Linked by Thom Holwerda on Tue 5th Oct 2010 21:56 UTC
Legal So, Google has finally officially responded (thanks for hosting, Engadget) to Oracle's patent and copyright infringement lawsuit against the search giant's Android mobile operating system. Apart from boatloads of pages on how Google pretty much denies any and all claims, there's a lot of interesting stuff in there - stuff that doesn't seem to bode well if the courts do decide Google is infringing Oracle's patents. It also makes it crystal clear that anyone who values Free and open source software should avoid Oracle products like the plague.
Order by: Score:
Sorry Google stuffed up.
by oiaohm on Tue 5th Oct 2010 22:21 UTC
oiaohm
Member since:
2009-05-30

Sun was refusing to give the TCK due to the fact apache was wanting to use a more open license than GPL.

http://jcp.org/en/participation/membership 5000 USD would have given google access to what you need to test TCK conformance ie Java Compatibility Test Tools (Java CTT). To individuals cost is nothing.

Even so android implementation would still fail. Reason 1 and its a big one. Out the box android implementation cannot run native Java bytecode. Even if it was just shipped with a on the fly converter to Googles. So the implementation fails.

Remember one of the key objectives of java is 1 binary everywhere. Google android fails this. Anyone show much an android app that is using Google byte code system that will run on windows and linux.

Sorry Oracle is no the only bad guy here. Google is also a bad guy due to the way they have done things.

Reply Score: 4

RE: Sorry Google stuffed up.
by wargum on Tue 5th Oct 2010 22:33 UTC in reply to "Sorry Google stuffed up."
wargum Member since:
2006-12-15

Look, if all you do is take a piece of FOSS and manipulate it in ways the license allowes you to and you can still be sued, something is terribly wrong.

I live in the EU, Germany to be specific. Software patents are illegal over here and an attempt to legalize software patents got rejected by the european parliament. It was the first time ever that the parliement rejected a bill that was proposed by the European Council.

So, at least for now, this is the sane part of the world I live in ;)

Think of the worst thing you can imagine, multiply this by infinity and you'll see just how bad software patents are.

Reply Score: 4

RE[2]: Sorry Google stuffed up.
by gnufreex on Tue 5th Oct 2010 22:51 UTC in reply to "RE: Sorry Google stuffed up."
gnufreex Member since:
2010-05-06

That is the problem, Google didn't manipulate it in ways that license allows. Oracle (or Sun) never released Java under ASL. They released it under GPL. Google decided that they will like to use reverse-engineered and ASL-licensed Harmony as base of Dalvik. GPL patent clause only work to protect GPL'd modification. Dalvik is viewed by Oracle as competitor because anybody can take it and make proprietary fork. IBM would like to do that. So that is probably biggest reason why they are suing.

Google could still rebase on OpenJDK and get out of this mess. OpenJDK class library is superior to that of Harmony.

Reply Score: 5

RE[3]: Sorry Google stuffed up.
by JAlexoid on Thu 7th Oct 2010 13:44 UTC in reply to "RE[2]: Sorry Google stuffed up."
JAlexoid Member since:
2009-05-19

reverse-engineered and ASL-licensed Harmony as base of Dalvik.


What!?!?!? Harmony is not reverse engineered it's written to publicly available specification.

Reply Score: 2

RE[4]: Sorry Google stuffed up.
by gnufreex on Thu 7th Oct 2010 19:53 UTC in reply to "RE[3]: Sorry Google stuffed up."
gnufreex Member since:
2010-05-06

My mistake. It doesn't change anything; I wanted to say it is a clone. Count it as I typo, all else stands same.

Edited 2010-10-07 19:54 UTC

Reply Score: 2

Java :-)
by evert on Tue 5th Oct 2010 22:29 UTC
evert
Member since:
2005-07-06

Actually, I never understood the popularity of Java. C++ with some libraries like Boost and Qt is great, also more open, and delivers good performance. Google should have bought Trolltech before Nokia did.

Reply Score: 4

RE: Java :-)
by Dr.Mabuse on Wed 6th Oct 2010 01:41 UTC in reply to "Java :-)"
Dr.Mabuse Member since:
2009-05-19

Exactly! It's all proving to be far more trouble than it's worth. Now, where's Meego at...

Reply Score: 1

RE: Java :-)
by bitwelder on Wed 6th Oct 2010 06:56 UTC in reply to "Java :-)"
bitwelder Member since:
2010-04-27

I disagree. Google's home field is the web (that's where their ads cash-cow machine works) and Java lives just in that field.
Qt is much more 'general purpose', and I think to develop to its full potential for Google would mean to invest well outside of its core interests.

Reply Score: 4

lemur2
Member since:
2007-02-17

http://www.groklaw.net/article.php?story=20101005114201136

I am presuming that we can shortly expect the same kind of response to Apple and Microsoft.

For example: Microsoft asserts its FAT patents, but the Linux kernel has work-arounds for those, where it does not include the patented functionality:

http://arstechnica.com/open-source/news/2009/07/vfat-linux-patch-co...

I think there will be equally strong responses to Microsoft and Apple in due course.

The only question in my mind is if Google will bother to try to take down Bing. Is it worth it?

Reply Score: 5

molnarcs Member since:
2005-09-10

Great link there to Groklaw. I'm not very keen on reading legalese, but some parts of Google's counterclaims are very clear, to the point, plain English common sense claims. This is my favourite:

"The deficiencies in Oracle's copyright infringement claim are both striking and telling in view of the fact that Android is an open-source platform, and that all of the relevant source code and documentation for Android is currently — and has for some time been — publicly available. Oracle's inadequately-pleaded accusations are striking because Oracle has had complete access to every piece of information that is relevant to any possible assertion of copyright infringement, yet Oracle still failed to meet even the minimum pleading standards. Oracle's inadequately-pleaded accusations are also telling because, notwithstanding Oracle's access to the allegedly infringing work, Oracle has not articulated a legally sufficient claim that puts Google on notice of either the factual basis or the substance of Oracle's copyright claim."

