EFF Files Amicus Brief In Google v. Oracle, Arguing APIs Are Not Copyrightable (eff.org) 147
Areyoukiddingme writes: EFF has filed an amicus brief with the U.S. Supreme Court in Google v. Oracle, arguing that APIs are not copyrightable. From the press release: "The Electronic Frontier Foundation (EFF) today asked the U.S. Supreme Court to rule that functional aspects of Oracle's Java programming language are not copyrightable, and even if they were, employing them to create new computer code falls under fair use protections. The court is reviewing a long-running lawsuit Oracle filed against Google, which claimed that Google's use of certain Java application programming interfaces (APIs) in its Android operating system violated Oracle's copyrights. The case has far-reaching implications for innovation in software development, competition, and interoperability.
In a brief filed today, EFF argues that the Federal Circuit, in ruling APIs were copyrightable, ignored clear and specific language in the copyright statute that excludes copyright protection for procedures, processes, and methods of operation. 'Instead of following the law, the Federal Circuit decided to rewrite it to eliminate almost all the exclusions from copyright protection that Congress put in the statute,' said EFF Legal Director Corynne McSherry. 'APIs are not copyrightable. The Federal Circuit's ruling has created a dangerous precedent that will encourage more lawsuits and make innovative software development prohibitively expensive. Fortunately, the Supreme Court can and should fix this mess.'" Oral arguments before the U.S. Supreme Court are scheduled for March 2020, and a decision by June.
In a brief filed today, EFF argues that the Federal Circuit, in ruling APIs were copyrightable, ignored clear and specific language in the copyright statute that excludes copyright protection for procedures, processes, and methods of operation. 'Instead of following the law, the Federal Circuit decided to rewrite it to eliminate almost all the exclusions from copyright protection that Congress put in the statute,' said EFF Legal Director Corynne McSherry. 'APIs are not copyrightable. The Federal Circuit's ruling has created a dangerous precedent that will encourage more lawsuits and make innovative software development prohibitively expensive. Fortunately, the Supreme Court can and should fix this mess.'" Oral arguments before the U.S. Supreme Court are scheduled for March 2020, and a decision by June.
Oh my god. (Score:2)
Re:Oh my god. (Score:5, Informative)
Re:Oh my god. (Score:4, Funny)
I'm sure Groklaw's coverage will be insightful.
Re: (Score:3)
Still one of the years where the legal system is basically a problem and not a solution...
This would also affect the GPL. (Score:5, Interesting)
Re: (Score:2)
Not really. It allows you to make an API compatible library under a different license so that you are not tied to the GPL'd library. But if the GPL'd library is the only option for running your program, then what you have is a derived work, whether you distribute the library or not.
Re:This would also affect the GPL. (Score:4, Informative)
But if the GPL'd library is the only option for running your program, then what you have is a derived work, whether you distribute the library or not.
But that's exactly the question that is being decided in this case. Is a work a derived work cause you are using its APIs?
Re: (Score:2)
Once that is done, GCC can argue that Clang is a derivative. But wait, that would make GCC a derivative of the C compiler in the original Unix.
Re: (Score:2)
And this is the root of the problem, what is an API and how far do we go before we say it's copyrightable. If you declare API's public domain, then all that is left is a specific implementation of an API which could be indistinguishable from another implementation. Especially in Java and other programming languages, there's only so many ways you can represent a string in memory on x86 or do a comparison. For programming languages, the API is the language so if all API are public domain you can't copyright a
Re: (Score:2)
And this is a problem how, exactly? If there is only one possible implementation of an API then both the API and its implementation are purely functional elements completely lacking in creative expression, and thus not the sort of thing copyright was ever meant to cover.
Re: (Score:2)
The right wing judges will come down on the side of companies who will win oodles of money going after API users. They'll argue it removes a barrier to fair trade. Any counter-argument will rely on the public good, something the right wing judges reject as existing.
