Jury mulls verdict in Oracle-v-Google Java spat The jury in the copyright trial between Oracle and Google over the use of Java in Android has retired to consider its verdict after closing arguments from both sides. Judge William Alsup gave the jurors a 21-page document with guidelines of how to review the case, telling them that they should ignore the arguments of lawyers …
"Just because Sun didn’t have patent suits in our genetic code doesn’t mean we didn’t feel wronged. While I have differences with Oracle, in this case they are in the right. Google totally slimed Sun. We were all really disturbed, even Jonathan: he just decided to put on a happy face and tried to turn lemons into lemonade, which annoyed a lot of folks at Sun."
http://nighthacks.com/roller/jag/entry/my_attitude_on_oracle_v
Ah yes, a disgruntled ex-employee who did his best work at Sun, of course he must be a legal expert on the case,,,
How about the guy actually in charge of Sun when Android was released and they negotiated to make a joint Java phone, even talking about doing a joint venture using an Android device, Jonathan Schwartz...
"Schwartz described that the community talked about open APIs and competing implementation. He affirmed that basically everyone would have the same set of APIs, but each company would create its own products, the virtual machines specifically, to go off and perform.
Getting into what is at the heart of Google’s case — that the 37 Java APIs in question were free for its engineers to use on Android — Van Nest asked Schwartz if the Java APIs ever sold or licenses separately from the language.” Schwartz replied instantly, “No, of course not.”
Van Nest also asked Schwartz if the Java APIs were considered proprietary to Sun, to which Schwartz also replied no, adding that “we would have worked very hard to say that wasn’t true.”"
This first part of the case is actually about the copying the naming convention and structure of the Packages and Methods that form the API after all - not really about dubious copying. But then how can you implement the Java Programming Language without using the names freely published in the API? Hmmmm...
Where do you think any form of Java would be headed today on mobile devices if it wan't for Android? It wasn't as though if there is one phone OS that allows writing apps in Java then there is not allowed to be another (one running true, certified Java would surely have been far better under that thinking). JavaME was never a big draw for phones, nothing like in the same league as IOS or Android or Blackberry.
Thumbs all around. I had a phone with the Java on it, actually. An old Sony Erricson, it was... Games and stuff. The most godawful dog it was to use Java on it too, as that was a separate module that was only loaded while playing games, and unloaded after. Oh, and the music player. 20s wait to get it up. Faster than a pill, but certainly not faster than waiting, waiting, then pulling the battery in annoyance.
Posted Anon ?
You asked: "Where do you think any form of Java would be headed today on mobile devices if it wan't for Android? "
Where its always been. Licensed to Sun now Oracle.
If Google loses, its a double edged sword. Can you say indemnification?
This post has been deleted by its author
Gosling seems to have a self-serving view of patent law. When Sun was taken to task by IBM for nicking their RISC IP, Gosling said that all the patent meant was that "if you make something simpler, it will go faster"... which is on its face ludicrous. It is like saying that Java should have no patent protection because if you enable something to be written once, it will be more efficient. John Cocke, IBM inventor of RISC, didn't win every engineer prize in the world for writing down "if you make something simpler, it will go faster" on a napkin. RISC was a genuine innovation.
Now that the shoe is on the other foot, with a much more ambiguous Java IP licensing structure, Gosling thinks Sun/Oracle is clearly in the right. They may be, but Gosling seems to be fine when he or his friends borrow other peoples work, not so fine when other people borrow their work.
How is that statement "well out of context"? It is exactly what Gosling said about Cocke's RISC patents and instruction set architecture. The same architecture which won Cocke the National Medal of Technology, the National Medal of Science, and every other award. According to Gosling, RISC was so obvious that instruction sets could not be created in any other manner (despite the lengthy history of CISC doing the task is a completely different manner) and Sun would have had no choice but to inevitably create RISC in their servers. Gosling and friends just threw a fit when they knowingly took RISC without payment. If you are Richard Stallman and you think all patent law is absurd and antithetical to freedom, etc, then you have an argument. If you then turn around and say that the technology you created is patentable and people should pay you for using it, then you are James Gosling. He likes patents when they have his name on them.