Reply Score: 4

Mouse Trap
by saidge@yahoo.com on Tue 5th Oct 2010 22:37 UTC
saidge@yahoo.com
Member since:
2007-11-06

Correct me if I'm wrong... but it sounds like Sun open sourced a mine field and kept the map under lock and key.

Now Google's stepped all over that mine field and never got the map, and Oracle armed the bombs.

That about right?

And remind me again, what good is open sourcing something without granting the applicable patent licenses?

It's like inviting someone to your house, saying 'mi casa es su casa' and then saying 'But if ye do summin I dun like, I'll shoot ya for tresspassin!'

Reply Score: 4

RE: Mouse Trap
by Kroc on Tue 5th Oct 2010 22:41 UTC in reply to "Mouse Trap"
Kroc Member since:
2005-11-10

“it sounds like Sun open sourced a mine field and kept the map under lock and key.”

Yeah, it’s a shame nobody saw that possibility. Oh wait, loads of us did. You would have to be very dense, or brave, to develop with Mono, likewise.

Reply Score: 4

RE[2]: Mouse Trap
by Thom_Holwerda on Tue 5th Oct 2010 22:46 UTC in reply to "RE: Mouse Trap"
Thom_Holwerda Member since:
2005-06-29

I've always been one of "meh" when it comes to Mono, but over the past few months, I've been growing closer and closer to the those critical of Mono and its Microsoft ties. Let me make it clear, though, that this doesn't mean I support those rabid anti-Mono and anti-De Icaza idiots - I just think that relying on Mono is a shortsighted mistake.

This Java Oracle thing only strengthens that idea.

Reply Score: 9

RE[3]: Mouse Trap
by Kroc on Tue 5th Oct 2010 22:51 UTC in reply to "RE[2]: Mouse Trap"
Kroc Member since:
2005-11-10

Exactly the same boat, I don‘t use the product so can’t scream hate at them--I just don’t care about it--but I do want to raise a hand and say that this is going to end badly and I told you so.

Reply Score: 1

RE[3]: Mouse Trap
by gnufreex on Tue 5th Oct 2010 22:59 UTC in reply to "RE[2]: Mouse Trap"
gnufreex Member since:
2010-05-06

Java is not same as Mono. Microsoft never open sourced .NET under GPL.

Dalvik is same to OpenJDK as Mono is to .NET. A unlicensed incompatible fork.

Except OpenJDK is free and it is stupid to use Dalvik when there is OpenJDK which is completely free and superior in every way.

On the other hand .NET is closed and Mono is a liability since it was not opensourced by patent holder. It is created by 3rd party and Microsoft can sue when they decide.

Reply Score: 8

v RE[4]: Mouse Trap
by graig on Wed 6th Oct 2010 03:42 UTC in reply to "RE[3]: Mouse Trap"
RE[4]: Mouse Trap
by nt_jerkface on Wed 6th Oct 2010 04:47 UTC in reply to "RE[3]: Mouse Trap"
nt_jerkface Member since:
2009-08-26

Mono is a liability because Miguel went past implementing C# and the CLI which are ECMA standards.

The liability risk is really with Winforms. MS didn't even push for Mono to include Winforms, it was Miguel that made the request. Once Novell gets bought out I'm sure there are all kinds of clauses that kick in.

Reply Score: 7

RE[4]: Mouse Trap
by dsmogor on Wed 6th Oct 2010 09:43 UTC in reply to "RE[3]: Mouse Trap"
dsmogor Member since:
2005-09-01

Dalvik is very different from a casual JVM. It just happens to utilize .class as an intermediate object format but that's it. It wouldn't make much use of OpenJVM code, so making it a GPL'd derivate of it would have to be a formal trick.

Reply Score: 2

RE[5]: Mouse Trap
by gnufreex on Wed 6th Oct 2010 23:44 UTC in reply to "RE[4]: Mouse Trap"
gnufreex Member since:
2010-05-06

Later part is not true. OpenJDK have HotSpot JVM (there is no such thing as OpenJVM), which is more geared towards enterprise. It is not great for phones. But, Harmony is not great for phones either. Harmony is reverse engineered copy of OpenJDK. What google took from Harmony is class library, not JVM. They wrote their own JVM.

In a same way they did that, Google could use OpenJDK's class library instead of Harmony's , and then write their own JVM. OpenJDK's class library is real deal, not hacked up clone, so they would even be more compatible with real Java. They could offer Oracle to adopt their stuff as Java ME, because Java ME sucks. Any kind of lawsuit wouldn't be possible because Oracle can't sue against derivative of it's own class library which is GPL. GPL forbids suing your users.

In fact, they could try to sue if they are crazy, but that case would probably already be dismissed by now.

Reply Score: 2

RE: Mouse Trap
by lemur2 on Wed 6th Oct 2010 01:13 UTC in reply to "Mouse Trap"
lemur2 Member since:
2007-02-17

And remind me again, what good is open sourcing something without granting the applicable patent licenses?


Open source java (openjdk) does have an applicable patent license, the GPL grants exactly that.

http://openjdk.java.net/
http://community.java.net/mobileandembedded/

For some reason that I cannot fathom, Google chose not to use openjdk.

Edited 2010-10-06 01:13 UTC

Reply Score: 3

RE: Mouse Trap
by graig on Wed 6th Oct 2010 03:40 UTC in reply to "Mouse Trap"
graig Member since:
2010-09-18

if you open source someone, you should basically have given them the rights to use all patents, as you gave them the right to use the code the patents cover, it's pretty much implied that you gave the right to the patents also.

Reply Score: 0

RE[2]: Mouse Trap
by lemur2 on Wed 6th Oct 2010 05:11 UTC in reply to "RE: Mouse Trap"
lemur2 Member since:
2007-02-17

if you open source someone, you should basically have given them the rights to use all patents, as you gave them the right to use the code the patents cover, it's pretty much implied that you gave the right to the patents also.


The GPL includes a patent grant that does exactly that.

The problem here is that Dalvik is kind-of-like a Java, but it isn't java. It isn't compatible with java, it is not the same thing functionally, the bytecode binaries are not inter-operable.

I'm pretty sure that Sun's patent grant for java required that one's implementation of java was bytecode compatible with Sun's java. If it wasn't, you didn't get the patent grant.