Re: (Score:2)
> The right wing judges will come down on the side of companies who will win oodles of money going after API users
That is incorrect. Right wing judges are notoriously constructionist. The timing of this case is important, because this is the ideal time to maximize the change that APIs will be ruled to fall under fair use, given the specific wording of the law.
Re: (Score:2)
"by securing for limited times"
its an important part of the law which you chopped out for some reason.
If you're going to copyright anything, it has to protect the author for as long as is needed to profit off it before it enters the public domain.
Re: This would also affect the GPL. (Score:3)
If you're going to copyright anything, it has to protect the author for as long as is needed to profit off it before it enters the public domain.
No, you need to protect the author for the shortest amount of time (and to the least extent of protection) needed to get the author to have created and published the work that they would not have created and published otherwise.
Remember, most works are not profitable. And I don't mean in a Hollywood accounting way, I mean that for the vast majority of works there is really just no profit to be had. We should not reward authors of unprofitable (ie unpopular, or unexploitable) works with effectively infinite
Re: (Score:2)
Indeed - but I think they actually *understated* their case. The first penny in excess of expenses is profit, at which point by their logic copyright protection should then end.
The reality is more complicated - copyright robs the public of all the good that would come of being able to freely share and make derivative works, in exchange for incentivizing artists to produce more works than they otherwise would - which is also a public good.
The sane goal under the constitutional justification for copyright wo
Re: (Score:3)
I certainly hope that the Supreme Court rules that APIs are NOT copyrightable.
You better make your peace with the fact that APIs are copyrightable because the Supreme Court already declined to hear this argument once, letting the Federal court's ruling stand. What does this mean for you? If you use an API, make sure you have a license to use it
This would also have an impact on the GPL, effectively eliminating the distinction between the GPL and LGPL, at least when it comes to libraries.
No, the LGPL is a license that allows you to do certain things, and would not be impacted by this (briefly the LGPL means you can link to the library without your own code becoming GPL/LGPL. That will not change if the court rules that APIs
Re: (Score:2)
You better make your peace with the fact that APIs are copyrightable because the Supreme Court already declined to hear this argument once, letting the Federal court's ruling stand.
Did you miss the part where oral arguments are scheduled for March? The Supreme Court has agreed to hear this case. That's why it's a case that the EFF can file an amicus brief on.
Re: (Score:2)
GPL code will still be licensed as GPL, LGPL code will still be licensed as LGPL.
You are still bound to the license of the code you're linking to, because you used it.
What this means, is that if you re-implemented GPL code interfaces and licensed it as BitterOak 1.0 License, then the copyright holders of the original code cannot sue you.
Re: (Score:2)
You are still bound to the license of the code you're linking to, because you used it.
No. As GPL proponents are always pointing out, the GPL is a redistribution license, not a usage license. As long as you aren't redistributing the GPL libraries, or any executables which CONTAIN portions of those libraries (as opposed to simply dynamically linking to them at load time), then the only hold on you is your code's use of the APIs through this dynamic linking. If the APIs aren't copyrightable, the GPL effectively becomes LGPL.
Re: (Score:2)
No.
Yes...?
As GPL proponents are always pointing out, the GPL is a redistribution license, not a usage license.
An oft-quoted oversimplification does not license terms make.
The argument goes, that linking to a GPL work (dynamically or statically) constitutes a derivative work (since your work is a combination of GPL code and your own)
The API is not in question.
And of course, the oft-quoted oversimplification is still right, in that if you don't ever distribute your work- then you can do whatever the hell you want with it.
I'm sorry, but you're simply wrong in this instance.
Re: (Score:3)
Google copied EVERYTHING from Java when they made Android. People think the case is about Google copying:
public int compareTo(String anotherString)
That's because it is, because that's what the district and appeals court decided on and that's the question before the Supreme Court.
Google should absolutely lose, and no sane ruling that makes Google lose would jeopardize "muh APIs". A sane ruling would involve looking at all the shit Google copied, in detail via discovery, then ruling that they did NOT just copy a high level description, then awarding Oracle fat stacks of cash.
