Java daddy says Sun engineers ran 'goofiest patent' contest
Sun engineers once ran an unofficial competition to see who could get the "goofiest" invention past the US patent office, according to former Sun man and Java founder James Gosling. In suing Google over its use of Java on Android, Oracle is waving seven Sun patents, and one of them carries Gosling's name. In a blog post sparked …
"In a blog post of his own, Java coder, JRuby lead developer, and former Sun employee Charles Nutter doesn't call Oracle's seven patents "goofy." But he does call them "laughable."
"The collection of patents specified by the suit seems pretty laughable to me. If I were Google, I wouldn't be particularly worried about showing prior art for the patents in question or demonstrating how Android/Dalvik don't actually violate them," he writes. "It feels very much like a bunch of Sun engineers got together in a room with a bunch of lawyers and started digging for patents that Google might have violated without actually knowing much about Android or Dalvik to begin with.""
-=-
The patents have already been granted. Showing prior art may be an attempt to invalidate the patent, however the patent is in place and Oracle can and is suing Google for a willful infringement. If Google wanted to challenge the patent on the grounds of prior art, then they should have done so to invalidate the patent.
The fact that on first blush, the lawyers and engineers can view the patents, view Android/Dalvik and make a reasonable judgment call... you have grounds for a lawsuit.
As Gosling said, patents good/bad/indifferent are a defensive measure to protect a company.
Just use the "RISC patent" as an example...
Oracle has more to lose than Google by not suing....
You're advocating that a company's best defense against patent suits is to hire an army of lawyers to permanently go to war against the patent office. That instead of dealing with accusations of infringement on a case by case basis.
Gee, Ian, I can't possibly understand why Google don't just hire you to fight Oracle for them!
"If Google wanted to challenge the patent on the grounds of prior art, then they should have done so to invalidate the patent."
Well why would they even attempt to do this is they did not think that they were violating them?
There are millions of patents out there and a very large proportion of them are very nebulous and a great deal could be invalidated with prior art representations.
But just who has the time, the will and the money to be bothered?
Isn't that what the USPTO is supposed to do and even they can't be bothered.
"You're advocating that a company's best defense against patent suits is to hire an army of lawyers to permanently go to war against the patent office."
Not really, the defensive patent argument goes more like this:
If I get sued or threatened on patent infringement, those having patents are in a more powerful position to counter sue or negotiate a cross license instead of paying royalties.
Some people say open source devs should acquire patents and donate them to an organization like the FSF to help open source developers defend themselves.
The fault with this argument is that acquiring patents cost real money and time. Sure it's possible to acquire patents purely for defensive use, but this represents an overhead that produces no benefits (other than lawsuit insurance).
People like me, who are against software patents, believe that the entire patent system directly and indirectly diverts resources away from legitimate R&D. It's really painful to pay money to support a known broken patent system.
In the end Gosling is right, "Even though we had a basic distaste for patents, the game is what it is, and patents are essential in modern corporations"
Oracle bought Sun to control (almost kill) MySQL. They also got Java which they can use against SAP and they are trying to protect the billions they make from old sun hardware with Oracle licenses.
Which brings the point of Java vs. Google. They hire the same lawyer who got rich off SCO and burned it to the ground.
All I can say is F oracle and Larry can take his gondala (yes he has one) to hell.
I'm not really sure what it is that I said you disagree with. It wasn't my statement, it was Ian's. He was saying that Google should have known well in advance that Oracle was going to buy Sun and acquire Java and use any acquired patents to troll for cash, then (deep breath) Google should have proactively attacked the patent office for issuing idiotic patents, SPECIFICALLY the ones that Oracle were going to sue them over.
The logic. It is flawed.
While I too think software patents are absolutely ridiculous the concept that a company should predict the future and engage a bureaucratic government entity in what equates to a land war in Asia instead of gambling on their lawyers skills in court is absolutely ludicrous.
If you want to talk about a system that diverts resources away from legitimate R&D, imagine a world where software companies are actively trying to get each others patents nullified by the USPTO because there's the risk that someday somebody might sue someone else. That concept is an even bigger joke than software patents themselves.
Exactly the point.
I only disagree with the cost of getting the patent. When comparing the amount of money to be gained, the patents are relatively cheap.
The only time patents fail to protect a company is when a competitor creates a better product because they had to think of a way to get around an existing patent.
I have always respected Joy and Gosling. I lost respect to Schmidt when he went over to Google. People don't understand that Schmidt is smarter than he sounds. He's just trying to spin BS to make their goals seem less *evil*. He's just a bad liar.
No,
I am not advocating anything in terms of how to defend against patents.