However, Oracle's main problem is that in the past they have argued directly against this provision, and they are on record as saying so. Google has a right to have taken them at their word.

Reply Score: 2

It is google's fault.
by gnufreex on Tue 5th Oct 2010 22:50 UTC
gnufreex
Member since:
2010-05-06

They only need to put some of OpenJDK code in DalvikVM and move it to GPL. That would give them patent grant from Oracle that comes with GPL. Apache Foundation complaints are in fact IBM complaints. IBM wants to be able to make proprietary fork of Java, so they advocate changing Java license to Apache license v2. Note that Red Hat don't have any problems because thir IcedTea is based on OpenJDK and licensed under GPL. Oracle would like to sue Red Hat (JBoss is killing WebLogic in the market). But they can't because they already given out the patent license when they distributed the thing under GPL.

As for claim that Schwartz convinced Oracle to sue: simple, De Icaza is lying. He is only source for that, he made that up. You can see from James Gosling's comments that Sun had plans to sue anyways. Java is free as always, but it is not under giveaway license. It is under GPL. Oracle is probably anoyeed that Apache-licensed remiplementation of Java would let IBM embrace and extend Java, cutting Oracle out of the picture (and that copy would be proprietary).

I don't like Oracle and I condemn any patent litigation, but fact is that Google chose not to use only thing that protected them from patents - OpenJDK. Even FSF pointed that in their response:

http://www.fsf.org/news/oracle-v-google

Edited 2010-10-05 23:05 UTC

Reply Score: 8

RE: It is google's fault.
by TemporalBeing on Tue 5th Oct 2010 23:26 UTC in reply to "It is google's fault. "
TemporalBeing Member since:
2007-08-22

IBM already had their own clean-room JVM (http://en.wikipedia.org/wiki/IBM_J9)which was proprietary. They discarded it a few years ago. So skip the IBM bs.

Also, Gosling stated that Sun never had the heart to sue over patent infringement, so Sun wouldn't likely have done so. Nothing he said speaks to Schwartz's role. That said, from the little evidence present you are still correct that it seems like De Icaza is the only source.

Per FSF, they are pushing the GPL. It wouldn't necessarily resolve the problem for Google to have used IcedTea (per FSF) or OpenJDK as Google's environment would still have had to pass the TCK to gain license, especially if they only used a subset of the OpenJDK or IcedTea implementations (as they did with Apache Harmony) if I understand everything correctly.

IcedTea could be in just as much legal trouble despite the GPL. Remember, Sun used GPLv2 which is more ambiguous about patents than GPLv3 - 99% of the reason for GPLv3 was to clarify the whole patents thing.

Don't get me wrong, I'm all for the GPL. Just pointing out there is more to it than simply using OpenJDK or IcedTea.

Reply Score: 3

RE[2]: It is google's fault.
by gnufreex on Thu 7th Oct 2010 00:46 UTC in reply to "RE: It is google's fault. "
gnufreex Member since:
2010-05-06

I know about IBM j9, it is consisted from code licensed from Sun (that is also in OpenJDK now, just under GPL) and Harmony code. IBM is trying to remove all code they got from Sun/Oracle (because they are paying some royalty for that, and they are contracted to stay compatible as long they are using that code), and replace it with Harmony code or their own propriatary code.

http://news.slashdot.org/comments.pl?sid=1752790&cid=33249630

I heard that story from few other IBM employees, so there is truth to that. But that is not all I heard. I also heard they plan to port some IBM proprietary languages to their proprietary JVM. So they are creating proprietary superset of JVM which they, wont share with Oracle or anybody else. It is important to note that Apache Harmony was started by IBM in 2005, before Java was officially open-sourced, but *after* GNU Classpath have become working 3rd-party implementation of Java. That is to say, Harmony is created as competitor to GNU Classpath. If IBM didn't plan to go proprietary, they would not start Harmony. Instead, they would help with GNU classpath which was already at advance stage at the time, and it was putting pressure on Sun JDK.

About FSF and GPL, I don't think they would lie to push GPL. I think they really beleive GPL would solve the issue, and I agree. In regard to java, GPLv2' implicit patent grant works same as v3's explicit would. That is because bulk of the code is contributed by patent holder, and some other reasons that would take much typing to explain here. So version doesn't make real difference. What GPLv3 help is to stop treacherous deals lie MS-Novell, and to make patent grant more explicit which helps when there are lots of contributors (like Linux kernel for example).

Again, for TCK, I think you got it wrong. You get patent license if:

a) you make implementation from scratch, without any OpenJDK code and go trough TCK to certify it (this doesn't apply for mobile, they excluded that; also, thre is a fee for using TCK, but that is not royalty)

b) You fork OpenJDK. No need for TCK or anything else, GPL gives you patent grant that apply to mobile, embeded, or even space ships or sharks with laser beams attached to their heads. You could additionally go with TCK, if you want to be compatible*.

c) You do same as a) but if you would like mobile license too, you need to pay Oracle additionally. First two are obviously royalty free.


*Don't be confused that Red Hat did b) with IcedTea, and then put it through TCK anyways. That is perfectly fine, that just mean that Red Hat got two licenses at same time. Also, I think Red Hat paid something to Sun so they could use OpenJDK name which is trademarked, and only available for use by OpenJDK project if not licensed separately. That has nothing to do with patents. Simple nymshifting solves that problem.

Reply Score: 2

RE: It is google's fault.
by igf1 on Sat 9th Oct 2010 05:09 UTC in reply to "It is google's fault. "
igf1 Member since:
2008-11-17

man, I gotta say it, I find the FSF's position on this matter, terminally disappointing. I mean, coming from a group that literally did exactly the same thing (cloned) to C and C++ and Python and Pascal and other languages. Google did, what everyone since the dawn of civilization has done. They saw something they liked and forked it. Hell, Java (my #1 lang BTW) is a derivative as is everything else. Google, played fair here, they re-designed it from the ground up, to suite their needs and they should have EVERY right to do so. They don't have to be bound by a licence that imposes one groups version of "freedom" on others. I personally think freedom has to fight for freedom regardless of (gulp) "alliance".

