Slashdot Mirror
Google Says 3rd Parties Would Be Liable For Java Infringement
angry tapir writes "Third parties, not Google, would be liable for any Java copyright violations in the Android mobile OS, according to a filing Google made in the US District Court for the Northern District of California. Oracle sued Google in August over a number of alleged Java patent and copyright violations in Android."
Free as in "free to pay microsoft and oracle".
I wonder how Oracle will go suing Android integrators in Korea, Taiwan and China?
http://michaelsmith.id.au
so basically leaving the customers (phone makers) twisting in the wind?
Microsoft promises to take legal responsibiliy in the case of patent lawsuits resulting from use of their platform. Interesting times we live in...
Download free e-books, lectures, and tutorials at bookgoldmine.com
Mayber some of the following paragraphs from tfa would have fit in the summary:
"""
"Any use in the Android Platform of any protected elements of the works that are the subject of the Asserted Copyrights was made by third parties without the knowledge of Google, and Google is not liable for such use," Google attorneys wrote in one of 20 defenses to Oracle's amended complaint.
Another defense states that Android was "created independently and without reference to any works protected by the Asserted Copyrights."
Elsewhere in the document, Google discusses the formation of the Open Handset Alliance, the coalition of vendors, including Google, that worked together to develop Android.
The filing also notes that Android can be freely downloaded and developers are free to modify the source code to suit their needs.
In addition, Google states that Oracle's patent claims should be denied under the doctrine of misuse. "Oracle has come to the Court with unclean hands due to its practice of requiring licensees of its purportedly open software to pay for licenses to items not covered by Oracle's alleged intellectual property in order to receive a license under Oracle's alleged intellectual property."
"""
?
Some of my favourite people are from th US; Vonnegut, Chomsky, Bill Hicks.
From TFA, Google filed 20 defenses taking an 'everything but the kitchen sink' approach. In other words, they listed every defense they could conceive of, so that Oracle has to defeat each individual defense. If one fails, Google will then rely upon the others.
Furthermore, it's a strategic move - if the others were responsible, Oracle could find itself in the position of trying to sue either companies with much smaller bank balances like the Open handset Alliance or some 20 year old student. That's a lot less attractive than a bumper payday from Google.
It helps to have "friends" in "high places."
ELOI, ELOI, LAMA SABACHTHANI!?
Here's what the article actually says:
"Any use in the Android Platform of any protected elements of the works that are the subject of the Asserted Copyrights was made by third parties without the knowledge of Google, and Google is not liable for such use," Google attorneys wrote in one of 20 defenses to Oracle's amended complaint."
That is, "we are not responsible for any violations added by the third parties". Well, duh.
It may not work, or it might be full of bugs and you have to pay more for it... But since they cover you for patents, that makes all of the difference? yea, thats the route I would recommend. NOT. I look for solutions. If I can not find solutions, I write them! Go ahead and try to hunt down all of the corps that might infringe your patent.... Their software is not shared. Your access to the source is non-existent. Open source only exposes the code if it is shared. It doesn't have to be. And programmers are not that hard to find. They are cheaper than most software, if kept in house. Which would give you all the advantage any how. Custom answer to custom request. Arrg...
Having a flow chart would clear up so many issues surrounding patents. Here is the question:
What line of reasoning must hold water before a patent is deemed valid?
If you read this story, you realize that each party is asserting their position as the valid one. To me, the confusion surrounding this topic is hitting me hard. A flow chart would help out a lot.
Is it "third parties" as in "anyone remotely connected to Android, including users and developers", or "third parties" as in the Open Handset Alliance which comprises 78 companies?
Methinks Google isn't saying, "Look, don't sue us, you should go after the users", but rather "Oh, you want to sue us? You'd better be prepared to include 77 other defendants with big pockets."
"We are Microsoft. You shall be assimilated. Competition is futile."
GO FUCK YOURSELF!
