a programming language is to ...". We know that making a programming language doesn't require duplicating Java (from the users' point of view), but most languages don't. Most programming languages do a lot different, a lot smaller, and a lot better.
When Google says "in the Java language there is only one way", what Google is really saying is "the only way to copy Java is to copy Java". Well it doesn't work that way.
Let's try looking at this from another angle.
I have a right to make a space fighting movie. The general idea used in Star Trek, Stars Wars, etc isn't copyright protected. What is protected is the storyline, the particular collection of characters, character names, catch phrases, etc. So let's try translating what you wrote to such a situation. You said:
--
Google's claim is that the abstract Java API is a "system or method of operation", and that it includes the specific names as part of that.
--
Do I have the right to use the specific names of Star Wars characters such Yoda, Han Solo, Princess Leia, etc in my space movie, because all of the character names are part of the idea of a space movie? Obviously not. I can make a movie about space, I can have a hero. I can't name my here Han Solo.
Yeah that's a more realistic analysis. The court decided the second factor in favor of Google though. It's an interesting read, even though it is long. I strongly suggest reading the decision.
"First they came for the slanderers and i said nothing."
Not just that, both companies are Left leaning companies. So even assuming that the justices would be biased towards entities that are pro Republican (not that it means much these days), they'd have little to choose from in this particular case.
The thing that's interesting is - how much of tech expertise do these justices need? Obviously, none of them know the first thing about coding. So the most I can assume is that the plaintiffs would have to highlight high level of appropriation in order to make the justices even begin to understand the issues at stake. But when those concepts are at that high a level, chances are that they encapsulate extremely broad concepts (kinda like Amazon's one button shopping), enabling industry-wide monopolies. From what I understand, the root of Google-Oracle's dispute is the Java VM. While Snoracle's VM was essentially memory based, Google's is register based. Would that not be significant enough a departure for Google to have made from Oracle, since it would have required some significant code re-writing to pull that off? Or would simply changing the memory mode that the VM operates in be inadequate a change? And most importantly, are the likes of Sotomeyer or Gorsuch or Roberts gonna grasp any of that?
Nice trolling, but even if Linux were to die, there are still the BSDs that are available. Besides, Google has been looking more at the BSDs since their license is more compatible w/ what they need, and so have the other major companies. Also, is GPL 2.0 revokable? Even the FSF and Stallman haven't tried it, as that would end 99% of the GPL market - namely Linux.
Maybe they should seriously resume work on the HURD?
Original Java is a programming API
Java is a programming language
The API (actually APIs, there are many) is a list of function/method names and types, and their parameter names and types.
Oracle's claims are that copying the text describing the API is copyrighted and copying the description is not protected by fair use.
Allegedly, Google distributed copies of the API description with the Android development kit.
I suppose one possible alternative Google could have used was have their development kit installer automatically download and run the Java development kit installer.
Don't try to out wierd me, three-eyes. I get stranger things than you, free with my breakfast cereal. --Zaphod Beeblebr
The problem is that software falls in this grey area between discovery, invention, and creative work. However, the creative work aspect is weak despite the code as art examples because you can make an artistic toaster without toasters generally falling under copyright.
Being what it is functionally, an API should not be copyright-able and neither should software. Most software shouldn't be patent-able either. Very little is actually inventive and most algorithms are better argued as a discovery than an invention. Complicated math was true before you stumbled on it and the same is true of algorithms. None-the-less I doubt that is what the court will rule.
Which was actually a great way to get acquired by Oracle and make money for shareholders. Mission accomplished.
Java is slow and shitty, software written in it almost always leaks and it's a nightmare to support.
Do you know why people use java? Because schools were obsessed with teaching java.
Btw thanks for the link to whichever decision that is. I'll read through it a bit later. I take it that's a court that ruled in favor of Google. I'm curious to read that because while I *want* Google to win, I can't imagine how they can under current law.
I take it you're not a programmer, or not a very good one. There is a VAST difference between a very good programmer and a beginner - the problem does not in fact define how the solution must be written, and every programmer would write the same thing.
Linux and Windows are vastly different because Microsoft didn't *discover* an operating system, they *created* one. (Somebody else created DOS). Creating new things is called being creative.
> Complicated math was true before you stumbled on it and the same is true of algorithms.
"The first search result should be Snopes" isn't a mathematical truth. It's not only not a mathematical truth, it's not even a truth, but an opinion, an editorial decision. Editorial opinions can be written using letters of the alphabet (which pre-existed!), and other characters on the keyboard such as "+" and âs. The fact that the concept of square root exists doesn't make it a mathematical fact that you should buy a stock when we the square root of the trading volume is greater than the CEO's age; that's an opinion. (A stupid opinion).
The fact that algorithms can easily give dumb answers demonstrates that they aren't mathematical truth - because they aren't even truth, they are opinion.
ReactOS requires 96 MB, so I'm going to say it's not just like Windows. :)
In fact, it was discovered that ReactOS developers copy/pasted code from Windows (via a disassembler), so indeed there are copyright concerns. Microsoft didn't care enough to even comment about it. Perhaps that was because ReactOS has dozens of users, but anyway Microsoft didn't indicate that they *care* about ReactOS doing blatant copyright violation, copy-pasting from Windows.
If Google were making billions of dollars distributing ReactOS on Chromebooks (Winbooks), then I bet Microsoft would care, and a court could decide how much ReactOS infringes, and if any of that infringement is fair.
The court actually brought that up. Google makes their own index of the web. if I were to steal a copy of Google's index and sell it, they may be able to sue me for copyright infringement.
Google's index is fair use of the snippets they use, the court has ruled, because it's transformative. You make a web page about iPad repair, Google makes an index of the web. Their use of a snippet from your page doesn't replace your page, in fact it makes your page more valuable for having been indexed.
I've read most of the opinion (ruling) and so far it's page after page of why Google should lose. I was thinking you had linked to a ruling FOR Google. I kinda want to skip to the the ruling to find out, but I kinda like the suspense too. :)
This ruling could affect interoperability, which is why I'd like to see Google win, but I don't think current law is in their side. I may come across a solid argument of why the law is on the side, but I haven't yet.
So far I've come across why the law *should* allow it (why someone wishes it lawful), and people repeating Google's very flawed arguments, arguments that very clearly misstate the law.
Found the Java coder.