Reusing software 'interfaces' is fine, Google tells Supreme Court, pleads: Think of the devs Google last night strode into the last-chance saloon of the US Supreme Court, warning judges (PDF) that if they did not overturn a Federal Circuit ruling in Oracle's favour over its use of Java code in the Android mobile operating system, it could "upend ... the computer software industry." "New entrants into a software market …
Do you prefer Google taking its place? If the part were reversed, I don't think Google would have let Oracle took advantage of Google's IP.
Anyway Google admitted it copied Java without licensing it just because it need a lot of developers quickly to try to lock-in the market into Android as soon as possible - cutting costs and time reusing as much code and tools as possible.
Google is also free to publish all the hidden code in its Google services in Android, if it's so worried about interoperability and let developers free.... and let user remove its applications.
Apparently you are new here. There's a long history of API's being *outside* of the bounds of copyright. Going all the way back to x86 compatible CPUs. Oracle made a 'Hail Mary' play to try and squeeze a windfall from Google from their purchase of Java from Sun. They included a patent claim with the cynical goal of getting it before the one appeals court that never missed out on a chance to expand intellectual property, whether it made sense or not. The original judge came to a well documented opinion that API's aren't protectable, even going so far as to learn to code himself. When that was overturned and sent back to determine if it was 'fair use', the jury ruled that it was. Once again overturned, disregarding the jury's verdict in its belief that absolutely everything is patentable/copywritable. This particular court's already been _slapped_down_ by the Supreme Court for allowing patents on things that were clearly unpatentable. Hopefully they do the same again in this case.
Are you sure? Feel free to link to the relevant documents.
AMD had a license from Intel, since it was one of its second-source suppliers, on the x86 instruction set. Still, that's not an API.
The fact that you all tremble because Linux copied the Unix API doesn't matter. Sun won against Microsoft when Microsoft Java implementation didn't follow the license.
When Microsoft will make a full copy of the Linux API you all will start to cry out it was unfair and MS should be blocked and fined.
Anyway, if you didn't discover that Google is far worse than Oracle, you'll discover it soon. Oracle was never so tentacular like Google is. Keep on licking the breadcrumbs it throws in front of you... while biding you fully into its profits-churning machine.
And open source developers are fully discovering it when monsters like AWS start to deliver services with the same API of their software and will push them out of any market as they please.
I suggest you go and re-read the history about Sun v Microsoft. For starters, Sun didn't "win" - the case was settled, albeit with Microsoft paying Sun a sum of money. Furthermore, the case was predominantly about the alleged breach of licence that Microsoft had from Sun, and not about any API-related matter - other than Microsoft's changes to Java which (in Sun's eyes) meant that they shouldn't have been calling things Java.
After the case, there was wide speculation that Microsoft would in fact go and do what Google have basically done - namely re-write Java and just call it something else.
> there was wide speculation that Microsoft would in fact go and do what Google have basically done - namely re-write Java and just call it something else.
Which Microsoft of course did: the result is called C#
If you insist:
17 U.S. Code § 102.Subject matter of copyright: In general ( https://www.law.cornell.edu/uscode/text/17/102#b ):
pay particular attention to section b:
"(b)In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work."
An API is a just such a 'system or method of operation'.
Then there's Baker v Shelden ( https://en.wikipedia.org/wiki/Baker_v._Selden ) back in 1879, where a system of accounting was ruled unprotectable by copyright. This came down from no less than the Supreme Court. His _book_ (a particular expression of the system) was copyrightable, but the system wasn't.
Then there's Bikram's Yoga College of India v. Evolution Yoga, LLC ( https://www.eff.org/files/2015/10/09/yoga-copyright-opinion-ca9_0.pdf ) in which the ninth circuit court held in 2015 that the sequence of poses was not the proper subject of copyright. This decision is important in two ways, first it provides a plethora of references to the various decisions underlying why an API isn't copyrightable including the previously mentioned Baker v Shelden. Second, the Federal court was _supposed_ to apply the rulings of the court where the case originated, which in this instance was the ninth circuit. Instead they chose to ignore precedence and rule that APIs are protectable by copyright.
As for your contention that AMD 'had a license' from Intel, that was for the use of their patents, not any copyrights (spoiler, there weren't any).
