I do wonder
..which Google patents Oracle is angling for, and is expecting to negotiate the rights for, behind closed doors.. Something relating to massively scalable distributed processing or the like? Answers on a postcard.
So, the open sore of Java test-kit licensing that Oracle inherited from Sun Microsystems has come to this: Oracle is suing Google. Oracle has claimed in a legal filing lodged with a US court Thursday that Google's increasingly popular Android operating system infringes on seven patents it owns regarding Java. This is no …
It's most likely that Oracle is only in it for the money... not seeing Oracle wanting to enter the search or online advert biz.
OTOH, maybe Google has a couple of database-tech-friendly bits locked up that Oracle could put to good use in its own products (be it Oracle the product, or perhaps MySQL)?
BTW, to the mods - where's the demonized Larry Ellison icon? If any living soul ever deserved one...
Sorry but Google DO evil. The engineers may be technically brilliant but they are so brainwashed that it's ridiculous, starting with the recruiting process, which is famously known for being weird (while what many people don't realize is that it is designed to attract a certain type of personality).
Let's not forget that for years Google did censor search results for the Chinese government and the internal justification they used is that it would be a bad user experience to click on a censored result and get an error message. Can you believe that? All the employees that did not leave after that are seriously brainwashed or don't have moral scruples.
And let's not talk about the wholesale rape of our privacy, even the people that don't use Google (traveling WiFi-snooping Googlecars anyone?)...
So I seriously hope that Oracle f*** the living daylights out of Google. Sadly, they will probably just settle for some money.
"So I seriously hope that Oracle f*** the living daylights out of Google. Sadly, they will probably just settle for some money."
So the enemy of your enemy is your friend?
Google certainly has some evil streaks, no question. But you also need to have foresight to see that there will be many casualties other than google resulting from oracle's power play.
For all the major "IP" battles that make it on radar, there are many more where medium/small developers have had to exclude features and re-implement their own source code to avoid algorithms. Write software = violate patents. This inefficiency for developers encourages trolling behavior over R&D.
We need to stand up and say no more software patents. I'm afraid even the UK is at risk of defecting from reason despite the evidence of how unworkable these things are in the US.
Mono/C#/CLR is a ECMA/ISO Standard and can be implemented without this BS from anyone since Microsoft has to license everyone on a non-discriminatory basis in order to have it's intelectual property included in a standard. Microsoft DID retain control over certain classes which interface directly with windows internals for a number of good reasons (would have to have licensed internal windows IP).
Java is NOT a standard and is not being considered as a standard. Sun withdrew from the standards process when told they had to give up control and license the IP. Instead they have the "Java Community Process", individuals and groups submit changes, upgrades and Sun (now Oracle) decides what it will include (it now feels like steal) to be a part of the official releases.
As far as I am concerned Java is dead. I am in the process of converting all of my products over to C#/.Net.
Note also please, C# is derived from C++, Java is derived from the Pascal/Modula tree. C# benefits from a longer development cycle (it fixes all of the problems with Java, most notably the idiotic internal implementation of strings in Java). C#, since it compiles (like PHP) works much better with operating systems.
What are you a Novell executive? Mono is garbage (performance and compatiblity wise) and worse if it ever does get popular its status under M$ patents are still very unclear. Why do you think open source projects avoid it like the plague to the point of not allowing Mono apps into free distros just because of patent issues.
*cough* *gasp* *choke*
Sorry, I've trouble breathing. *phew* Laughing so hard. You mean like the OpenXML document standard?
"Converting all my products". What Java products does your astroturfing agency produce?
Also note please, humans derive from fish. Monkeys are carbon based life forms. Whitespace works better on 32 bit hardware. Therefore all your base belongs to Microsoft.
For reasons related to Microsoftts previous behaviour, I'm normally a fan of everything non-Microsoft, and for that reason would prefer that Java lives on in preference to C# or anything else emerging from Seattle.
However the previous post was a rational and well argued response to the issue at hand, so I'm a little surprised it has been downvoted so much. If you really disagree I'd rather read your reasons why than giving this a thumbs down.
Yes, C# is a pun on C++++. But the name Java was chosen because it was better than D, which was hoped to be the successor to C/C++. However, who's the leading light in the .Net/C# dev group? Anders Hejlsberg. Where did he come from? Borland. What did Borland make? C++ Builder and Delphi. What was Delphi? Object Pascal+VCL+IDE. Let's also look back and see that C was inspired by Pascal.