Your sane ruling is probably out of the picture. The Federal Circuit absolutely ruled in a way that put "muh APIs" at risk by ruling that google copying those descriptions is both a copyright violation and not fair use; neither how it was coded nor what it did internally were part of the analysis and the JVM/Dalvik issue were factors. The questions before the Supreme Court a
Re: This would also affect the GPL. (Score:5, Informative)
Re: (Score:2)
Copyright on API would destroy all interoperability. Hell if this holds Oracle might be paying out billions to someone else for using SQL, another API.
The ruling, if it stands, would literally make it impossible to interoperate with any API without violating copyright.
Re: (Score:2)
Oracle copied everything making their SQL server from IBM. They are playing with fire and if they get the ruling they want they will be bitch slapped into oblivion. The Nazgul at IBM are likely following the case with interest.
Re: (Score:3)
Just completely wrong. "muh APIs" is a Google talking point that they're STRONGLY PUSHING onto trash "tech journalism" websites.
Read the court filings and previous decisions. Google FUCKED UP and is going to get bitch slapped. It will NOT greatly affect the industry, it will be a narrow ruling because Google copied EVERYTHING. An "API" is the description and implementation, but all you young retards who don't actually program think it's just the description.
If you code, you don't know the history.
Much of the software that exists as a standard today exists because someone decided to re-implement someone else' API. MS-DOS v1 was a CPM clone, BSD recoded AT&Ts Unix. SAMBA, , WINE, SQL, Libc, The entire industry is based on re-implementing other people's stuff when you have a better idea.
Re: (Score:2)
You know, if you use sockpuppets the trick is to vary the tone and word usage between them. Posting with three different accounts that use the same words is just making it obvious.
Anyway, you continue to be wrong and I can only assume at this point that you work at Oracle.
Re:This would also affect the GPL. (Score:5, Insightful)
A specification is not an implementation.
Re:This would also affect the GPL. (Score:5, Informative)
An "API" includes the implementation.
Saying it over and over doesn't make it true. Google rewrote or used Apache's implementation of all but less than a dozen lines of code for the initial release, and then later rewrote those few copied lines of code (range check) almost immediately after.
Oracle's whole billion dollar case eventually hinged on the trivial range check code, which the original trial judge lambasted as being the dumbest claim in human history (I'm paraphrasing). Oracle, knowing that they had lost, then focused their claim on the API (which is a specification, not an implementation), which they also lost at trial -- twice. Then they successfully baffled with bullshit some stupid judges on the appeals court to agree with them.
Now we're all caught up.
Re: (Score:3)
Re: (Score:2)
What about when you call it? Then it is code. An API is both specification and code.
int foo(int a, int b) { // specification ...
}
void bar() {
int temp = foo(10, 20); // code implementation
}
That's not implementation of the API.
Re:This would also affect the GPL. (Score:4, Insightful)
As nagora said, that is NOT an implementation of the API, that's the *usage* of the API.
Specification: (this is what is part of the API) // returns the sum of the parameters
int sum(int a, int b, int c)
Note that it doesn't tell you how it's done - just what the calling form is, and a broad summary of what will happen if you call it. Even the specific verbiage used for the description is probably under copyright as well - it's only the functional information included in that description that's part of the API. Like recipes - you can't copyright the recipe itself because that is functional information specifically excluded by copyright law. But you can copyright the specific presentation and verbiage used to describe the recipe.
Implementation: (this is NOT in the API, and IS under copyright)
{
return a+b+c;
}
Note that this is the part actually tells the computer step by step how to do what the specification claims it does, and in most cases would be something far more complicated.
In Java the specification and implementation are located in the same place in the code, but they're still different things. If you've properly documented your code for javadoc, then the entire API will be included in the documentation, while none of the implementation will be.
You could automatically recreate the entire class heirarchy with all its methods using just the javadocs, and create a drop-in replacement for the Java libraries that would compile correctly... but it wouldn't do anything, because it would be an empty shell with no actual code implementing the described functionality. That's the API. To be useful you would then need to write all the code that implements that functionality.