It's funny that you mention AWS. Do you know that Oracle has implemented Amazon's S3 API without permission (which is perfectly legal) ( https://arstechnica.com/tech-policy/2020/01/oracle-copied-amazons-api-was-that-copyright-infringement/ )? Oracle is claiming that it's OK because the Amazon SDK is licensed under the Apache license. But that's the *SDK* _not_ the API. Java's SE SDK was licensed under the GPL, but Oracle is arguing that license only apply's to the SDK, not the API.
Hmmmmm.......
No, that statement does not mean what you think it means. If it did, then software could not be protected at all, which would be an untenable situation.
If the original creation meets the threshold of originality then the affirmative defence raised by the maker of a derivative work fails. So all APIs are in copyright, and always have been, but not every attempt to protect it against copying will succeed - as we learned from the SCO case.
This is Google’s big lie.
I do hope this helps you, as you are really struggling here.
This post has been deleted by its author
>And open source developers are fully discovering it when monsters like AWS start to deliver services with the same API of their software and will push them out of any market as they please.
Funny you should say that, Oracle's cloud service uses AWS API, oh the irony.
@LDS
You are being downvoted because you correctly understand copyright. That is discouraged here. Only posts which misrepresent the law receive many upvotes.
It is important for all developers that Judge Alsup’s mistakes are corrected, and that Oracle prevails here. Open source software needs strong copyright to protect it from abuses such as you describe. I cannot comprehend how any supporter of open source wants to change the law to make abuse of open source easier, which is the consequence of a Google victory here.
The APIs issue is a misdirection. Google created this to distract poorly informed people from the facts of the case: which is that Google used a lot of Java code (tens of thousands of line), Google knew it needed a license, and should probably go and get a license (all documented in emails). But instead it decided to chance it. It has now run out of road.
I suggest you actually learn some of the facts from the case and from the judgements.
This has been going on for over a decade and you really need to understand the underlying issues and case law
Both Google and Oracle are tossers for sure. That said, the law is actually on Oracle's side.
There is IP to take advantage of and in some of the earlier testimony from a decade ago damns Google.
Down voting myself or others doesn't change that fact. Its not just an issue of copying the published APIs for interoperability w Java but the intent was to replace Java and its JVM so that Google wouldn't have to pay royalties. Google's 'clean room' version wasn't clean and that's what damns them.
Sorry Ian,
I have read the facts and the case law since this farce's inception. There is a fundamental difference between *source code* and an *API*. It is only by conflating one for the other that any of the cases cited by either Oracle or the appellate court would be anywhere close to on point. Personally I don't believe source code should be copyrightable either, but that's neither here nor there.
An API is a method or process, it has to be the same for interoperability. As I quoted earlier, methods are expressedly excluded from copyright protection.
--------------------------------
An API:
int add(int a, int b)
---------------------------------
Some source code:
int add(int first, int second){
return first + second;
}
---------------------------------
int add(int a, int n){
int result = a;
if(n > 0){
for(int i=0;i<n;i++){
result = result + 1;
}
}
}
---------------------------------
The first isn't copywritable, either of the source code examples _could_ be protected by copyright. So a method isn't while a *particular expression* of that idea can be. Your reference to a 'clean room version' is misplaced in this instance. That only applies to _patents_ NOT _copyrights_, the subject of this case.
As to your remaining points;
Sun released three types of Java; Java EE, Java SE, Java ME. Sun had already released Java SE source code under various permissive licenses. Sun wanted mobile developers to use Java ME for mobile, Google rightly concluded that it wasn't fit for purpose. So Google based it's Android language on Apache's Harmony ( https://harmony.apache.org/ , https://en.wikipedia.org/wiki/Apache_Harmony ). As long as Google didn't call the result 'JAVA' that was perfectly legal. Lots of companies did the same. They threw out the portions they didn't want, added android specific bits and wrote their own bytecode and runtime. The result being a language that would be familiar to existing Java developers and would run on their new mobile device.
It's the same way that database developers can run *most* of the same SQL statements across different database systems. "SELECT * FROM table; " will work on any database that supports the SQL standard. Oracle, IBM, MySql, PostgreSQL, etc. How each database *implements* that statement in code may be different, but no one is allowed to monopolize the;
SELECT command followed by one or more fields, followed by FROM followed by the table, syntax.