Reply Score: 1

RE[2]: It is google's fault.
by gnufreex on Sat 9th Oct 2010 21:03 UTC in reply to "RE: It is google's fault. "
gnufreex Member since:
2010-05-06

I think you misunderstood what they said.

FSF didn't say Google shouldn't make Dalvik. They said Google could defend better from a lawsuit if they used some of GPL'd code from OpenJDK (which Oracle distributes, and gives patent grant), and if they released Dalvik under GPL. And that is factual statement.

Then they criticized Oracle because any offensive use of patents against free sofware is inexcusable. They practically sided with Google, just mentioned that Google could avoid this.

Reply Score: 2

RE[3]: It is google's fault.
by igf1 on Sun 10th Oct 2010 20:47 UTC in reply to "RE[2]: It is google's fault. "
igf1 Member since:
2008-11-17

I fully understand the comments made by the FSF. I'm saying that they used this law suit as a promotional opportunity. That's not in the spirit of freedom. One of the more questionable assertions is that the GPL2 is a "strong defense" against patent attacks. It's not, GPLv2, has an implied (open to interpretation) patent license. It doesn't contain explicit licensing as the Apache 2 and GPL3 licenses do. This move was quite intentional by Sun, as it feared forking more than anything, and would have had little control over java's development had they gone with v3. Don't get me wrong, Google's no saint here, and they certainly tend to bow out when it comes time to take a stand against the dark side of big business. Oddly enough, process patents are one such example. I find their presumed neutrality on the issue very strange indeed, as they have no patent extortion side business like Oracle, IBM, MS etc., are constantly extorted by others holding imagination monopolies and have a relatively anemic portfolio.

Just so were on the same page though, please remember that the FSF is a corporation, as such it panders and promotes. I just hate the tacky intellectually dishonest undertone of this statement... but that's just me ;)

Reply Score: 1

...
by Hiev on Tue 5th Oct 2010 23:04 UTC
Hiev
Member since:
2005-09-27

Avoid MySql? I don't think so, you guys including OSNews don't have the balls to drop it.

Reply Score: 5

RE: ...
by Lennie on Tue 5th Oct 2010 23:29 UTC in reply to "..."
Lennie Member since:
2007-09-22

PostgreSQL 9 is really nice. :-)

Reply Score: 5

RE[2]: ...
by Hiev on Tue 5th Oct 2010 23:29 UTC in reply to "RE: ..."
Hiev Member since:
2005-09-27

So is Firebird, I use it.

Reply Score: 2

RE[2]: ...
by Magma on Wed 6th Oct 2010 14:20 UTC in reply to "RE: ..."
Magma Member since:
2006-03-07

Postgres violates my bit-bucket view of databases. When you become enamored it's hard to switch.

Reply Score: 1

RE[3]: ...
by Lennie on Wed 6th Oct 2010 20:18 UTC in reply to "RE[2]: ..."
Lennie Member since:
2007-09-22

PostgreSQL seems to adhere to more standards than any other database system I know. Well, OK, maybe SQLite has a bigger percentage of standards-compliance, but it has a much smaller commandset.

Reply Score: 2

RE: ...
by bryanv on Tue 5th Oct 2010 23:33 UTC in reply to "..."
bryanv Member since:
2005-08-26

I'm currently working with Postgresql 9.0 at work, on a rather large system and data set. I couldn't be happier. I have found the community support, toolset, and actual software to be far, far better than anything mySQL ever offered.

The last I'd worked with postgres, it was version 7.1 back in 2001. Has it ever come a long way since then!

Reply Score: 3

...
by Hiev on Tue 5th Oct 2010 23:08 UTC
Hiev
Member since:
2005-09-27

Oh, and add VirtualBox to too.

Reply Score: 3

RE: ...
by Lennie on Tue 5th Oct 2010 23:25 UTC in reply to "..."
Lennie Member since:
2007-09-22

I'm actually looking forward to using btrfs. It's pretty much the only think I 'need' from Oracle.

Reply Score: 5

RE: ...
by 0brad0 on Wed 6th Oct 2010 06:25 UTC in reply to "..."
0brad0 Member since:
2007-05-05

Oh, and add VirtualBox to too.


I have no problem not using something that hasn't worked properly with the OS's I want to run and is buggy.

Reply Score: 1

RE[2]: ... - Which OS?
by jabbotts on Wed 6th Oct 2010 14:02 UTC in reply to "RE: ..."
jabbotts Member since:
2007-09-06

I had the oposite problem, after years with VMware Server I had to abandon my VMs because VMware couldn't provide a functional install for Debian 64bit (wtf.. Debian isn't obscure enough to justify grief). By contrast, Virtualbox has worked like a dream thanks to a recent version included in Debian 6.

(not that I'd use either for production servers; ESX or an otherwise dedicated host if it was production use in the racks)

Reply Score: 2

To clairity things...
by JPowers27 on Tue 5th Oct 2010 23:28 UTC
JPowers27
Member since:
2008-07-30

OpenJDK is a procedure/function library used by programs running on the VM.

Dalvik is a VM.

Any patents for Java VM are unlikely to affect Dalvik since they are totally different VMs. Dalvik is closer to the old PCode VM then the Java VM.

The only patents Google has to worry about are the ones covering the Java Standard Libraries (SDK). I'm not sure how you would patent a set of base functions.

Any patents in the Java compiler are also moot. Google doesn't include the compiler. Google has a java-byte-code to dalvik-byte-code compiler. They also have an assembler that generates dalvik-byte-code directly. Also, there are multiple java-byte-code compilers and Android's converter will be happy with the output of any of them.

Harmony doesn't include any Sun source code in the library (it's a clean room version). Google removed the user interface sections of Harmony and replaced them with an Android GUI. Thus they violate no copyrights.

Any patents in Java should be thrown out. You can't patent an abstract idea. Thus, how can you patent an abstract idea running on an abstract machine.

Reply Score: 2

RE: To clairity things...
by nt_jerkface on Wed 6th Oct 2010 05:23 UTC in reply to "To clairity things..."
nt_jerkface Member since:
2009-08-26



Any patents for Java VM are unlikely to affect Dalvik since they are totally different VMs. Dalvik is closer to the old PCode VM then the Java VM.