For actual coverage of Google's counterclaims, I suggest Groklaw: http://www.groklaw.net/article.php?story=20101111114933605
It's been widely reported that there's a duplicated file, and indeed it there is something close to that. BUT! One thing you'll find missing in Oracle's Exhibit "J" are the package headers at the top of the file. There's a good reason for that. On the Android side the file is in package org.apache.harmony.security.tests.support.cert, in directory support/src/test/java. You can see this in the git repository for android. It's sitting in a directory of test support classes.
So the matching file that we have here is part of the test suite to ensure compliance with the interfaces. It is NOT part of the implementation itself. So the real question is, is it OK to have this kind of file sitting in the test branch, to ensure that the real implementation of it complies?
The fact that the package headers have been removed and that this file is from the test suite can't be anything other than a deliberate attempt to deceive, well, someone. ;)
It's rather unbelievable that with thousands of stories out there on this file nobody is talking about WHERE it fits into the android tree.
Microsoft promises to take legal responsibiliy in the case of patent lawsuits resulting from use of their platform
What does that have to do with copyright claims?
Answer: absolutely nothing.
OK, but the Java license specifically demands that Java be implemented in full with no additions or subtractions (I para phrase). What Google have done is to implement subsets of "Java", and call it something else.
Everyone can see that though they do not call their implementation Java, the implementation is indeed Java to a great extent. That is where Google might lose in my opinion.
The headers haven't only be removed - which is a GPL violation by itself - there's a *new* header:
This is a blatant copyright violation, because you can't re-licence GPL code as Apache.
Dilbert RSS feed
What does the Dalvik name has anything to do with this? If you're talking about calling itself Java, this is an unrelated _patents_ matter, while the customers under a bus thing is about copyright - more specifically, the code Oracle says was copied from the OpenJDK and relicensed as Apache, which is a copyright violation.
"Third parties, not Google, would be liable for any Java copyright violations in the Android mobile OS"
Dilbert RSS feed
I read and re-read tfa and just coudn't see Sun *ever* doing this level of crap. Larry may be a "great capitalist" but he's a failure as a human being. Bad Larry! Baaad! (smacks Larry on head with rolled up newspaper)
Organization? You must be joking..
the arrogant @$$hole in the polyester suit.
It's a good thing we have choice in the market.
Prediction: They're going to lose all their open source franchises
as developers and customers walk out.
Where are we going and why are we in a handbasket?
Dalvik is not a java vm, it makes quite different byte code. Google is suggesting the open handset alliance members are responsible not their customers.
Yeah, yeah, but it is *Java* in the sense that a jury will understand it. That's all that really matters here.
That conclusion is really unfounded. You would have to know and understand Android and its developmental origins to understand what is going on here. Google participated in the development of Android and went to great extremes to protect itself from precisely what is going on here now. But as we all know, you don't have to do anything wrong to be sued. In this case, if you read the claims against Google and Google's responses, you would begin that Oracle's claims are rather similar in nature to SCO's against IBM. Many of the claims are completely non-specific while others are simply wrong.
But with all that said, Google made every attempt to protect itself from things like this but is helpless to prevent other parties from making themselves vulnerable to successful legal attacks. What does that mean? Let's put it this way:
Let's say you develop something using Microsoft technologies. Great. Microsoft will protect you legally if any of Microsoft's technologies violate another party's patents. On the other hand, Microsoft will NOT protect you if you add your own stuff that violates another party's patents or other IP. In this case, Google had a hand in creating the base Android platform. It does NOT violate any of the claims made by Oracle. What makers of Android based products may do to the Android base that does violate Oracle's claims is not the fault or failure of Google. If I make a Redhat variant distro and put commercial software in there that infringes on the copyright of another party, Redhat isn't liable for that either.
This is not Google throwing anyone under the bus. This is Google making clear its limits of liability.
My point was, ripping off Java and bypassing the licensing requirements for a mobile Java might have been a good way to save some cash, but it was slightly shady at best, and is going to kill Android as this struggle battles on.
They *call* it Dalvik and the bytecode differs, but the fact remains that it sucks in Oracle(TM) Java classfiles and spits out a proprietary, incompatible binary format. One of the general complaints in the broader struggle between Oracle and Google is precisely this.
The idea that Google copied some code isn't hard to believe, as the whole of Dalvik is just a ripped off Java, with just enough changes here and there to make it appear that it isn't a ripped off Java.