Now having said all of that, Google did copy a very small amount of testing code that got released. Google argued that the amount, in light of the code base was allowed as it was 'de minimus'. But that's an entirely separate issue from whether or not API's should be allowed to be copyrighted.
You do yourself a disservice accusing me, or anyone for that matter, of down voting your posts. I'm not saying that this one shouldn't be down voted (lacking the understanding you accuse others of, confusing such simple concepts as patents and copyrights and making ad hominum attacks ;> ).
Hopefully you read things more closely yourself in the future.
I am replying to correct myself. I was mistaken when I wrote:
"Your reference to a 'clean room version' is misplaced in this instance. That only applies to _patents_ NOT _copyrights_, the subject of this case."
Since there is no independent invention exception to patents, the above wouldn't apply. Which makes my other comment:
"confusing such simple concepts as patents and copyrights"
also inapplicable.
That being said, your insistence that: "Google's 'clean room' version wasn't clean and that's what damns them." is irrelevant as there was no such 'clean room' Google publicly acknowledged that they based their efforts on Apache's Project Harmony, which was licensed under the Apache license making what Google did perfectly legal, insofar as source code went. The API was never covered by copyright and technically (with the exception of a rogue appellate court) still isn't.
I apologize for any confusion I might have caused.
Correct, I remember there was some source code that was copied verbatim, with comments and all. Copyright applies to source code, written work etc. Patents apply to methods, processes, etc and *may* cover API's.
I clearly remember Google had lifted not just the API's (Which would not be covered by Copyright) but also the source code (which is covered by Copyright).
No no no no. You've got it wrong. If Google offers to save puppies, your life and spouse, give to charity, and donate millions away, you're suppose to complain because you have an iPhone.
/sacasm
2 wrongs don't make a right, but if you avoid all the rights, you might go a little bit crazy.
yeah I fear I have to side with Google *cough* and Microsoft *cough* on this one, because I can see the tea leaves being VERY VERY BAD if Oracle gets their way in this...
every API could suddenly become PROPRIETARY!!! Writing a competing product would THEN be IMPOSSIBLE... and say goodbye to Wine, and other open source alternatives to commercial products, maybe even Open and/or Libre office.
And THAT would be COMPLETELY *FELT* ('feel' being the OTHER 'F' word)!
(only the lawyers win when all of this CRAP happens)
I wouldn't worry, Oracle are beginning their slow swirl around the plug hole. Their business model is firmly stuck in the 90s and unless they change they're doomed especially given that all their software apart from their core RDMBS is junk and the latter is having its lunch eaten at the low end by OSS and no-SQL DBs. On top iof that very few people are interested in their 2nd rate cloud offering. Relying on high end customers to continue to pay for the RDBMS would keep *a* company in business, but no one of Oracles size.
So when I enter US territory would my Android phone stop working, or would I be billed on a pay-as-you-go basis by Oracle?
Scratch that, this is Oracle. They'll bill me AND the phone will stop working. They'll tell me it's my fault because I'm not on the latest version of the software.
No, I want to see Google crush Oracle, to drive their salesmen before them, and to hear the lamenting of their shareholders.
Actually, I'd be happy to just see Android purged of all traces of Java, just to see the look on Larry's face when he realises that he's managed to knock yet another reason to learn Java out from under the pool of potential developers for "his" platform.
At this point you would have to be mental to develop a phone app using Java, given that if Oracle somehow wins, Google will almost certainly deprecate Java and phase it out, leaving you high and dry.
buying Oracle to get them to go away would still leave that on the books for future court cases
Is this really a problem? Buy Oracle, preserve the lawsuit against their new Oracle subsidiary, and then don't let the subsidiary defend its position. It's still a matter of convincing SCOTUS, but without a bunch of pesky arguments from the opposition. Plus they get to punish everyone at Oracle who supported this infringement case.
If I ran Alphabet I'd be tempted. It'd be expensive, though. And very likely couldn't be completed in time.
I am surprised the Nine Seniles are even looking at this. They must feel something is badly amiss here to even take it as normally they do not take cases like this. My hope would be to reinstate the original ruling from many eons ago with damages awarded to Google for misuse of the courts.