Oracle contends that Davlik is not a clean room implementation.

There was a blogger a while back who not only predicted this would happen but also correctly identified the part of Davlik that Oracle would target.

Google's defense questions the motivations of Oracle and their position on open source which tells me that they are afraid of simply refuting the charges.

Reply Score: 2

RE[2]: To clairity things...
by wirespot on Wed 6th Oct 2010 07:45 UTC in reply to "RE: To clairity things..."
wirespot Member since:
2006-06-21

Oracle contends that Davlik is not a clean room implementation.


And that is the gist of the matter, the only solid thing in the Oracle claims. But it has to do more with copyright than with patents. Oracle will have to show what pieces of code Google copied verbatim from the official Sun JDK.

I personally find it hard to believe there are any. If Google says it did a clean room implementation, it must have. The only claim Oracle could have is with the fact Google lured developers from Sun in order to do so. But that falls in the realm of trade secrets, NDAs, unfair competition and so on, not copyright. Not unless those developers managed to replicate entirely from memory verbatim copies of large pieces of code they previously wrote while at Sun.

Google's defense questions the motivations of Oracle and their position on open source which tells me that they are afraid of simply refuting the charges.


You must have read the summaries wrong (or had them explained wrong to you). Google is refuting the claims (not "charges"). Not only this, Google is going for a summary dismissal of the case altogether and asking for Oracle to pay their expenses. That's a very aggressive stance for such an early point in such a case.

Why would that work? By piecing together a couple of issues, some of which you touched on above:

1) The Dalvik code is completely open, so Oracle could have very easily pointed out exactly what pieces infringe their copyrights. They didn't, why? The law in the US says the accuser has to show all the facts so the defendant can defend themselves, not keep them guessing. PJ at Groklaw gave a nice analogy, which was actually used by a magistrate in the SCO trials (SCO tried this with IBM): it's like a store detective grabbing a person, claiming they stole a product, but instead of saying what product and that they saw them take it, they give them a catalog and tell them "find the product in here, you know what you took!" You can't do that, yet that's basically what Oracle's doing. (And SCO was punished later in the trials for doing this, may I add.)

2) Oracle used to be on the same side as Google, criticizing Sun for the terms they chose to make JDK open. And now, after taking over, they do an 180 turn. Don't thing that such ill-intent and hypocrisy isn't relevant in a court of law. The judges are not blind and not so easily swayed by technicalities. That Oracle did this is saying something, both about their copyright and their patent claims; see estoppel for instance.

It is perfectly feasible for Google to succeed in dismissing the case for the above reasons. That means Oracle's claims would go *poof* instantly, no litigation necessary. Not to mention paying Google's expenses in all this.

Reply Score: 4

Wirespot you have it badly wrong.
by oiaohm on Wed 6th Oct 2010 07:59 UTC in reply to "RE[2]: To clairity things..."
oiaohm Member since:
2009-05-30

Java patent grant is dependent on either being tck or being . Dalvik is neither.

Its not copyright infrignement it is patent infringement. Google simple has no patent grant.

Google and Oracle criticizing Sun on not being open enough true. But that still does not give Google the right to disobey the rules everyone has had to.

http://en.wikipedia.org/wiki/OpenJDK remember Redhat pulled off passing the TCK and so legally having a patent grant using non Oracle stuff.

It would have been simple for redhat as well to say stuff the rules.

Google aggressive stance is that Oracle is not playing light either. Remember Oracles words not payment but complete destruction of infringing product. That would be Dalvik and androids java design infringing on Oracles patents that no longer produced. Very big and costly to Google.

Both are out for pure Blood.

Reply Score: 2

lemur2 Member since:
2007-02-17

Java patent grant is dependent on either being tck or being . Dalvik is neither.


Or being what?

Whatever ... as you say Dalvik is not java, and Dalvik is not TCK.

Since Dalvik makes no claims to be either, why exactly does Dalvik need a "Java patent grant"?

Dalvik is a non-java VM. So is .NET, it too is a non-java VM.

Why then didn't Oracle sue Microsoft rather than Google?

That would be Dalvik and androids java design infringing on Oracles patents


Dalvik is not java, and makes no claims to be.

http://en.wikipedia.org/wiki/Dalvik_%28software%29
Dalvik is the name of the virtual machine in Google's Android operating system. Dalvik is thus an integral part of Android, which is typically used on mobile devices such as mobile phones, tablet computers and netbooks. Before execution, Android applications are converted into the compact Dalvik Executable (.dex) format, which is designed to be suitable for systems that are constrained in terms of memory and processor speed.

Dalvik, like the rest of Android, is open-source software. It was originally written by Dan Bornstein, who named it after the fishing village of Dalvík in Eyjafjörður, Iceland, where some of his ancestors lived.

...

Unlike most virtual machines and true Java VMs which are stack machines, the Dalvik VM is a register-based architecture.


The Dalvik VM doesn't even have the same fundamental architecture as Java VM. It is fundamentally different.

Edited 2010-10-06 08:34 UTC

Reply Score: 3

RE: To clairity things... - abstract idea
by jabbotts on Wed 6th Oct 2010 14:05 UTC in reply to "To clairity things..."
jabbotts Member since:
2007-09-06

"You can't patent an abstract idea."

Rational Thought was patented early on. The USPTO and similar do not have rights to make use of that patented process.

Reply Score: 5

lemur2 Member since:
2007-02-17

"You can't patent an abstract idea." Rational Thought was patented early on. The USPTO and similar do not have rights to make use of that patented process.


If the USPTO genuinely granted a patent on an abstract idea, rather than on an implementation of an idea, then that is merely an error on the part of the USPTO.

The law is clear enough ... one can apply for patents on new methods of doing things, but not on the abstract idea of doing that thing itself.

Reply Score: 2

jabbotts Member since:
2007-09-06

You did catch the joke in that comment right?