This should teach everybody a good lesson about trusting Google, and also about coding for proprietary VMs be they from Oracle or Google or Microsoft even.
Then the jury is poorly educated. Google should have from the beginning implemented several language to Dalvik VM compilers. I know that would have cost more, but they could have then removed support for java the minute Oracle bought SUN. We all know this is what Oracle does and would do.
That's low! It should be on you!
Besides, Java is just the input language - change it for Android 4.0 - change the input language to Basic, make a new compiler and change the virtual machine accordingly. Problem solved.
If Google really cared they would fix Android Chrome to reflow text, instead of discriminating
So Google rips off most of Java, makes an incompatible bytecode format, calls it something else, and it's magically *NOT* Java?
Google could have protected itself from this fiasco by starting fresh and not using any of the Sun (Oracle) Java implementation, and calling it something else. What they did instead is take a buttload of Java stuff and call it something else so they didn't have to pay up on the licensing fees.
They were betting on massive amounts of applications being written for Android deviced by Java coders because ***___--- IT'S JAVA ---___*** for all practical intents and purposes!
Evil
Rips off or makes an inter-operable language?
There is a difference.
Class names are not artistic and do not deserve protection.
If this is true then Wine is in the same boat.
They implement a compatible Win32 layer.
Which is it?
How about the functions in the classfiles?
Are you saying Google reimplemented every Java class from scratch, in a clean-room fashion, without having seen it first?
So then you think Wine should be shutdown as well?
They implement Win32 and now Win64.
The real lesson here is Google should have either bought SUN or stayed the hell away from anything Oracle touched. Dalvik is not proprietary, it is Apache licensed. Do you think Apache is proprietary?
That might be, but considering that Google has advertised the language as "the Java language" it's an honest enough mistake.
I am saying seeing GPLed code does not prevent me from writing Apache licensed versions that do the same thing. No where in the GPL does it say that by reading this code your brain is now property of someone else.
I am not saying it is impossible that they clean roomed it, it is in fact quite possible, since docs could be easily written by one team from that code to be given to another team. Again the GPL never says thou shalt not write docs from this code.
Since when is a VM licensed under the GPL (OpenJDK) or one licensed under Apache 2,0 (Dalvik) "proprietary"?
They _are_ free. The problem is patents, which apply to any code, free or not, and the copyright violations, which are simply a result of incompatible licenses.
Dilbert RSS feed
Actually, if you simply copy the GPL'd code and relicense it as you wish, you are committing a violation of the GPL at the very least.
If you clean-roomed the code, you may not have a problem. However, if tiny details in your Apache code are identical with the GPL'd code, you are clearly messing around with the GPL.
This all ignores the fact that Dalvik was created as a way to bypass Sun's mobile Java licensing, without actually giving up compatibility with Sun's language files, passing their compliance tests, etc, and all the while fragmenting the Java platform.
Oracle is not pursuing Trademark violations, which is all that sounds like.
If this stuff holds up how long will it be before MS kills Wine?
Is Oracle officially the new SCO?
If you don't want to do the research to verify that I am correct, at least read this part from Groklaw:
I must have pointed out about 20 times here on Slashdot that Android is not Google's OS any more than Linux is a Red Hat OS. It is an OS produced by 78 different companies who are members of the Open Handset Alliance and also has numerous unaffiliated contributors.
Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
I never said I could simply copy it. Tiny details of creative work seem not to match. Else Oracle would be not pulling an SCO. If tiny details really did match we would see reams of matching code.
Dalvik was created to be an open VM that can have java type language compiled for it.
This nonsense that you must pay rent to read a file needs to end. I should not pay to decode h264, nor should google pay to translate some java. If the originator of some filetype thinks it is so special he should not show it to anyone.
So they're just throwing their partners under the bus? I bet that is going to make everyone want to build more Android phones.
or else!
sun / oracle fragmented the "java platform" by trying to keep desktop and mobile java isolated from each other
Snowden and Manning are heroes.