Anyway, in the .Net world, C# is compiled to IL, which is then JITed at runtime. In the Java world, the source is compiled to bytecode (javac?) and JITed at runtime. There are differences in these, but the implementation of a compiler is more important than the fact that it is compiled. so leave PHP out of this, it's just a wannabe Perl (which is also compiled and interpreted) :)
I'm not well up on the implementation of strings in Java, but if that's a reference to their immutability, then C# does it too. Though I did read about problems with String and StringBuffer using lots of memory.
"Let's also look back and see that C was inspired by Pascal."
No it wasn't. C was inspired by B, which was inspired by BCPL. Both BCPL and pre-date Pascal. The common ancestor is ALGOL. Brian Kernighan, who developed the C programming language alongside it's originator Dennis Ritchie, even wrote a famous critique of the Pascal programming language where he declared it "not suitable for serious programming".
I don't know what's in the patents but wonder if they will stand up to post-Bilski scrutiny. That changed the patent landscape so it's possible that the patents will be invalidated. many of Microsoft's would be if they went to trial on them, which is why they target smaller companies that can't afford that.
Taking on Google is a big risk (deep pockets). But that's Larry's style.
Bring 'em on !
...they claiming a patent on recompiling code to optimize performance on new hardware?
I call BS. Compiling to hardware had been the paradigm since the invention of high level languages such as WATFOR FORTRAN 77 in the middle of the last century.
Unless there is some sort of expiration date on prior art, while there might be a design patent on a specific implementation, the IDEA of machine specific code is public domain.
Java's (open source) license proscribes the use of certain versions on certain classes of devices? Wow, that's messed up.
Can someone please define 'desktop' and 'mobile' in context? Is my laptop permitted to run SE? How about my wife's netbook? Or my daughter's iPad?
This really jars with my understanding of the open source ethos. I'm cheering for Google FTW on this one.
Because of the Java Community Process, people always ASSUMED Java was open source. This is very wrong. Java is less open than Open Solaris and we now know how open Open Solaris is.
Mono is an implementation of an open standard. It really is open.
Way back in the day, Java ME (throw in a few [tm] marks) was as much as a mobile device could run... just barely. Desktop size Java was out of the question. Too big, etc....
Queue in the ME licensing and high $$$ royalty deals of those times.
Later on, mobile devices got better, and they continue to become more powerful. Quite impressive these days, I must say - and there is no end in sight.
Some extremely clever and somewhat reckless people elsewhere in the ecosystem began to believe the new devices might even be good enough to run desktop Java [tm] and avoid the ME =~ /license/royalty/every-deal-is-different/ marketplace entirely.
Such radical and disturbing thinking could kill off a nicely lucrative revenue stream that was once directed towards Sun Microsystems, and is now owned by ... Oracle.
"This shall not stand"
Q.E.D... here we are.
"This really jars with my understanding of the open source ethos. I'm cheering for Google FTW on this one."
I guess it depends on the license. I believe Java was released under the GPL2, which is from the early 90's and predates a lot of the stupid software patenting issues we have now, so it may be a little inadequate.
Hypothetically google may have signed agreements with sun so that it could use and modify java source code and not be liable under the GPL. If so, it may be under these agreements that oracle is now attacking google.
Legally, I don't think oracle can back away from any of the decisions sun made to open source java. However oracle may have their own patents which are not subject to the GPL. Even if the code was written by someone else at sun, under US patent law it's infringing upon "oracle's technology". If java infringes on oracle patents, oracle can use the courts to coerce all java users in existence to pay up.
As much as I'd hate the outcome, I imagine oracle has a very good shot at winning this one.
"[.NET is open] as long as you get it from Novel since only then you're covered by a license to the patents."
Wrong. You are mixing things up here. The core .NET (C# and Common Language Infrastructure) has been ISO and ECMA standardized. They are *also* covered by Microsofts "Community Promise" which has legal ("estoppel") value. Microsoft has forfeited any right to sue any implementation for infringement on patents necessary to implement covered standards and specifications.
So right there, if Google had chosen C#/CLR they would not have been in this position. The community promise and the RAND requirements of ISO/ECMA covers this use case perfectly.
Mono goes beyond the core C#/CLI and implements a number of APIs developed by Microsoft for the full .NET. On top of the community promise, Microsoft has granted the Mono project right to distribute without risk of patent infringement. *This* is the pledge which "only" covers Novell customers e.g. anyone who downloads Mono/Moonlight from Novell. This is not to say that Mono *actually* infringes any MS patents - just that they or their customers will *never* be sued for something like that.
