Oracle blurts Google's Android secrets in court: You made $22bn using Java, punk An Oracle lawyer has blurted out in court how much money Google has made from Android – figures that the web giant has fiercely fought to keep secret. And those numbers are: US$31bn in revenue, and US$22bn in profit, since 2008, when Android was launched. This money comes from Google's cut from sales made via the Google Play …
Aye, I don't get this. IF Oracle did somehow win, the chaos that's follow, including against Oracle themselves from various language developers.
"nice output statement you got there, you take a string as input, and output it to a file/screen/stream/all of the above? that looks very familiar. If fncRangeCheck was worth 6.5Billion to you, then printing 'hello world' to the screen must be around.. 26Billion?"
"If you allows APIs to be copyrightable, you're going to have a bad time."
That boat has sailed; an Appeal Court's already ruled Oracle can copyright its APIs. However AFAIK it's not been decided whether Google's use constituted fair use or was infringing.
They are all from AT&T. Maybe now SCO?
This would be the end of Linux etc.
How can a Java API be different from a C API, in this respect?
And Google just copying the Java APIs like they did must be one of the dumbest things they ever done. Especially as most of those APIs are not very good.
> And Google just copying the Java APIs like they did must be one of the dumbest things they ever done. Especially as most of those APIs are not very good.
---
This is plain silly. Only someone without any clue about the topic can write down something like this.
If Java's APIs weren't very good, then Java never would have become so successful as it is today. It is not a coincidence at all that Google chose it as the main programming language and runtime technology in Android. Even after 20 years after its birthday, Java is the most popular programming language, by a wide margin, see TIOBE (also http://motherboard.vice.com/read/2015s-most-popular-programming-language-was-good-old-java). The ecosystem is just that good around it (third party libraries, compilers, IDEs, debuggers, runtime instrumentation tools...etc) Other, fully open-source/free toolkits come close to it in several aspects but never in all of the aspects.
It is also not a coincidence that Google is in the process of rebasing Android on OpenJDK (the standard, open-source implementation of Java which is also the base of Oracle's own Java version). It has now become obvious that Google will have to pay for using Java, so there is no point in using an inferior, old implementation like Harmony. They might as well update Android to Java 8 and use the latest language features and performance improvements + the miriad advanced tools based on standard Java.
> They are all from AT&T. Maybe now SCO?
> This would be the end of Linux etc.
Ironically, it is Google that could lead you finding the answer to such questions as posed in your title. POSIX is a family of standards that were developed by and belong to the IEEE Computer Society.
> And Google just copying the Java APIs like they did must be one of the dumbest things they ever done. Especially as most of those APIs are not very good.
Google didn't 'copy' the API, they _used_ the API, specifically a subset of Apache Harmony.
"Google pleaded with the court to remove the financial information from the hearing's transcripts, and place the numbers under seal. "
Good luck with that. The numbers are now in the public domain. What are you going to do google? File a takedown request with yourself?
After literally a picosecond's diligent hacking, I found a little-known site that is making the information public, the scofflaws:
https://www.google.com/search?q="%2422Bn using Java"
Top link. Boxed out, no less, to highlight it.*
Thank goodness Google are safe from the risk of anyone disseminating this information.
*- I wouldn't put it past Google to censor the results and spoil my joke, so for anyone in the far-distant future reading my comment: at time of posting, that search term returned umpteen links to this story, with this very page on the Register at top of the list.
Sadly, it's not a misuse anymore. http://theweek.com/articles/466957/how-wrong-definition-literally-sneaked-into-dictionary
Methinks you (and the author of that article) do not understand the rôle played by dictionaries. A dictionary is descriptive -- it describes the language as it is used, possibly with some comment on whether that usage is considered correct or appropriate, but often not. The fact of a word being listed in a dictionary with a particular meaning means that some speakers of the language do use that word with that meaning -- not that they are correct in doing so.
Even so, the article you cite does note that some of the dictionaries the author consulted did give usage guidance, and that one of the Oxford dictionaries noted that the word is sometimes deliberately misused for effect (which doubtless leads to its being listed with the incorrect meaning in lesser works, and propagates the error).
"A dictionary is descriptive"
I clearly remember being taught in grade school - and this is one of the very few specific bits of knowledge which I recall learning in grade school - that some dictionaries are descriptive, and others are prescriptive. I presume, possibly mistakenly, that this distinction still holds.
Whether those prescriptive dictionaries are actually successful in causing people to use the language "properly" is a different matter entirely.
I clearly remember being taught in grade school - and this is one of the very few specific bits of knowledge which I recall learning in grade school - that some dictionaries are descriptive, and others are prescriptive. I presume, possibly mistakenly, that this distinction still holds.
Some dictionaries claim, or aim, to be prescriptive. They are successful to the extent that they control language use by actual speakers and writers, which is close to not at all.
Depending on the patient, prescriptivism varies between a wistful fantasy and a derangement of the mind. While there are some good arguments (and many poor ones) for preferring certain usages over others, and while even the most dedicated descriptivists have opinions on what constitutes appealing, careful use of language, attempts to bully people into conformance in its use have never succeeded, and declarations of "correctness" have always been ill-founded.