Re:This would also affect the GPL. (Score:5, Insightful)
> Of COURSE an API is copyrightable
> how it's coded
BULLSHIT.
1. A specification is NOT an implementation.
2. We ALREADY went through this in the 1980's with clean room reverse engineering of IBM's BIOS.
3. If API's are copyrightable then every retarded company will try to claim that EVERY function API in the std C lib is copyrightable. This is beyond fucking retarded. It would set the ENTIRE computing industry back 100+ years.
4. API's being copyrightable would be like a Mathematician claiming that the PLUS operator, or ANY OTHER operator is copyrightable. This is fucking retarded beyond belief.
When does this insanity end???
> but Google absolutely copied nearly everything
FTFY.
5. You keep using that word "stole". I don't think it means what you think it means.
6. You can't fucking "STEAL" an API -- all you can do is COPY the specification.
If the specification is PUBLIC then you "legally copied it."
If the specification was PRIVATE then reverse engineering has LONG been held as a RIGHT.
Stop being an Oracle shill.
Re: (Score:2)
Re: (Score:3)
While I agree with your other points, please note that "It would set the ENTIRE computing industry back 100+ years" is not an argument --
It is, since the objective of copyright - and it is stated to be such in the relevant laws - is to encourage innovation.
Re: (Score:2)
It's an argument of what the law *should* be to implement its constitutional justification, but not of what it actually is. The same basic argument can be leveled against the outrageous copyright durations that cripple artistic innovation - but that doesn't change the fact that they are in fact currently law.
Re: (Score:2)
2. We ALREADY went through this in the 1980's with clean room reverse engineering of IBM's BIOS.
Google didn't do a clean-room implementation, or anything close to it, which is why they are in this mess.
Re: (Score:3)
Nope. But the Apache Harmony project did - and they're the ones Google copied the implementation from. As they were explicitly allowed to do by the Apache license.
Re: This would also affect the GPL. (Score:2)
Re: (Score:2)
But that's fine, because Google didn't use the trademarked Java name, which is what passing the Java compatability test was required for.
Oracle's legal recourse is limited to copyright claims - which don't apply because Google didn't copy anything protected from them, and trademark claims, which don't apply because Google isn't using any of their trademarked names.
Microsoft Java (mostly) died because they didn't want to pass the test, but had to use the Java name if the language was to be of any value to t
Re: This would also affect the GPL. (Score:2)
Re: (Score:2)
In the 1980's, IBM had not published every detail of their BIOS, they had some interfaces but most of the stuff was hidden in chip. It then required a clean room implementation to write the actual code/chip.
In the case of Google Android 'Java' this was completely different. Not only did it copy/paste pretty much the entire Java documentation, it did the same for the interface and given Java documentation defines the implementation, it's hard to argue that Google wasn't just copying Java wholesale, even impl
Re: (Score:3)
In the 1980's, IBM had not published every detail of their BIOS, they had some interfaces but most of the stuff was hidden in chip. It then required a clean room implementation to write the actual code/chip.
WTF are you talking about? The IBM PC and the PC/AT shipped with a FULL BIOS LISTING in the manual. They were made with standard TTL logic chips -- there were no custom ASICs on board.
Re:This would also affect the GPL. (Score:5, Informative)
It was different. Quote from wikipedia:
"Sun Microsystems, which held Microsoft in violation of contract for including a modified version of Java in Microsoft Windows that provided Windows-specific extensions to Sun's Java language; Microsoft lost this decision in court and were forced to stop shipping their Windows-specific Java virtual machine. Microsoft eventually ceased to include any Java Virtual Machine in Windows, and Windows users who require a Java Virtual Machine need to download the software or otherwise acquire a copy from a source other than Microsoft."
Microsoft had a License from Sun, but violated the terms, implementing a incompatible jvm. It was more about the implementation than the api. No copyright of api involved at all.
Re: (Score:2)
More importantly the entire SUN-MS lawsuit was a contract dispute, not a copyright claim.