IMHO software patents in general are a bad thing. (But that is a whole different discussion)
What I am commenting on is that Nutter's comments seem to trivialize the fact that software patents today are used to inflate the barriers to entry and to limit the competition giving their products time to mature and gain a foothold.
If you're the size of Google, Oracle, IBM, etc ... you have to have Intellectual Property Attorneys on staff and in house. (You also need to hire some of the best and brightest if only to establish a conflict of interest)
You have to recognize the fact that companies are using patents not to protect IP but to reduce competition. Once the barrier is created, then the burden of removing the barrier is upon the competitor. In this case Sun created the barrier. Google willfully tried to circumvent the barrier. Because of the millions of dollars (billions even) at risk, Oracle has to spend money to litigate. The net result... Google comes to terms or they fight it to the end. IMHO Google will lose unless they can sell the idea that the PDA phone is not a 'mobile' device but a mobile desktop.
Were Android a tablet and not a phone... they might have had a chance.
@ Ian
First you wrote:
"If Google wanted to challenge the patent on the grounds of prior art, then they should have done so to invalidate the patent."
Then you wrote:
"I am not advocating anything in terms of how to defend against patents."
Oh! I see! It's all completely clear now. Oh, wait, except the opposite of clear.
"I am not advocating anything..."
"If Google wanted to challenge...then they should have done..."
Oh, oh wait...
"...in terms of how to defend against patents."
"...challenge the patent...they should have...[tried]...to invalidate the patent."
All right, I got it, I got it now! Shit, I lost it.
Somebody please help me here, I'm obviously dealing with somebody from the Ministry of Truth.
I think you misunderstood...
I'm not saying how Google should defend itself against the lawsuit, meaning I'm not advocating any strategy.
My comment about when to invalidate a patent is meant that any company, in this case Google wants to argue about the validity of the patent, they need to do so before they are getting sued by the company that holds the patent. Regardless of the validity of the patent, at the time of the lawsuit, its in place and that is what the court will base their rulings on. They don't care about the validity of the patent just that the company being sued did infringe upon the company.
Clearly you don't understand the law (US Law).
You also don't understand the logic behind using patents outside of patent trolls.
I'm fairly sure, damn near positive in fact, that I'm not the one misunderstanding or failing to comprehend anything in this situation.
This deserves a giant "WTF"....I'm surprised that piece of nonsense got past the article's author.
All instruction codes are interpreted by a processor. Sometimes these codes are interpreted by a formal microcode engine, sometimes they're not, sometimes its a bit of both. It depends on what you're trying to achieve -- trade offs between complexity, size, power, die size.
Nothing to see here....
Even the concept of portable VMs that can run native code is older than anything mentioned in the patent - I was writing business software in 1990 that "compiled" into code that was most definitely interpreted by a VM, as it would run unchanged on machines as diverse as pdp11s, 8088/80x86 or 68000 boxes, etc, as long as they had the suitable runtime, but into which you could add in native code subroutines. And the OS was (c)1982 on the login screen.. (and it's still going strong, although it looks more like a windows app these days..)
SAP: Please send fat cheque for using our patented JVM patents you blatantly copied and used for your new SAP/R3 components. What did you smoke when you believed in the free/cross platfrom/nice guy hype we told you ??
Besides, you are a nasty competitor in the ERP space. We hate you.
IBM: We request at least 700 million dollars/year for using our Valuable Patents you use in WebSphere. Alternatively, a 30% cut of WS sales. We should also talk very seriously about you using our precious Java tech in that nasty Derby database.
Peugeot Citroën: Please fork over 50 Euros/year for each OpenOffice seat. That's at least 1 million Euros/year according to our statistics. And when you are at it, could you please tell us the number of French Gendarmerie Procurement ? They owe us 70000*50 Euros/year.
Using our Precious StarDivision Patents without coughing up !
All of http://wiki.services.openoffice.org/wiki/Major_OpenOffice.org_Deployments: You are a bunch of retards who will believe in fairy tales and we will sue you like Peugeot and the Gendarmerie.
Wordpress, Wikipedia: We suggest you enter negotiations for Royalty Payments for blatantly using our Valuable Patents in MySQL without license. Maybe 50% of wikipedia donations would calm us down. Or 15 million US$, whatever is higher.
Redhat: Please see our demands to IBM. As a courtesy, we will demand only 300 million $/year for JBOSS' use of JVM patents.
Alfresco: How much money do you make and would you be fine with forking over a 50% cut of revenues ?
==========
My comment:
We are seeing this very, very clear now: GPL and most of the other funny licenses are actually a TRAP, that will allow Retroactive Payment Demands. Microsoft is a much more honest company than SUN/ORACLE, who will first give you a sweet cookie, let you live under the impression it is free and then decide to charge you while the cookie is being digested.