The Académie Française is charged with maintaining a dictionary of the French language. This is an authoritative definition of the French language and so must be considered prescriptive. Not even the august members of the Académie Française themselves take the idea very seriously.
While I bare no sympathy to what Google has become, I find it difficult not to notice that:
Neither the play store, nor ads are fundamentally bound to java, infringing on java copyrights (if that happened in the first place) or anything to which Oracle may be theoretically entitled for a chance for remedy.
While we all know that Google uses Android as a scorched earth tactic to ensure that nobody can get even close to their dominance in search and ad slinging, you have to prove that this is its exclusive function via appropriate discovery. Oracle has failed to do that (I do not think they would be allowed by most judges to go on that fishing expedition anyway). So from that perspective, even if there is infringement (which I doubt), this all smacks of racketeering. Oh, such a good 22Bn business you got there, wonder if you would like something to happen to it. Even if Oracle is entitled to any remedy (*), it should be derived from Android licensing agreement numbers, not from Playstore and ad slinging revenue numbers. Oh, by the way, it will be interesting how much do those contribute to G00G bottom line.
(*) A raft of recent cases in Eu went solidly the other way. If it is essential to implement functionality for a system, you cannot copyright the shape essential to implement said functionality. It fails the expressiveness test. Lego, London Taxy company, etc - you name it.
Bare = take. Thanks - lesson for the future - should not post before 4th espresso. Damn... there is no coffee icon. Probably so that Oracle does not ask El Register to hand all of its measly profits for violating their "hot java" copyrights.
Oh well... beer as closest equivalent (especially on a Friday afternoon).
Intolerable!
Me seem to recall Red Hat, having to change their patching procedures a while back because another software company (naming no one, mind you) would just mooch off their patches to maintain their own, "Unbreakable", distribution.
Now, the legality thereof in GPL land is probably on the moochers' side, but when that same company cries grievances to high heaven on usage of a supposedly open sourced language, I find that ironic ;-)
The figures I was able to find shows Google making about $60 million in revenue from the Play Store in Q1 2015. So it is unlikely to add up to more than a billion or two since it started, or less than 10% of that $22 billion in profit.
Another source showed Google claiming last year they'd paid $7 billion out to developers. That would imply $3 billion in profits since they take a 30% cut, but since Google pays a share of that revenue to carriers to encourage them to support Android they probably get less than that so that also points to a number of no more than $2 billion or so.
That leaves $20 billion in profit from ads alone.
And how much does Oracle owe Microsoft, and Apple and IBM?
And how much does Microsoft owe the POSIX lot?
And how much does every software company owe every OS company?
And how much does every one who has ever written software owe anyone who has ever provided an API to that software?
Or... we could just leave it as it is, interface is interface, not code, not copyright-able. The attempt to use patents to confuse the issue pretty much shows the weakness of Oracles own case, so even a judge that doesn't really understand the technical aspects can see the game playing.
Oracles (subsidiary Sun), once cloned Windows:
https://en.wikipedia.org/wiki/Wabi_%28software%29
"Wabi was a commercial product from Sun Microsystems that implemented the Windows Win16 API specification on Solaris; a version for Linux was also released by Caldera Systems. Wabi supported running applications developed for Windows 3.1, Windows 3.11, and Windows for Workgroups."
If Sun can clone Windows API then Google can clone Java API, or rather use Dalvik VM that implemented a Java compatible API.
Oracle cannot pick and choose its history.
Why do you think everyone who can is abandoning Solaris? Or Oracle's version of MySQL? Or Oracle's database and software platforms for that matter? When Oracle bought Sun, I knew that was pretty much going to be the end of new Solaris deployments, and that's exactly what's happening. The problem is that Oracle never gives anything away long-term; eventually they will take it back if they see a way to do so.
Oracle is doing the same thing with Java -- they're now changing the licensing of the runtime environment for customers who use it directly in certain embedded devices -- it used to be a free platform, still is for most uses, and that's why software was written in it, CS programs are based around it, etc. I know a few embedded device manufacturers who are taking the painful step to move years of Java code onto something else just so they don't have the uncertainty of having to suddenly pay millions of dollars when Oracle comes to collect license fees.
Java was never a free platform for embedded devices. Even though Java SE was open sourced, there were field of use restrictions that prohibited it's use in embedded devices. For that you needed a separate license, and pay royalties. Google didn't want to pay pennies per phone in royalties so they just ripped off Java instead, against the advice of some of their senior (and former Sun) staff
The Lions commentary on UNIX 6th edition has prominent copyright notices for the kernel interfaces.
http://www.lemis.com/grog/Documentation/Lions/
This means that fork() open() read() write() and close() are APIs owned by Bell Labs (which I think finally ended up in Nokia via Alcatel/Lucent).
Is Nokia now able to make a claim against Oracle? In this case, that is the one claim that the courts should allow, in full, plus penalties for willful infringement.
Android and the OpenJDK project
"Google intends to address the fact that, on December 24, 2015, Google released new versions of the Android platform that are expressly licensed by Oracle for use by Google under the free, open source license provided by Oracle as part of its OpenJDK project."
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