Google became vulnerable to patent litigation from Oracle because the patent grant of OpenJDK only covers *full* implementations (no super/subset). Google chose to implement a VM and only support a number of core classes. Presumably to wiggle around licensing (or require device vendors to license) Java ME/MIDs.
Sun *never* relinquished control of Java. They only open sourced the *implementation* of Java SE. The spec was always controlled fully by Sun (and now Oracle) even though they appeared to take advice from the community through JCP. Had Sun allowed Java to be standardized through ISO or ECMA this would never had happened.
"Because of the Java Community Process, people always ASSUMED Java was open source. This is very wrong."
Java is open source, under the GPL license. The separate compatibility test suite is not open source, and the name "Java" is trademarked. You cannot call your software "Java" unless it passes the tests and receives official blessing from the trademark holder (now Oracle).
That was my thought too. If Java is licensed under GPL (even if v 2), no one can restrict where or how it can be run -- that's one of the four basic freedoms. You might not be allowed to *call* it Java if you use in a way Oracle does not like, maybe (I haven't seen the relevant licenses, not want to, honestly), but that would be a trademark issue and not software. Just like you could fork Linux, but could not call it Linux. What about, say, something robotic sounding... :-)
"It seems that Oracle is arguing that Google's Android has violated its Java patents by running on a mobile device, where it's not allowed, instead of sticking to the desktop, where it's permitted."
Now, they seem to be trying to go around that by suing on *patents*... but if the patent forces you to break one of GPL's implied freedoms, then how does it work? Does the license trumps the patent (I'd think so), or vice-versa? If it was GPL v.3, that would be a different story...
"If Java is licensed under GPL (even if v 2), no one can restrict where or how it can be run -- that's one of the four basic freedoms."
If I write code, and it happens to contain patented algorithms, the patent holder has the right to control the use of my code until their patent runs out. It doesn't matter what license I use for my code.
As you can tell from the downvotes on my earlier post, it is unpopular to even suggest that the GPL may be inadequate for legal protection. Regardless, the fact remains that nearly all modern software, open source included, is subject to patents not belonging to the copyright holder.
There is no license which can be invented to override the patents of others. At best, a license can revoke all rights to the patent holders who would try to enforce their patent against the license, which is what the GPL does.
Many OS projects are in a state of denial when they claim "X does not infringe any patents". GPL code does not exempt developers from patent obligations. All this is why software patents should be abolished, people should never be prevented from using and licensing their own code however they see fit. It was wrong to ever grant software algorithms patent protection in the first place.
"Does the license trumps the patent (I'd think so), or vice-versa? "
It's not about one or the other taking precedence, they're different protections. Software patents protect the maths/algorithms, copyrights protect the source code. Owning a patent does not give the holder rights to other's source code, but it can control the commercial use of it.
What article claims (that Dalvik is a implementation of Java SE for mobiles) is far from true. It doesn't implement neither SWT, nor Swing for instance. It is another 'type' of JVM, situated somewhere in between ME and SE, being incompatible with both. It is actually much more similar to the Sun vs Microsoft than this article tries to suggest - Google has created it's own custom version of Java, incompatible with most end-user software; efficient - yes, but totally against any regulations Sun and now Oracle imposed, and also mostly incompatible with other software written in Java.
I expected more from El Reg.
I'm pretty sure the Android implementation has nothing to do with which version of Java is running where. Dalvik is a completely new implementation, it just so happens to use Java source as in input. So it doesn't need to adhere to any licensing restrictions Sun put on their implementations. It's why Google wrote it.
The patents are aimed squarely at how Dalvik operates, the processes and methods used. It's a pretty weak shot by Oracle. Its also pretty risky, Google have a LOT of their own patents, many which Oracle most likely infringe themselves.
Really though Oracle have kicked themselves in the teeth with this one, Android was making Java cool again. Now they do this which just scares people off.
I am surprised by how stupid Google was with the way the did Java in Android. The business thing to do was to license j2ME and build their APIs on top of that. If they had done this, there would be no case. More importantly, this would be transparent to developers/users since it would work the same. Now regardless of the merits, its going to cost them one way or the other.
And it's only marginally about Android. It's not about any intellectual property Oracle acquired from Sun.
It's about the launch of Windows Phone 7 coming up in a few weeks. Bill and Larry are co-billionaires for charity now. They're playing on the same team - for the children, for their legacy. That doesn't mean they've given up playing dirty. Far from it: it gives them a moral certitude that they're working toward a worthy end that justifies almost any means. Bill may not be CEO of Microsoft any more, but that's where the bulk of the billions he's giving to charity come from.