If Oracle sued Red Hat claiming Linux violates their patents, would you say that Red Hat was acting in a "low" fashion for pointing out the fact that Linux is not Red Hat's OS? Of course not. So why would you expect Google to take the hit when Android is not their OS? (Hint: It is FOSS, and is an Open Handset Alliance OS)
Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
Well, I don't know the entire complaint. But I did check on the PolicyNodeImpl.java file. That code was released by Sun under the GPL with the classpath exception; it isn't proprietary source code. At worst, Google would be guilty of a GPL violation by changing the license. But that's not even what's going on. Google isn't shipping the code as part of Android, the shipped it as part of a test package. Furthermore, that code appears to have to reuse even private variable names because those appear to be accessed via reflection by something.
In general, APIs shouldn't be, and aren't, copyrightable. And since in Java, even private members form part of the API (due to reflection), Oracle really has no leg to stand on in claiming copyright over their choice of private variable names.
I agree that google probably should have bought sun. Perhaps though, settling may cost less than an acquisition might have.
Throw the makers to the lions.
Got Code?
"This is a blatant copyright violation, because you can't re-licence GPL code as Apache."
You're assuming this has been copied.
This is pretty much the same as saying someone copied a .h file because they are remarkably similar, sometimes there is only one correct or one common way of writing code.
For this reason, not all code is copyrightable, there are several tests to determine this. I'll be hugely surprised if you could persuade a judge that the similar code in these files is anything other then two implementations of the same standard using common programming practice.
Non-programmers can see that though they do not call their implementation Java, the implementation is indeed Java to a great extent. That is where Google might lose in my opinion.
ftfy. Don't bother telling us "what you think is obviously a Java-VM implementation" until you have (1) read the Dalvik source code, and (2) written a simple compiler/vm.
"...All there is is bullshit, pardon my vulgarity here. Layers of it. One layer of bullshit on top of another. And what you do in life like when you get older is, you pick the layer of bullshit that you prefer and that's your bullshit, so to speak. "
- Bernie LaPlante (Hero)
So whose lawyer has the more preferred layer?
Microsoft could indeed (re)sell you a copy of OSX if they possessed one...
It's like Microsoft demanding you pay Microsoft a license fee for OSX in order for you to buy Windows, but in return for that license fee for OSX you get... nothing... Well you get a paper from Microsoft that says that as far as they are concerned it's okay if you get OSX by some other means and then use it. But since Microsoft doesn't have any stake in OSX that paper doesn't actually mean anything.
Now if you replace the copyright/patent (improperly "intellectual property") bits of the claim with the word/idea "Safety" the nature of this as a "protection racket" becomes clear.
You pay Oracle some money to protect you from any claims Oracle might make against you for things Oracle doesn't own...
You pay (entity) some money for protect you from any (action) (entity) might take against you for things (entity) doesn't otherwise have any right to influence (like your kneecaps or your store-front downtown).
[See "The SCO Group".]
Innocent people shouldn't be forced to pay for inferior software development.
--"Code Complete" Microsoft Press
The objection is, to my understanding, that "Exhibit 'J'" doesn't consist of any of the actual source code at all. It's got nothing to do with the headers or locations. Exhibit "J" is a decompile. To that end it has all the expressive part stripped. It may be that both halves of the comparison are decompiled from their respective objects using the same tool.
This would naturally strip the result of any indicators as to whether any code was copied because the _tool_ would pick the variable names, and the indenting style and so on.
So consider two implementations of some function "int add_two_integers(int, int)".
One guy goes in for the one-liner: "int add_two_integers(int l, int r) { return l+r; }"
Another guy does the long haul: /* giant copyright notice */ /* motivation for writing code */ /*specific integer sizes selected to constrain results within 32 bits for sure */ /*with comments and everything */
int32
add_two_integers(int32 Left, int32 Right)
{
return (Left + Right);
}
In no way did either party "copy" the other.
But you compile both on a 32-bit platform, then decompile them both, and then say look, that second guy just stole the first guy's code.
The outcome of the above would be different code were both compiled for, say, an amd64. The intent is clearly different. The amount of effort clearly different again. The actual act of copying isn't even in question when the source is examined.
But cook it right and use decompiles and whatnot and you produce a misleading sense of similarity.