Reply Score: 2

Weak
by joshv on Wed 6th Oct 2010 01:41 UTC
joshv
Member since:
2006-03-18

"As Google gleefully points out several times, Oracle used to be squarely on Apache's side in this debate, urging Sun quite strongly on several occasions to provide said license to Apache. For instance, Oracle supported a motion that "TCK licenses must not be used to discriminate against or restrict compatible implementations of Java specifications by including field of use restrictions on the tested implementations or otherwise.""

This line of argument seems particularly weak. "See, see, Oracle used to be on our side!!!"

Yeah, and then they got what they wanted by purchasing it, instead of convincing Sun to give it away. Now they don't want anyone else to have it. So freakin' what? Unfair? Why didn't you buy Sun then Google?

Reply Score: 0

RE: Weak - not at all
by lemur2 on Wed 6th Oct 2010 02:01 UTC in reply to "Weak"
lemur2 Member since:
2007-02-17

"As Google gleefully points out several times, Oracle used to be squarely on Apache's side in this debate, urging Sun quite strongly on several occasions to provide said license to Apache. For instance, Oracle supported a motion that "TCK licenses must not be used to discriminate against or restrict compatible implementations of Java specifications by including field of use restrictions on the tested implementations or otherwise."" This line of argument seems particularly weak. "See, see, Oracle used to be on our side!!!"


Au contraire, estoppel is a perfectly valid legal defense.

http://en.wikipedia.org/wiki/Estoppel

Estoppel in its broadest sense is a legal term referring to a series of legal and equitable doctrines that preclude "a person from denying or asserting anything to the contrary of that which has, in contemplation of law, been established as the truth, either by the acts of judicial or legislative officers, or by his own deed, acts, or representations, either express or implied.


The argument is that because Oracle are on record as saying that "TCK licenses must not be used to discriminate against or restrict compatible implementations of Java specifications by including field of use restrictions on the tested implementations or otherwise" then Google have a perfect right to take Oracle at its word.

Edited 2010-10-06 02:06 UTC

Reply Score: 6

RE[2]: Weak - not at all
by saynte on Wed 6th Oct 2010 06:08 UTC in reply to "RE: Weak - not at all"
saynte Member since:
2007-12-10

I don't think estoppel can be used against Oracle here. Oracle said that Sun shouldn't use the TCK license restrictions in a certain way. This is an opinion about Sun's behaviour, not a promise by Oracle. Even then, Oracle's statement was about the TCK, not about the patent licenses (to which, yes, the TCK is a gateway).

It's certainly dirty what they're doing, but they probably can't be attacked on these grounds.

Reply Score: 3

RE[2]: Weak - not at all
by joshv on Wed 6th Oct 2010 10:46 UTC in reply to "RE: Weak - not at all"
joshv Member since:
2006-03-18

No, if Sun had granted royalty free, perpetual patent licenses to the Apache Harmony project and Oracle had tried to withdraw them, you might have a point.

Oracle published an opinion about Sun's mechanism for controlling access to those patents. Nobody is disputing the fact that the Harmony project never obtained those patent rights.

Reply Score: 2

RE: Weak
by joshv on Wed 6th Oct 2010 13:53 UTC in reply to "Weak"
joshv Member since:
2006-03-18

Jeez. Post a rational comment that violates the local groupthink and get downmodded.

Reply Score: 1

Google should have used Qt
by nt_jerkface on Wed 6th Oct 2010 05:01 UTC
nt_jerkface
Member since:
2009-08-26

They went with a non-standard Java and Dalvik to discourage porting to other Linux based phones.

I also don't buy their claim that JME wasn't good enough for smartphones.

Google is not the internet savior or even a supporter of Linux. They are just good at convincing everyone that they are not like MS or Apple.

By using Qt or JME they could have encouraged ports to desktop distributions. Their Dalvik stack is all about lock-in, sorry if that is too harsh of a reality for some people here.

MeeGo is what a portable Linux should look like. You Linux advocates jumped on the wrong bandwagon.

Reply Score: 3

RE: Google should have used Qt
by lemur2 on Wed 6th Oct 2010 05:04 UTC in reply to "Google should have used Qt"
lemur2 Member since:
2007-02-17

They went with a non-standard Java and Dalvik to discourage porting to other Linux based phones. I also don't buy their claim that JME wasn't good enough for smartphones. Google is not the internet savior or even a supporter of Linux. They are just good at convincing everyone that they are not like MS or Apple. By using Qt or JME they could have encouraged ports to desktop distributions. Their Dalvik stack is all about lock-in, sorry if that is too harsh of a reality for some people here. MeeGo is what a portable Linux should look like. You Linux advocates jumped on the wrong bandwagon.


What exactly has any of this rant got to do with the thread topic?

PS: BTW: AFAIK, Meego includes Mono by default.

http://abock.org/2010/05/26/banshee-for-meego

Edited 2010-10-06 05:13 UTC

Reply Score: 1

RE[2]: Google should have used Qt
by molnarcs on Wed 6th Oct 2010 05:31 UTC in reply to "RE: Google should have used Qt"
molnarcs Member since:
2005-09-10

Google is kinda cool, most people here seems to think so at least. Me included. The lock-in tool is kinda nonsense considering all their tools are open source, and seeing how wonderfully hackable my Nexus One is (built in unlock functionality, root, hell I saw people run the desire ROM from HTC on the Nexus). Not to mention that they provide tools that make it easy to migrate your data from their services (gmail, calendar) - how many companies do that? You wanna switch to a competing product? Fine, click this button to export all your contacts (or calendar data, blogspot, etc) into a standard, portable format (this functionality was added after gmail became popular).

Most companies dedicate a lot of their efforts in locking you in. Companies that at least don't put effort in it are OK. Companies that actually make it easy, in my book at least, are cool. A lot of people like Google for the same reason. Or other reasons. So there always must be someone (it's a feature of all discussion boards) who screams in your face STUPID! Don't you see? You are just buying their propaganda. They are evil because [insert cliché here]. It's usually the "all companies are evil because they want profit" argument.

So he just wants to say that Google is EEEVIL, that's all.

Reply Score: 4

nt_jerkface Member since:
2009-08-26


What exactly has any of this rant got to do with the thread topic?