Microsoft needs to sow some Fear, Uncertainty and Doubt in the mobile and tablet markets so that they can market their products as being free of that taint and so gain an opening that frankly isn't there otherwise. If they can't get a leg up here they're facing barriers to entry that are insurmountable. People Luurve their iPhones and iPads - 95% of owners would recommend the product. Android is swinging up with a market share growth that defies gravity and owner satisfaction ratings that are nearly as high. Without a rash of IP suits against iOS and Android, some legitimate security disasters there, and so on, they've just got nothing but a product that on its face is, er, inadequate.
The Virtnet lawsuits are about the same thing, by a different route. Given a near certain loss in the lawsuit but the potential for a long delay, it's easy to see the negotiating team work out a settlement that goes something like "Look, we could tie this up in the courts with appeals and whatnot for another decade or so... or we could settle for a decent figure as long as you agree to go after our preferred target next." The settlement was probably cheaper than their ad budget for Windows Phone 7, and Microsoft is not one to shy away from pouring money on the bonfire in the name of strategy.
There will be more of these suits filed from every angle. And the puppet press will cover each one from the same "Evil Google, Evil Apple" point of view. They're all bunk.
Stiggy - Every Java contract had/has a "field of use". This was so that you could limit J2SE to desktops only and not have it creep into mobiles (as some customers tried) and also stop J2ME (the mobile version) making its way up onto desktops and servers. One of the issues was that J2SE was royalty free and the growth was in the mobile/handset side so the FOU was effectively protecting your revenue stream.
Now in the age of tablets, large screen handsets etc the divide between what is a desktop and what is a mobile device is very, very blurred and the FOU is maybe out of date with current technology.
SCO had a lot of problems:
* an incompetent legal team (going after IBM's own Nazgul? WTF were they thinking?)
* no money to speak of (in the tech field, and in spite of Microsoft's $149m cash injection)
* shaky legal grounds (turns out they never owned the copyright)
* they weren't really suing over IP issues - the board at SCO just wanted to kite their stock one last time before it face-planted for good.
Oracle has none of these problems, and it is run by a calculating bastard with an ego the size of Canada (that, and Oracle has a bank account big enough to actually -buy- Canada).
That is what should scare the unholy shit out of any company doing business in the tech field (at least in the US, and in the EU before long if things continue like they appear to be).
"Why not include his holiness Saint Jobs in your conspiracy theory as well? "
It seems Larry has been taking a few lessons off Saint Jobs in taking previously licensed stuff back into private hands. Like Apple with its operating system, Larry has started to close off Solaris and Java.....
I'll let someone else build on this :-)
Dalvik doesn't run Java byte code. It's not a subset of the JVM or a superset of the JVM. It's a completely different runtime that uses a register based virtual instruction set as opposed to the stack based set underpinning the JVM. It's analogous to the JVM and to .NET's CLR and to the iPhone's LLVM.
The only way Java comes into the picture is because it is the programmer's interface to Android. The programmer writes Java code which is turned into Java bytecode. Then an additional compile time step transforms the .class file into a Dalvik .dex file. Then all the .dex files are packaged together to form an application package .apk file. All this happens on the developers machine. By the time it touches an Android device there isn't single piece of Java in it.
Android does implement a subset of Java public APIs (which have been implemented countless times before, e.g. Classpath, Kaffe, Harmony, Skelmir etc) and therefore the runtime environment is familiar to the programmer.
That's about the extent of it. Google have never claimed it is Java and it isn't Java. This is why Oracle can't sue on trademark or licensing grounds. Instead they're trying their luck with certain patents. I expect many of these so-called patent violations don't even cover Android at all. They cover the SDK where much of this translation happens. I also expect that Google has plenty of its own patents, especially concerning web technologies that Sun / Oracle have violated. The most likely outcome therefore is Google tell Oracle to get lost, a public shouting match occurs and then things get settled before anything reaches court.
making Java "cool" is not the same as giving Oracle revenue as a result.
I'm honestly not sure where I stand with this one, as my HTC Desire is indeed very cool, but I can see things from Oracles point of view.
Someone else suggested that Java licensing seems to be a little confusing as a result of the increased "blur" between what is a mobile device and what is a full fledged computer. I hope this is sorted out not only between Google and Oracle, but hope that it is made clear to all developers exactly how open Java is, so we all clearly know under what conditions we can use it or not.
Biting the hand that feeds IT © 1998–2019