Reply Icon
As pointed out elsewhere here, their cloud offering violates AWS' intellectual property rights if they're correct in their legal assertions.
-=-
Uhm no.
You really need to either talk w an IP lawyer or start to learn the law you think you know.
There is more to this case that what you think you know about copyright law and IP protection.
When the arguments are said and done, it would be remarkable for the conservative courts (Note the little c in 'conservative') not to side w Oracle. The case law is on their side and Google is attempting to play on the heart strings with their amicus brief from Microsoft.
The issue, and sadly so is that whichever side wins, the ruling will be abused by others.
Abused in that it will be misapplied and the only winners will be the lawyers.
Oracle bought Sun because of the Java ME licensing and Java SE licensing.
It would have been cheaper for everyone in the long run if Google had purchased a Java license from Sun.
I suspect that a certain exec from Google didn't want to do just that due to spite.
There is a school of thought that suggests that as capitalism matures it becomes 'rentier' in nature -- that it, it ceases to produce anything useful but relies on extracting ongoing payments -- rent -- for its income stream. There's a lot of historical precedent for this observation although I'd say its not necessarily the system that breeds this mindset, its part of the evolution of corporations.** Anyway, trying to squeeze rent out of owning piece of so-called intellectual property is the modern way to make money so what Oracle's doing is just its business model, it fits in with its attempt to own Linux and whatever else it gets its hands on. We need to push back against this mindset because it does lead to stagnation and decay, we've all seen and experienced innumerable examples of how destructive this is to industry. Its not as if Java is a particularly novel concept as languages go, its just a particular implementation of a Pascal like language that had strict licensing conditions specifically to prevent it fragmenting and so screwing up its 'write once, run anywhere' purpose. The licensing wasn't designed so that a monopolist could lay claim to it and then try to squeeze everyone (after the fact, of course -- you never do this up front because you need to get the base established).
To me, trying to patent a computer language or assert copyright over it is like trying to patent English.
(**I've always thought that corporations would love to adopt the business model of he the IRS or HMRC -- taxation is the best business to be in because its relatively small overhead for the revenue stream collected and no messy products to make, warehouse or ship. Its not such a fanciful notion, either -- selling of taxation powers was historically a way for kings to raise sums of money. (...and some of the outsourced government functions in the UK seem really close to being a modern version of this practice.))
No it's definitely capitalism.
Corrupt third world countries start off like this, but capitalism doesn't. It has been gradually getting heading in that direction.
The taxation analogy is a good one. These super-rich people want to do no productive work at all and receive an income stream from the patents, rights and licences they have purchased.
Where it differs is that tax gets spent on defense, health, schools, roads which benefit a lot of citizens, not just one rich person and their family.
Is it a sign of a moribund system or simply of a moribund corporation? I'd expect the latter. At some point its products, including the IP it tries to collect rent on, become obsolete and businesses which can innovate will displace it. Behaviour like this should be a warning to shareholders to look to their exit strategy.
True, but if it's the latter they're burning money to do it and even if they're buying them up to replace the innovation they can't do themselves they'll smother the companies with their procedures and policies and end up still burning money but more slowly.
"There is a school of thought that suggests that as capitalism matures it becomes 'rentier' in nature"
What you're describing is the natural transition from capitalism to mercantilism that occurs if capitalism is not ruthlessly controlled and constantly slashed at.
Indeed, which Adam Smith warned explicitly about and which members of the Adam Smith Institute appear to have forgotten about. Loads of self-declared "fans of Adam Smith" I have come across accuse me of quoting Marx when I quote Smith to them.
It's a great shame because Wealth of Nations is a surprisingly good read (and free, of course!)
> its just a particular implementation of a Pascal like language
Agree with most of what you said, but to nitpick an arguably inconsequential point- Java's basic syntax and general structure is (as with a large percentage of modern languages) broadly copied from that of C, not Pascal.
"No it's the Google that carried out the biggest copyright theft by storing copies of everyone's website.
If you want to view the content of my website you can write to me and ask permission - address is in the contact page"
1. How do people find your contact page if not through search engines? Or are you content to have a really small user base? (And how do you find those potential users? since most people on the internet seem unable to actually type a URL into the address bar of whatever browser, rather than typing some or all of that same address into a Bing or Google search box, and you don't seem to want to appear anywhere near them - or is it only Google?...)