So don't go looking at the files from the two distributions and how similar or different they are. The objection is to the particular details of an exhibit we don't have that has, according to Google, been produced or redacted or just plain old manipulated to remove the obvious dissimilarities in a way that Google thinks the court should see as dishonest or biased against a correct finding of fact.
Innocent people shouldn't be forced to pay for inferior software development.
--"Code Complete" Microsoft Press
can Oracle now win this lawsuit without simultaneously destroying Java as a product that any developer may choose to use, since it creates the possibility for Oracle to extend the scope of their patents just by the application of the language itself on the part of the developer?
Larry must be fuming at the lawyers who convinced him this was a great strategy to take. Now he may have put his entire company in jeopardy by proceeding with this lawsuit.
but who will be eager to buy a license to a product that Oracle can sue you if you use it?
If this policy begins to bleed into its Java - database offerings, it could sink Oracle as a company within a few years time.
Larry must be freaking out over the lousy legal advise given to him by a few lawyers eager to find a new gravy train.
Yes this has been conventional wisdom. However, it looks as if Google may well use it to beat Oracle over the head with it, since it forces anyone using "open source" Java to pay Oracle, when they don't actually own the right to force developers to pay them for what is essentially open source software. That is an illegal extension of the technology that the patents cover, which only strictly pertain to just how similar to the JVM any VM has to be and be permitted to call itself Java. Google makes no pretense that its VM is a JVM, only that it can use the open source Java language, among others, to produce VM Dalvik bytecode. It seems that Google lawyers had done their homework well before Oracle bought Sun and that Larry got some bad legal advice as to the strength and scope of Sun's patents.
What makes you think there will be a jury rather than just a judge?
Everyone can see that though they do not call their implementation Java, the implementation is indeed Java to a great extent. That is where Google might lose in my opinion.
Uhhh, no. The Dalvik VM is a completely different beast; comparing it to Java is like comparing a sterling engine to an electric engine. Yes, they are both engines. Yes, they might both run on clean energy, and even get you from point to another. However, the damn things don't even have the same guts.
everyone seems eager to smear themselves in s__t as under all that BS is a pot of gold. Fro that either Larry will be glad to paint themselves brown.
The file is not "duplicated". Somebody used a decompiler (probably "jad") to decompile the corresponding Java 5 class. Java 5 here, not OpenJDK.
That somebody then tacked the Apache license on top of the decompiled code.
Pretty obvious if you compare the file to PolicyNodeImpl.java from the publicly accessible Java 5 sources. Again, Java 5, not OpenJDK.
Sun MADE the Java platform, and it seems that Oracle as their successor can do pretty much what it likes with Java.
Again, this should be taken as a good, hard lesson - don't bet the farm on a proprietary language that is encumbered with patents and copyrights and etc.
Yeah, I know that Java is GPLd now, but in the strictest sense it still isn't Free as there are other gotchas and pitfalls.
Indeed, if you look at the two files in question, there are many differences that aren't obvious at first glance. I don't know if one of them changed over time, or if one was influenced by the other, or if class definitions were copied (but not the impl), or what... but it should be pointed out they aren't exact duplicates.
Well, yes, for the sake of argument. I don't know if it has or not, that's for the court to decide. But if it has, it's a copyright violation.
Dilbert RSS feed
>>> But you compile both on a 32-bit platform, then decompile them both, and then say look, that second guy just stole the first guy's code. >>>
Copyrights protect expressions. You have two functions clearly different as they are expressed but after compilation they look the same. So what?
It is quite weird to see the Slashdot community support Google when they are trying to do exactly what Microsoft was trying to do when they aimed to destroy the Java platform through Embrace, Extend, Extinguish. Puzzling. Is there something about Google that makes them more "holy" than Microsoft?
Java is three different and increasingly separate things:
* The Java programming language
* The JVM
* The Java libraries
The JVM can run many languages that are not Java. Many of those languages can use Java libraries. There can be compilers that compile code in the Java language to something other than JVM-bytecode. Dalvik is a VM, but not a JVM. You can compile Java, but also other languages to Dalvik bytecode. Harmony is (as far as I understand) a re-implementation of Java libraries, and they're available to any language running on Dalvik.