Google wouldn't have been sued if they went with Qt or JME.


PS: BTW: AFAIK, Meego includes Mono by default.


Banshee is replaceable. The risk would be with a Linux distro that is packed with irreplaceable Mono software that uses Winforms or an SDK that makes use of it.

Reply Score: 2

lemur2 Member since:
2007-02-17

Banshee is replaceable.


Absolutely.

So why have Banshee as the default, and thereby require Mono, when having any of a number of alternative media players would have avoided the need for Mono entirely?

Hence, Meego has lost the plot, and therefore, any support they otherwise may have had.

Reply Score: 2

emilsedgh Member since:
2007-06-21

Meego hasnt lost the plot. Actually, Meego has not joined the game just yet.

Android was a rush into market. And it has been very successfull, relatively.

However, Nokia's plans were more long term. The GTK+ -> Qt switch is just a proof of the fact that they care about technology they use and its quality. They want no rush into the market, so they spend good time on getting there.

I think Meego will truly join the game in 1-2 years and it will kick ass.

Reply Score: 4

lemur2 Member since:
2007-02-17

Meego hasnt lost the plot. Actually, Meego has not joined the game just yet. Android was a rush into market. And it has been very successfull, relatively. However, Nokia's plans were more long term. The GTK+ -> Qt switch is just a proof of the fact that they care about technology they use and its quality. They want no rush into the market, so they spend good time on getting there. I think Meego will truly join the game in 1-2 years and it will kick ass.


Yes, Meego has lost the plot, because it currently completely un-necessarily includes Mono.

Strangely enough, if there is going to be uptake of Meego, then I think a variant called Smeegol just released by of all people OpenSuSe has a better chance.

http://www.phoronix.com/scan.php?page=news_item&px=ODY1Ng
This openSUSE team is now announcing the first public release of Smeegol, which combines MeeGo with openSUSE into one netbook-designed Linux distribution.

Making Smeegol different from upstream MeeGo or openSUSE is that this distribution is based upon openSUSE but with the MeeGo User Experience while making a few changes along the way like dropping in the Mono-based Banshee music player, NetworkManager-powered network configuration, a newer version of Evolution Express, etc. Any independent users can also build their own Smeegol-customized OS using SUSE Studio.


Now that appears to be a Meego variant "finding the plot again" in my view.

Edited 2010-10-06 23:11 UTC

Reply Score: 3

lemur2 Member since:
2007-02-17

Now that appears to be a Meego variant "finding the plot again" in my view.


OOps, I spoke too soon it would appear.

http://news.opensuse.org/2010/10/06/announcing-smeegol-1-0/

Smeegol offers the latest Banshee’s powerful music player


Note to self - lessons to be learned:

1. Meego has still lost the plot.

2. Don't trust Phoronix reports.

Reply Score: 2

Richard Dale Member since:
2005-07-22

"Now that appears to be a Meego variant "finding the plot again" in my view.


OOps, I spoke too soon it would appear.

http://news.opensuse.org/2010/10/06/announcing-smeegol-1-0/

Smeegol offers the latest Banshee’s powerful music player


Note to self - lessons to be learned:

1. Meego has still lost the plot.

2. Don't trust Phoronix reports.
"

You are replying to a story about Oracle suing Android over reimplenting bits of Java, and you are more concerned about whether MeeGo includes Mono to run a music playing application. And you talk about 'losing the plot'?

Reply Score: 2

emilsedgh Member since:
2007-06-21

I dont like Mono too and I hate the fact that its included in Meego, however:

1) The inclusion of Mono has nothing to do with Meego losing the plot

2) I guess sooner or later Meego will drop it too. As soon as they have a proper replacement for Banshee.

Reply Score: 2

gnufreex Member since:
2010-05-06

There is already proper replacement for banshee. Amarok for example. It is even QT-based, old Nokia's favorite.

But reason why new Nokia is pushing Mono, is Stephan Elop. ExMicrosoftie who is now Nokia's CEO. He will probably kill MeeGo because of Microsoft Jihad against Linux. Also, more Microsoft intruders are coming to Nokia. Chief of MeeGo group got fired (or quited after seeing writing on the wall).

Reply Score: 2

RE: Google should have used Qt
by r_a_trip on Wed 6th Oct 2010 09:49 UTC in reply to "Google should have used Qt"
r_a_trip Member since:
2005-07-06

Their Dalvik stack is all about lock-in, sorry if that is too harsh of a reality for some people here.

What is so "lock-in" about a VM which is licensed under Apache V2 and for which anyone and their dog can download the source code and integrate it in whatever platform they like (providing the licenses don't clash)?

What is to stop Linux distributors from shipping their distro's with Dalvik? If GCC was to get a back-end that compiles to Dalvik bytecode, we could even write Dalvik executables in non-java languages.

Reply Score: 4

Innocent until proven guilty?
by r_a_trip on Wed 6th Oct 2010 09:34 UTC
r_a_trip
Member since:
2005-07-06

After reading the comments one thing strikes me as odd. Most people here assume that the claims Oracle America makes regarding their patents and copyrights are valid and that these claims cover Android and Dalvik.

There are a few problems with that. Google didn't use any code from Oracle America (F.K.A. Sun Microsystems). They bought an independently crafted, non-Java VM dubbed Dalvik. They used a set of cleanroom implemented java class libraries from the Apache Harmony project for their translator. (One can assume these are cleanroom, because the Apache Foundation isn't sued for violating the license on the Java specs.) So where is the copyright violation? No verbatim Oracle America code seems to have flowed into Android.

With relation to the patents. Have these patents been challenged in court before and are they valid, non-obvious inventions? If so, does Dalvik infringe on them? If these questions haven't been answered yet, we must assume that Google is innocent and that Dalvik doesn't infringe.

Just because you put in source code with Java syntax in one end of the development, it doesn't necessarily mean that the output somehow must be a kind of Java.

Reply Score: 4

Isn't this what MS did when Sun sued?
by OSGuy on Wed 6th Oct 2010 09:49 UTC
OSGuy
Member since:
2006-01-01

Isn't this what MS did with Java ages ago when Sun sued? Remember Microsoft J++? I think MS added extensions to the language which made apps written in J++ incompatible with other platforms. Sun wasn't happy so they sued. I think MS ended paying heaps of cash...