Re: (Score:3)
No they did not. The court case clearly showed that Android did not contain any code directly written by Oracle. Not a single line. (There were 7 lines used by mistake in an internal Android test framework, but that was not distributed as part of Android).
The court case is about the fact that because Google wanted to have api compability with the existing java code, so some of the code Google wrote, looks like the code which Sun/Oracle wrote with the exception of whitespace things, such as tabs, spaces and
Re: (Score:2)
So sorry either you support anybody being able to rip off anything as long as you give it a different name,which means you can kiss FOSS goodbye as any corp can just rip it off by doing some copypasta and calling it "Frinix" or some shit, or Google needs to get a good bitchslapping for just wholesale ripping the whole damn thing off...so which is it?
You are incoherent. If anyone can rip off anything then FOSS will be the only game in town.
But that's not even remotely the case here. At most an API could be covered by a patent. Copyright law has nothing to do with it and is irrelevant.
Re: (Score:2)
>At most an API could be covered by a patent.
I don't think it could. Patents cover specific functional implementations - exactly the part of the code that is excluded from APIs.
Re:This would also affect the GPL. (Score:4, Insightful)
Copying the code is an issue. Copying the behaviour (teh API) is not. Which one are you arguing about? Do you know?
Re:This would also affect the GPL. (Score:4, Insightful)
Not from Oracle they didn't. They copied the implementation from the Apache Harmony project, which is a clean-room implementation under a license that specifically allows such copying, and over which Oracle has no legal authority.
Oracle's entire claim is about the API - the high-level description.
Re:This would also affect the GPL. (Score:5, Insightful)
A sane ruling would involve looking at all the shit Google copied, in detail via discovery, then ruling that they did NOT just copy a high level description, then awarding Oracle fat stacks of cash
This tells me you haven't kept up, because that was what the prior trial was all about. The current case is about if API's are copyrightable or not, which the copyright law and prior precedents explicitly says they are not. If the ruling is upheld, expect lawsuit bonanza with all the big ones dukes it out until they come to agreement and cross-license each other which means all the smaller software companies are SOL unless they pay license fees for the privilege of using API's.
Re: (Score:2)
Perhaps read the ruling that was handed down, because it directly contradicts what you are saying. plus the trial behind it actually did what you wanted - went through all the evidence.
Perhaps you are arguing that API's are functional code which was what Oracle's lawyers argued. Look up Anette Hurst and her statements regarding what she thinks API's are and think really hard about if you really want to agree with her misconceptions on technology.
Re: (Score:2)
Re: (Score:2)
By your reasoning if it is eventually found in favour of Oracle a firm once called Relational Software is going to have a lot of money to hand over to IBM for copying it's SQL language.
Novell will also be in for a large pay day for all that copying of AT&T's Unix API. The list is endless.
Re: (Score:2)
Well, it's a point of view.
It's also bollocks. An API is a description of expectations. There's nothing even remotely copyrightable about publishing your expectations; it doesn't even make sense.
Re: (Score:2)
The case has far-reaching implications for innovation in software development, competition, and interoperability.
Re: (Score:3)
Google copied EVERYTHING from Java when they made Android.
And Sun (Oracle), at least conceptually, copied everything from Objective-C when they made Java.
People think the case is about Google copying:
public int compareTo(String anotherString)
Compares two strings lexicographically. The comparison is based on the Unicode value of each character in the strings. ...
And other than the name itself, this is semantically identical to
- (BOOL)isEqualToString:(NSString *)string;
This is about Google copying that description, what it specifically does internally when running, and how it's coded.
There's exactly one way for an isEqual method to be coded. So you have a choice: Either Google did not violate the law or Sun (Oracle) did. Either A. the current landscape is the right way for things to end up, or B. Apple deserves a monopoly on cellular phone operating systems. There is no plausible uni
Re: (Score:2)
Sorry, my bad. It's semantically identical to:
- (NSComparisonResult)compare:(NSString *)string;
And the name is nearly identical, too.