James Gosling also went on to boast that his job at Sun was making sure all the J2ME implementations stayed compatible. A job Sun & Gosling failed on an epic scale.
Instead of WORA we have companies making money from testing J2ME apps against thousands of different devices for other companies unable to afford their own device library.
Meanwhile Sir Trolls A Lot ponders where Java would be without Android. It would be exactly where it was before Android, sitting almost|totally unused on billions of low end phones, running the OS on none of them. It would be running on no iPhones, no WP7 phones, no Win8 devices and there'd be no Android for it to run on. Java would be dead on mobile.
Gosling shares Suns lack of vision at the time, unable to see past the licence fee for a box tick on the advertising that no-one actually uses because it's not good enough to use. Unable to see that lifting those artificial restrictions could allow so much more. At a time when Apple was quietly creating iPhone and completely discarding Java, Sun & Gosling were intent on selling an ageing crock.
Sun made Java what it is today. A barely relevant mess.
Agree, Sun could never decide if they wanted to be Apache, Red Hat, or Microsoft. I am so thoroughly confused by what is licensed in Java and what is not licensed. You need a license for the core packages and the JVM (sometimes), but not for all of the class libraries and APIs (sometimes). What good are the class libraries and APIs if you cannot use the core packages and JVM? Sun makes people think fondly upon companies that just charge you for everything. Give me a quote so I can make a decision, but don't let me use something only to find out I have to pay for it after the fact... which might have led me not to use it at all. Trying to split the difference between Richard Stallman and Bill Gates is going to be confusing.
Just SAMBA? Think of anything and everything.
If Oracle wins this practically prohibits any development of interoperable software without taking a commercial license. That will be the end of the software industry as we know it affecting both free and paid for alike. It is from the "I love the smell of monopoly early in the morning, it smells like 300% profit margin" book so Oracle will not be letting go and it will get some Amicus Curae from the usual suspects in the next rounds.
If Oracle loses on all counts it will have an interesting effect as well - mostly on open source software. Presently a lot of the "do not link", etc is enforced through the gratuitous rubberstamping of GPL notices on includes, headers, etc which are in fact API definitions. If Oracle loses, commercial software will be able to ignore these with impunity and communicate with GPL software in ways that are considered "unacceptable" at present. Ditto for the other way around which is usually less of an issue at present.
Except interoperability is protected by law, and in SAMBA's case doubly so by the antitrust measures applied to Microsoft.
So there's nothing to worry about in SAMBA's et al cases. SAMBA is completely legal and even protected by law.
On the other hand if your goal is to copy the API to fragment the platform, like Google did in this case, then you should start thinking of seeking legal advice.
Interesteing, so the reason that Google copied the API wasn't to allow the Java Program Language to be used for e.g. interoperability it was so that they could fragment Java - perhaps destroy Java completely?
Missed that bit in the trial.
Samba can only work by copying the API. Where is the legal definition of one company copying the API is copying, another is interoperability when they are doing the same thing? Does the law have a section such as "2.b) ii) Unless the program creates fragmentation, under which case it will be illegal".
Do you think that the goal of Samba was that it should be 100% compatible so that it did not cause fragmentation of the protocol or that it was just trying to do a workable implementation?
@AC
I never said destroy Java - ACs overdramatising, what's new - but the Android's Java clone is certainly fragmentation since it doesn't align with either the Java SE or Java ME platforms - it's a new platform.
As for Samba, they didn't copy the API since they don't have the source code for Microsoft's file sharing services. They reverse engineered it. Reverse engineering for interoperability is covered under fair use.
Google is not attempting interoperability or a "workable implementation" of existing Java platforms, you can't take a Java SE or ME app or even compile it's source code and run it on Android, you must code it specifically for Android. That's not interoperability.