2. do Google try to claim copyright to all websites, or to be allowed to use any and all website content as they see fit without having to ask permission, or claim any sort of ownership of those websites?
3 do Google really have copies of *everyone's* website on the internet? Including the so-called 'Dark Web'? I can imagine a few ne'er-do-wells not being overly happy at that prospect...
3a Copies of everyone's website? Cop*ies*? Bobby Newmark's on the phone - he wants his Aleph back.
That would be the Authors Guild, Inc. v Google case ( https://en.wikipedia.org/wiki/Authors_Guild,_Inc._v._Google,_Inc. ).
The courts found that Google did *NOT* commit copyright infringement (there is no such thing as copyright _theft_) with it's Google Books search database. So, to answer your question, nope.
Ethical developers and businesses around the world continue to recognize the value of Java and take advantage of our licenses to drive innovation and profit.
The only innovation is on how Oracle decides to screw over the installed customer base this quarter, and the profit is all Oracle's. Ethical developers have moved on from Java and into other platforms, especially now that Oracle wants to charge a per-desktop fee for it. Could you imagine the outrage if Microsoft decided to start charging a per-desktop fee for the Visual C Runtime libraries to each developer or user? "Run Everywhere" has turned into "Get it out of here".
As has been said umpteen times, yes Oracle charge a license for Oracle's Java but you don't have to use Oracle's Java, there are other JVMs out there, use OpenJDK and it's free, there are also other commercial Java versions besides Oracle's.
I guess haters are always blindly gonna hate...
As a basic copyright issue, google's argument fails for the following reasons:
1) oracle owns the java community. While they cannot claim copyright ownership over third party java applications,
they still can claim ownership of the java community.
2) google cannot claim that reimplementation of java api is required for smartphones to work. There exists smartphone platforms
that do not have java apis available, thus killing google's theory that there is only one way to write those declarations.
3) google has history of skipping copyright's limitations and explicitly trying to bypass copyright rules. For example their book scanning operation.
this java api issue is similar google operation where they simply do not care enough about following established rules.
Thus oracle should be on the winning side of this argument.
1) Doesn't matter in this case.
2) Sure, you can have a phone without Java APIs. But then you have a phone without Java at all, as there is no way to do anything with it.
3) See an above comment for details of of the case. Again, that specific case doesn't apply here.
Aside from the rest of your points being misunderstandings or simply back to front (API cloning was the established practice), I'm curious about this:
they still can claim ownership of the java community.
What do you think is the actual basis in law for "ownership of the java community"?
> What do you think is the actual basis in law for "ownership of the java community"?
When sun/Oracle spent the money on marketing their technology to java programmers, the money they spent must have value. The value must be somewhere in the output of their marketing activity, i.e. when they got java programmers to use java technology, that change must have owner. Sun/Oracle is the owner of java community when they caused many people to make decisions to use java.
As the owner of people's collective java usage decisions, sun/oracle can control the java technology.
Let's see where do we start:
1) Who "own's" the java community is completely irrelevant to the case at hand.
2) They didn't, they claimed that implementation of the Java API was:
a) legal, it's an *unprotectable* API
b) required to be compatible with the Java *language*
3) Google's supposed history is also irrelevant, and their 'book scanning operation' was declared legal by the courts.
So about the only thing you've gotten correct/relevant in your post is that this case is a 'basic copyright issue'. An API is a method or system and so is categorically excluded from copyright protection.
See, that was simple.
This post has been deleted by its author
There is a difference between implementing libraries and providing an API and 'using' (calling) an API. That difference is at the heart of the case, you appear to have missed it. Terms and conditions don't come into it, the issue is whether the interface to a library is protected by the same copyright as the library itself. If the interface is protected then nobody can provide a replacement for a given library without infringing on copyright.
Actually the interface (or API) was understood to be categorically excluded from copyright protection (as written about many times previously in this thread itself). Therefore you can't protect an API with any copyright license. So *anyone* can replicate an API regardless of what the creators of that API might have wanted. At least until this pigs breakfast of a ruling.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