Re: (Score:3)
That's why patents and copyrights are two different things.
Re: (Score:2)
People have always been free to write new code that does the same thing as existing code under another license. Copyright protects the actual code, not its functionality.
Re: (Score:2)
No they did not. Not a single line.
All 100% of the code in Android is either:
a) Written by Google, or people Google paid to write the code.
b) Taken from project apache harmony(Or other projects) which have a license which allows usage of their code. (Not from oracle projects).
Re: (Score:2)
You should take into account the fact that you're a total ignoramus when you post your drivel. Might be a tad less embarrassing if you just stop.
I copyright APIs with the letter E in them! (Score:2)
I copyright APIs with the letter E in them!
Has Oracle ... (Score:5, Interesting)
Re: (Score:2)
Re: (Score:3)
Better yet has Oracle paid IBM for copying their SQL language...
I don't know what to think... (Score:4, Interesting)
... I hate Google, and I detest Oracle.
Is there a way for them both to lose?
Re: (Score:3, Funny)
Re:I don't know what to think... (Score:4, Insightful)
Re: (Score:3)
At this point, Congress would just cave to the company lobbying. The biggest companies can afford the most effective lobbyists. The biggest companies are in the business of squishing any small upstart companies and any more roadblocks the biggest can throw at the upstarts will be welcome from the biggest's point of view.
Re: (Score:2)
Everyone seems to be forgetting that Oracle is primarily a database company, who's primary product is an SQL database. They also seem to be forgetting that Oracle have never paid a single cent in licensing fees to IBM for copying their SQL language. So if Google need to pay billions of dollars to Oracle for copying Java, Oracle need to pay many more billions of dollars to IBM for copying SQL. Oracle are playing a very very stupid game, because if they win this case they loose, big time to the point where th
Re: (Score:3)
For example, if you try building and selling a CPU with the x86 instruction set, Intel will sue you with a multi-billion dollar lawsuit -- because the x86 instruction set is copyright owned by Intel.
Generally - this is false.
People have implemented x86 independently. Generally, Intel only had patent law (the clone CPU was alleged to implement a patented procedure) and contract law (the clone CPU maker was alleged to violate terms of an earlier contract with Intel) to fall back on. The raw binary inputs to an x86 CPU are not copyrighted, nor copyrightable.
The only situations in which copyright would apply to CPUs specifically, is where the documentation was illegally copied, or the photo-lithography of
Re:I don't know what to think... (Score:5, Insightful)
... I hate Google, and I detest Oracle.
Is there a way for them both to lose?
Instead of deciding based on picking teams, maybe think about the issue at hand. If you're not a programmer, maybe it's not so clear... but if you are, it should be pretty obvious that alllowing APIs to be copyrighted will do incredible damage to the software industry and help basically no one.
Re:I don't know what to think... (Score:5, Insightful)
...and help basically no one.
Except the lawyers...
Re: (Score:2)
Re: (Score:2)
it should be pretty obvious that alllowing APIs to be copyrighted will do incredible damage to the software industry and help basically no one.
It won't be too bad. The key is to be sure to get a license to any API you use, and all the major ones are licensed in a reasonable way (including Java! The only reason Google got sued is because they didn't want to use the GPL for their implementation of Java).
Re: (Score:2)
it should be pretty obvious that alllowing APIs to be copyrighted will do incredible damage to the software industry and help basically no one.
It won't be too bad. The key is to be sure to get a license to any API you use, and all the major ones are licensed in a reasonable way (including Java! The only reason Google got sued is because they didn't want to use the GPL for their implementation of Java).
As others have pointed out, if APIs had been copyrightable in the 80s, Compaq could never have created a PC-compatible computer. Consider the implications.
Re: (Score:2)
An API is the entire fucking interface. It's the code that runs and does shit.
Please provide an example of an API that can functionally run. I'll wait....
Re: (Score:2)
Repeating your lies over and again does not make them true.