Google is not attempting interoperability or a "workable implementation" of existing Java platforms
But does the use of the API necessitate that interoperability? The problem, of course, is that Java is both a language and a platform. If the language comes with no strings attached but the platform is encumbered then this is very much like having your cake and trying to eat it. I seem to remember similar discussions years ago when Sun was forced to open the development of the platform as IBM and others threatened to bless a clean room implementation and, thus, deny Sun any future say in matters.
"you can't take a Java SE or ME app or even compile it's source code and run it on Android"
I couldn't compile my Java SE game editor build on ME at the time Sun & Google were negotiating. Apparently it's OK for Sun to artificially fragment Java. Today I couldn't guarantee the Swing UI source would compile under ME, Swing support supposedly is there but partial.
At the same time more than 75% of the Java source is shared between my games editor and the Android game build, exactly the same source - this shitty language doesn't support conditional compilation, which normally pushes that shared source above 90% in other languages. Only the UI layer changed - as it would have to since it's doing a different job and there's no point reading sensors my PC doesn't have and a mouse my phone doesn't have!
Some of the tools I use only need a shell to inject command line parameters and an Activity stub to make them launchable on Android. They compile and run quite happily with that. Not a useful thing to do though.
You grossly exaggerate the level of incompatibility and ignore the fragmentation baked into Java itself.
"As for Samba, they didn't copy the API since they don't have the source code for Microsoft's file sharing services. They reverse engineered it. Reverse engineering for interoperability is covered under fair use."
Wow, so if the information is published, in the public domain, in a book and GPL'd then you can't use that information unless you do it in 100% compatible mode, however if you hack (in the traditional sense) a protocol, api etc then you can do what you want with it? Is that case law?
"Google is not attempting interoperability or a "workable implementation" of existing Java platforms, you can't take a Java SE or ME app or even compile it's source code and run it on Android, you must code it specifically for Android. That's not interoperability."
Android said it was trying to be compatible with the Java Programming Language for a subset and superset of the Java Specification. Sun explicitly and implicitly allowed that. They just stated that it must be fully compatible to claim it is Java and use the coffee cup logo.
Oracle aren't suing over fragmentation or saying that Goole explicitly needed a licence per se. They are saying (in this part of the trial) that they broke copyright law as the API is creatively names and structures (SSO) and therefore should be afforded similar protection to a novel.
So this affects every type of interoperability - protocols, APIs, interfaces, xml data transfers. In fact anything - even if you have been entitled and given permission to use it in the past, could be stopped using copyright law if this were to succeed (unless copyright was explicitly waived as well)
Samba can only work by copying the API. Where is the legal definition of one company copying the API is copying, another is interoperability when they are doing the same thing? Does the law have a section such as "2.b) ii) Unless the program creates fragmentation, under which case it will be illegal".
===============
Sorry, Samba is an implementation of protocol (CIFS), not an API.
They are fundamentally different, and not related.
Try again.
"Interesteing, so the reason that Google copied the API wasn't to allow the Java Program Language to be used for e.g. interoperability it was so that they could fragment Java - perhaps destroy Java completely?"
Then why did Google try so hard to licence Java? Not licence the artificially crippled J2ME crock but real J2SE, with real test compatibility. Android would have become J2ME + Android extras. Extras everyone including Sun would be free to bolt onto Java everywhere.
Sun deliberately fragmented Java with the ME *variants* long before Google came calling. But they crafted a licensing plan to prop up that fragmentation that failed to remember Moores Law, that eventually J2ME would have no technical justification to hide behind.
Sun was offered a way out of their own trap and lacked the vision to take it. They screwed themselves.
I dont know...
You have this thing called fair use... And that's not a blanket statement saying that you can't violate fair use.
As well you don't destroy interoperability ... You just hav to think a bit about it...
Did Amazon's one click patent destroy the web?
I'm still not sure what the consequence of that copyright would be. On the face of it, it is legitimate to want to be able to assert the authorship of the API which is just a specification, i.e. no one else can claim to have come up with it. This puts them up there with technical specifications like the HTTP protocol. As only implementations can be patented, it is, er, patently obvious that you cannot patent the specification. Copyright would allow for licensing of the specification for derived works but this might be considered to apply only to extensions of the specification itself rather than implementations of it. As is noted elsewhere, the point of APIs is to guarantee interoperability.