Reply Score: 2

r_a_trip Member since:
2005-07-06

Isn't this what MS did with Java ages ago when Sun sued?

Not quite. Microsoft licensed Java (specs and trademark) from Sun and with that got the contractual obligation to have their implementation be TCK compliant.

Microsoft did call their implementation Java, but they implemented incompatible, Windows only extensions and included them within the Java namespace. This was a breach of contract and that was the core of the suit against MS.

Google doesn't license Java, doesn't use Sun code and doesn't use a Java VM and doesn't call their implementation of a VM dependent development platform Java. That is why Oracle is claiming copyright and patent infringment as the core of their complaint and not breach of contract.

Reply Score: 3

lemur2 Member since:
2007-02-17

Microsoft did call their implementation Java, but they implemented incompatible, Windows only extensions and included them within the Java namespace. This was a breach of contract and that was the core of the suit against MS.


Precisely so. Because they tried on an "embrace and extend" tactic, so that some programs (using Microsoft's extensions) would only run on Windows, yet Microsoft wanted them to be called Java ... Microsoft broke a contract with Sun.

Google doesn't license Java, doesn't use Sun code and doesn't use a Java VM and doesn't call their implementation of a VM dependent development platform Java. That is why Oracle is claiming copyright and patent infringment as the core of their complaint and not breach of contract.


I can't see any way that Google can be violating copyright if they don't use any Sun code.

I can't see any way that Google can be infringing any patents if Dalvik isn't Java, doesn't use Java code, isn't cross-platform, doesn't attempt to be "write once run everywhere", doesn't work like Java does, doesn't run Java programs (bytecode), and cannot do what Java can do.

If Dalvik infringes Java patents, then so too does every other VM that isn't Java. It would seem in fact that if Dalvik infringes Java patents, then so too does every other program ever written, because the description in the paragraph above would fit them just as well.

Of course, as you point out, Google cannot have broken a contractual obligation with Oracle if there is no contract (Google doesn't license Java).

Edited 2010-10-06 12:04 UTC

Reply Score: 3

lemur2 Member since:
2007-02-17

Isn't this what MS did with Java ages ago when Sun sued? Remember Microsoft J++? I think MS added extensions to the language which made apps written in J++ incompatible with other platforms. Sun wasn't happy so they sued. I think MS ended paying heaps of cash...


Microsoft called their product the Microsoft Java Virtual Machine. It wasn't Java. Microsoft wanted to try to sell "a better java than java" ... which would only run on Windows.

http://en.wikipedia.org/wiki/Microsoft_Java
In 1998 a new release included the Java Native Interface which supplemented Microsoft's proprietary Raw Native Interface (RNI) and J/Direct. Microsoft claimed to have the fastest Java implementation for Windows, although IBM also made that claim in 1999 and beat the Microsoft and Sun virtual machines in the JavaWorld Volano test.

Microsoft's proprietary extensions to Java were used as evidence in the United States v. Microsoft antitrust civil actions.

Begging to be sued.

Dalvik isn't java, and it doesn't claim to be java.

Edited 2010-10-06 11:47 UTC

Reply Score: 3

Drop Java
by FunkyELF on Wed 6th Oct 2010 12:02 UTC
FunkyELF
Member since:
2006-07-26

Use your new Go language or Python, or finally provide first class C / C++ support.
Use LLVM as your VM.
For apps already out there create an bytecode translator.

Java is dead on the desktop / web anyway, lets make it dead on mobile too.

Reply Score: 1

Why care
by bfr99 on Wed 6th Oct 2010 13:24 UTC
bfr99
Member since:
2007-03-15

Someone should explain to me why I should care if large corporations trade lawsuits and monies.
As many others have pointed out if your religious beliefs require it then by all means avoid Java and Android and products using these technologies. Presumably there are phones available that are patent free.

Reply Score: 1

Alternative to MySQL (not Postgres)
by Magma on Wed 6th Oct 2010 14:10 UTC
Magma
Member since:
2006-03-07

Is there a suitable alternative for moderate sized databases of a few gigabytes? SQLite is good but has trouble with very large tables when I tried a few years ago.

Or Maybe a combination:

SQLite = small-ish datasets and config
Firebird = for million+ row tables?

Reply Score: 1

Hiev Member since:
2005-09-27

Firebird works OK for me with simple DBs (less than 100 megs.) and with databases with million of rows (about 4+gigs). I can actually embedded into my programs like with SQLite.

Give it a try.

Reply Score: 2

Magma Member since:
2006-03-07

I'll have to try Sqlite again for large datasets.

Reply Score: 1

Why don't you hate Microsoft even more?
by Sabon on Wed 6th Oct 2010 17:53 UTC
Sabon
Member since:
2005-07-06

I just don't understand how anyone can hate Oracle and not hate Microsoft even more. The examples are numerous.

Reply Score: 2

Patents
by Moochman on Thu 7th Oct 2010 01:04 UTC
Moochman
Member since:
2005-07-06

I always felt like Android was a lock-in hack excuse for not supporting some "real" version of Java, for instance the woefully underutilized smartphone-friendly CDC variant of Java ME:

http://www.oracle.com/technetwork/java/javame/tech/index-jsp-139293...

Truth is, not just the Android platform, but also GWT and to some extent Google App Engine are effectively encouraging the development of non-standard Java code that's tied to Google's own platforms. So I can understand why Google and Oracle would have problems seeing eye to eye. (Of course, it's worth keeping in mind that "non-standard" really just means "not developed in the JCP", which of course is of somewhat dubious significance anyway. If Java developers prefer GWT over JSF, well honestly, there's nothing you can do about it--after all, GWT components tend to be a bit better. But I digress.)

But! Despite everything, I still want Oracle to lose this lawsuit, because I don't support the use of overly broad software patents to fight corporate battles. If Oracle wants to compete (or reach agreements) with Google, they're going to have to find better ways.

Edited 2010-10-07 01:24 UTC

Reply Score: 2