Google absolutely did not copy everything. This was made perfectly clear in the early trials. The implementations (which are the bulk of everything) are different. Google used a "white room" implementation. Yes, Oracle tried to claim some of the implementations were the same but they got laughed out of court.
Oracle (and you with your lies) are trying to outlaw white room implementation which would also outlaw reverse engineering interfa
Re: (Score:2)
An API is more that the fucking documentation you see on a website and the high level functionality of foo(int) returning another int.
No it isn't.
An API is the entire fucking interface. It's the code that runs and does shit.
It very specifically is not.
It's the detailed documentation of that code. It's the organization, naming conventions, examples, sample data, etc. It's the coverage of what functions were created and which were not (often because they can be done with specific use of one or more functions). In this case, it's also the fucking VM because we're not dealing with compiled code, but Java JIT bullshit.
Google copied EVERYTHING.
You are AN IDIOT.
Electric sockets are NOT made out of electricity (Score:2)
Re: (Score:2)
If you consider an isomorphic mapping equivalence, then anything that implements a software interface must contain that interface. I suppose you could argue that electric signals and magnetic fields aren't characters, but I don't think many would agree with you.
Re: (Score:2)
Re: (Score:2)
Google didn't copy the implementation, they re-implemented it which is completely different.
That's also why the judgement from the original trial only pertains to the API and not the implementation.
Re: (Score:2)
An API is always abstract. If you think I'm wrong, please provide an example of an API that can functionally execute.
If you can't, may I suggest you stop spewing bullshit.
Land of Wizards by Tom Wolfe (pg 127) (Score:2)
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
At that point you might as well say "the legal system in general is abused by the powerful," to which I say "no shit."
Re: (Score:2)
It's like they say... (Score:4, Funny)
A man that represents his shelf has a food for a client.
Extremely serious consequences (Score:2)
Ultimately what many may not consider is that the core API would be eligible for copyright protection as would any library or sub-component. In other words we're talking about the foundations of OSes and programming languages themselves, both standard APIs and ABIs.
int main(const int argc, const char **argv);
Who holds these copyrights? How long until they expire? Is it more than a century? APIs were never considered eligible as they are not creative works. So from the very beginning these foundations were b
Re: (Score:3)
I've got to disagree with one of your points. SOME interfaces MAY not be creative works, but almost all of them are. They are, however, functional requirements, and thus were never considered copyrightable.
Re: (Score:2)
Given your analogy, it would be perfectly fine to take a picture (photocopy) of the copyrighted picture because the picture itself has become a fact and is no longer creativity, basically creativity ends the moment you affix it in words, a painting or picture? Doesn't make sense.
Re: (Score:2)
Replying to self to elaborate a bit; A good example of copyright eligibility is a telephone number as I used in my description. Originally the choice of the format of such a number is indeed creative... something like 3 digits, hyphen, four digits (xxx - xxxx). The sole thing about that choice which makes it creative however is "3 hyphen 4" which isn't considered substantial. Now you might make up such a number, take 867-5309 for example. This number itself is also not substantially creative enough to warra
Re: (Score:2)
It's not about a function signature, description, the idea of implementating the Fibonacci sequence in a recursive way, etc.
It's about everything from the implementation (twice over, as Google was using JIT style VM bullshit, not compiled code) to the description, organization, examples, etc. Google copied EVERYTHING, as they always do.
Re: (Score:2)
int main(const int argc, const char **argv);
Who holds these copyrights?
In the case of the C standard library (a language itself isn't copyrightable), it would be ANSI or ISO, right?
Re: (Score:2)
"Documentation or record of fact is not eligible for protection. An example of such records of fact is a list of telephone numbers, names and addresses"
I don't know if you intended this but to my mind this is nearly an accurate description of interfaces. The only thing I would add is that their can be multiple entities behind the "telephone numbers, names and addresses", but I don't think that detracts from the meaning much.
Oracle are now a patent troll? (Score:2)
Kind of like words (Score:2)
Comments miss the point - not individual APIs (Score:2)