As far as I understand, compiling Java code to Dalvik bytecode is not the issue. What is an issue is whether Dalvik violates patents and/or requires a JVM license. Also an issue might be whether Harmony violated copyright by copying code from Java libraries.
Note: I'm no legal expert, nor do I know all that much about this case, but this is what it seems to be about to me.
So Google rips off most of Java, makes an incompatible bytecode format, calls it something else, and it's magically *NOT* Java?
Please. Define Java. It may (at the very least) refer to any of the following:
1 - the Java language specification
2 - the Java Foundation Classes (standard library)
3 - the Java Runtime Environment
4 - the Java Development Kit + Certification Suite
5 - the Java trademark owned by Sun/Oracle
Of these, 1) and 5) are not relevant to the discussion: the language specification is completely open, non-copyrighted and unencumbered by patents (I might hope), and Google is not violating the trademark, because they're calling their implementation Dalvik. It would be equally hard to argue that Google is using the JRE, because they're using an incompatible bytecode format, right?
So that leaves 2) and 4). The certification suite is irrelevant, Google is not trying to certify its VM and DK. As for the JFC, Google is using Apache Harmony with certain parts (introspection) removed. As for the JDK, Dalvik requires its own compiler.
So yes, if you narrowly define Java as the Language Specification only, or even if you include a subset of the java.* Foundation Classes, then Google is using Java. But regardless of what you're claiming: Java is a trademark owned by Oracle. That means that you do not get to decide what is Java and what is not. Only Oracle gets to decide that.
you pull a new defense out of the hat near the end of the trial.
http://www.youtube.com/watch?v=Pr3sBks5o_8. 'Nuff said.
The headers haven't been "removed" since the file is clearly not a copy. The file does appear to have been derived from Sun's source file, but not necessarily in a way that violates copyright.
Yeah, I know that Java is GPLd now, but in the strictest sense it still isn't Free as there are other gotchas and pitfalls.
While software patents exist, there is nothing that will not have 'gotchas and pitfalls'
The intent is what makes it completely different.
Who cares if there's smart guys on the jury to begin with anyway?
Lawyers almost invariably challenge and remove them. They *like* dumb juries.
Not really. The competing Java platform in this case is JavaME, which had essentially destroyed itself (with more than a little help from Sun) by being a hugely fragmented, expensive, closed-source mess that bore little resemblence to desktop Java. Then along came Google with something that desktop Java developers could actually develop code for and expect it to work consistently.
I know JavaME was seriously broken, the problem I have with what Google did was that fixing JavaME would have been a good idea, forking was a bad one.
Really? So the intent makes the result OK? If I intend to have fun in my car and drive around drunk, because it is fun, then it is OK if I kill you since my intent was not to kill you?
Intent is irrelevant, result is relevant. Forking means bad things for the Java Platform, which was badly broken for mobile, we all know, but still. Google forked, which kills the platform, irrespective of intent. Oracle can not, and should not, sit by and watch when another company destroys their platform. Sun didn't when Microsoft tried, no matter what intent Microsoft had.
Oh, and please get me right here. I love Dalvik as a platform. I think what Google has done, technically, is excellent. Finally we can actually develop in Java on the mobile platform. The fact that I like it and that you like it should not blind us to the fact that Oracle should not accept it. There is a huge difference between what might be good for you and me and what is good for Java as a platform and the owner of the Java platform.
IANAL, but ISTYOMLOS** so I have picked up a few things.
Ellison will have his legal team print all the JVM source code and ship it to every Android integrator by Sikorsky helicopter.
On top of each pile, there will be a note: "Show us your violations."
**I Squandered Ten Years Of My Life On Slashdot
Rich And Stupid is not so bad as Working For Rich And Stupid.
You probably should have read the first few paragraphs:
... and ...
The entire document relates to and describes terms of service for the website, not Android. If you want to know about the licensing terms for Android you need to look at the license(s) in the source.