If IBM would have acquired Sun, in my opinion, this whole mega-lawsuit threatening to change the nature of software development and the future of Java would not have occurred. IBM would have settled it quietly and wouldn't be going for the $6 billion jugular. IBM doesn't like showdowns whereas Oracle seems to always be looking for the next scorched earth battle.
IBM likes showdowns as much as the next but it has become extremely adept at that which Oracle is aiming for: leveraging hardware, software and services of each other. That it has a long history of co-operation with open source from releasing its own software as open source (Apache, Postfix, etc.) to contributing actively to existing projects (most notably Linux and Java and OpenOffice), shouldn't distract from the fact that if IBM seems a $ 6 billion opportunity it will go for it. However, such lawsuits can adversely affect customer relationships. It may have been in weighing up the two - costs and benefits of litigation versus affect on sales - that decided IBM against the purchase of another hardware division with uncertain software licensing business attached.
This post has been deleted by its author
I think we agree. IBM is not a charitable organization and wants to get paid, but they are not going to burn bridges trying to get every last dime to the point of pushing people away from their tech. If they can find a way to make a revenue stream while allowing whoever is licensing their tech to make money, they don't have any problem with it. IBM embraces the ecosystem approach. Oracle just likes to fight for fighting sake. Their recent wars will probably end up hurting them as well as who they are fighting. HP vs. Oracle will hurt Oracle and destroy HP Itanium. IBM will be the winner of that fight. Oracle vs. Google will hurt Java (Oracle) and Android (Google) to an unknown extent. Good luck getting any software providers to embedded Java at this point. The winner of Oracle vs. Google will be Apple and maybe Microsoft.
Assuming things like RangeCheck get tossed as de minimis, or at worst get a proportionate fine (ie small in keeping with the de minimis quantity of infringing code):
BS&F's Structure,Sequence,Organisation invention may be surprisingly easy for Google to work around, particularly if they can get delay with an appeal. Easy *because* the VM and classloaders don't really care how that hierarchy is organised or what names are used, only that the symbols in Dalvik bytecode match the symbols in the libraries on the device.
The toolchain already translates compiled Java to Dalvik, a perfect place to perform a lot of symbol remapping. Remap the Java API structures to an arbitrarily organised set of entry points - sorted flat lists would seem most immune to copyright claims and most apps wouldn't notice. But you've just knocked out Sequence and Organisation from the claim, possible Structure. Once you abandon total compatibility it's easy to carry on throwing it away.
Should buy enough time to make more radical changes like completely excising Java if that becomes necessary.
I don't think any company should be subject to IP theft regardless of who it is. The arguments here are trite. Being able to simply decide to misappropriate software will have a chilling effect on IP development. You think Google has created anything original? No read the news, someone is taking them to task almost every week over some misdeed, it is unprecedented. You can't keep defending the indefensible. Why should Oracle concede anything, after all both Java and ME have been around awhile supporting many devices, like it or not. They have a right to protect their mobile Java investment, not Google.
You're betraying a woeful lack of knowledge of US software copyright law. It has evolved exceptions that can be read as recipes for how to copy software and get away with it, if you want to look at it that way. They actually evolved to prevent the chilling effect absolute, unfettered copyright protection would have, principally the ability to lock out competition and unfairly lock in customers, to allow interoperability - which prevents an infinitely fractured market for software (which damages end users).
The issues in this case are not whether Google took Sun IP, it's whether they took more than the law allows, whether they followed the 'recipes' correctly so it was legal to do so, whether Sun pissed away it's own rights to stop them by its own behaviour and a flawed licensing scheme.
Google tried walking on the bleeding edge of software copyright law, this case is testing how well or badly they did. That BS&F had to recycle the invented crap from the SCO fiasco to get as far as a jury should tell you Google did pretty well. Don't confuse actions that seem wrong with actions that are illegal, they aren't always the same.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