Not just for the website, but for (and I quote the lines you just quoted, emphasis mine):
1.2 Your use of products, software, services and websites in connection with the Open Handset Alliance website
So, it is not just related to (and describing terms of) service for the website but:
related to the OHA website. Somehow I think Android is a) a product and/or b) a software related to c) the OHA website (and/or the OHA itself which is also a Google trademarked name.)
I never knew that Android was a fork of JavaME. Thanks for educating me.
All this talk about headers .... when the replacement of the header was a violation in itself.
No, the code is different. They obviously had different authors implementing the same specification. One author prefers while loops, the other for loops. Variables and members are named differently. See these examples:
Java:
private void getPolicyNodes(int depth, Set set) {
if (mDepth == depth) {
set.add(this);
} else {
Iterator it = mChildren.iterator();
while (it.hasNext()) {
PolicyNodeImpl node = (PolicyNodeImpl) it.next();
node.getPolicyNodes(depth, set);
}
}
}
Android:
private void getPolicyNodes(int i, Set set) {
if(mDepth == i) {
set.add(this);
} else {
PolicyNodeImpl policynodeimpl;
for(Iterator iterator = mChildren.iterator(); iterator.hasNext(); policynodeimpl.getPolicyNodes(i, set))
policynodeimpl = (PolicyNodeImpl)iterator.next();
}
}
-molo
Using your sig line to advertise for friends is lame.
The .dex format is very different from the .class format, and the differences are to optimize for memory and CPU on mobile devices so that you can run one lightweight VM per app. These aren't superficial changes.
The Java license, as I understand it, is requirements on how to implement something called Java. If they're not calling it Java, how is the license applicable?
"Presence of malice makes it wrong" != "Absence of malice makes it right"
... this sounds like a crap move. But you have to consider just why Oracle went after Google. Practically every cellphone vendor has some models with Java support. So just about anyone using Android already has a Java license of some kind. And it's hard to imagine Google didn't make this announcement only after talking to strategic partners in the Android business.
As well, as long as they're not actually making Android devices (Google-branded Nexus devices are made by existing phone makers, not Google), it's hard to imagine just how Oracle manages to really sue Google on this, regardless of the actual truth behind the open source licenses Sun put in place ages ago. Oracle seems hell-bent on making Java irrelevant, rather than making money on it. Google doesn't want to move off Java, but if their Dalvik JIT is violating the original FOSS license (which seems to be Oracle's entre into the lawsuit business here), they could migrate developers to other technologies using the same JIT. Or do a cleanroom Java, for that matter. Only a small hill to climb for a company as powerful as Google.
-Dave Haynie
I know JavaME was seriously broken, the problem I have with what Google did was that fixing JavaME would have been a good idea, forking was a bad one.
Android is - deliberately - not a fork of JavaME, if anything it has more in common with the desktop version. By the time you've fixed all the issues with JavaME that prevent it from being a practical option for smartphones, you may as well start again from scratch. It's not like existing applications would work anyway.
HURRR!
No, I don't think Wine should be shut down. Where the fuck did I say that?
The WINE devs have been careful to emphasize that they are performing a CLEAN ROOM implementation of the Windows API, without looking at the Windows source code. They are not simply copying parts of it that they like, or creating middleware that sucks in raw Windows .DLL files.
If you would wake up and pay attention, Google's difficulty started when they produced a mobile phone Java, while trying to both skirt around their licensing fees and also breaking the patent protection guarantee, which specifies that Java should not be fragmented. They decided to be cheap about their choice of a platform, and liked Java, but didn't want to pay up, so they got some lawyers and finagled up a solution that they believed would shield them from consequences.
Fact: Dalvik is just an incompatible Java under a different name, that uses actual Java files but spits out a proprietary format. You know, this is almost exactly the same sort of shit that MS did, and over which they lost in court.
I like open source software, but I don't think Google should be given a free pass to just do whatever the fuck they want because they are friendly to the cause. They fucked up, clearly, and could easily have created a new programming language or taken any number of roads other than cloning Java and deciding to skimp out on paying Sun (or Oracle) for their trouble.
The Google/Oracle patent litigation is really heating up. If judged in the court of public opinion, it's looking more and more like Oracle would be the loser.