What's next ??
Mr Larry could also buy
* Zend Labs (PHP)
* Alfresco (instead of suing for JVM use)
* Digium (Asterisk VoiP server)
* Ingres
* Canonical
and use their patent portfolio to suck the blood out of unsuspecting users of their (mostly) trivial patents. Does anybody know how many patents these companies have ?
What's a good strategy against this ?
LyX, Abiword, KOffice instead of OpenOffice ?
Postgres instead of MySQL ?
Python instead of PHP ?
Whatabout Canonical, Novell and Ubuntu oder SuSE ? Any patents filed ?? Go with BSD ?
I will right now deinstall Java and OpenOffice from my computers. All people who want to send me a document need to use PDF, TeX or something AbiWord understands.
I think you missed that Sun was a great company, with a good social concious.
Oracle bought Sun when their executives wanted a quick cash out, and now own the assets and can do what they want with them.
Despite all the talk here though, there is a difference between enforcing right you have on technology that you believe another company has stolen, and open source code.
Open office is open source code, Sun/oracle/Star/GOD can't take it back.
Same with MySQL and any other OPEN source software.
Java gets murky and I'm not an expert here. I believe the basic language tools etc, are free. but of course Sun also produced tools, and intellectual property that they DIDN'T give away.
I'm guessing that Oracle is going after proprietary software and code that is NOT part of the open efforts Sun has done in the past.
Given that its going to trial, there must be something there.
Now that said, its not to say they can't seriously slow down development for a bit in the open source area. Open office and MySQL were dependent on Sun for some guidance, and grunt work. Oracle can cut of this supply, but its not their only supply.
Volunteers, or other corporations would have to step up and manage their own distribution.
If everyone gets on board and moves forward this will not be a big deal, history teaches us reality will be different. There will be a fork or forks and infighting over what is the proper way forward. Eventually everyone will migrate to the 2-3 best camps and one will emerge as a front runner. At this point things might actually start to get better - look at how great RedHat, Ubuntu and say Debian have done for Linux, with more than one option available, and certain flavors getting better at particular niche roles.
There is an off chance they throw a party and no one comes, no one can be bothered to do the coding work, keep updates relevant to keep the software working, manage a repository to host the files etc etc. If no one steps up open source software can "die off" by just not being updated, sure the old code is there and available, but less usefull as time goes on when not updated. I really am sure that IBM, Google and others will either join forces to create a common distro for all these tools, or go it alone and compete with each other.
Oracle has made it clear that they will sue and license code and IP which they have released as open source. They have released Solaris and tech as open source, they have released Java and the Hotspot VM as open source, they have released Glassfish as open source and so on. It's not about proprietary code it's about being evil. Sun allowed unlicensed JVMs since before Java was a widely used platform. The reason for the MS suit is because they broke the agreement they made, not because of competition and splitting per say, they would have been fine with it if it wasn't called Java. But that was also based off Sun code. The agreement was that it would be compatible it wasn't. OpenJDK for example is compatible and Dalvik doesn't claim compatibility, neither is anything stopping Oracle from releasing a Java ME environment and JVM programmed via the NDK. Or stopping the phone makers from putting a third party JVM in there. Dalvik doesn't claim to be a JVM or Java runtime only a proprietary virtual machine.
Open source infringes on patents just as much as other software licenses does. There's open source code in Android which you explicitly can't use without a patent license. Source and IP isn't the same thing and open source isn't public domain.
> Open office is open source code, Sun/oracle/Star/GOD can't take it back.
I did a quick look at http://www.openoffice.org/licenses/lgpl_license.html for an example, and the LGPL there did not discuss patents at all. I am not a lawyer, yadda yadda, but suppose the following happens:
Suppose there's some patents that could apply to OO that Oracle owns. Or even makes some patents covering parts of OO explicitly. Users of OO are safe, I'd guess, but Oracle could claim that those that distribute a fork of OO are violating the patent because the fork is a different product. Since the LGPL doesn't mention patents, only licensing and transference of source code, all the vendors of the 'different products' might be called unlicensed violators.
And if Oracle were to discontinue working on OO, then all new versions of OO would be forks, and be targets. Effectively, an easy way to setup submarine patents with the results that was suggested.
Sure, the code can be seen, but with the patents in the way, it may be look, but don't touch. Please, please tell me I'm wrong.
@JavaOpenOfficeMySQLShaft:
OMG are you serious? This is just a suit about Google's attempt to circumvent the licensing that they would otherwise have to pay to Oracle for creating a non-compliant VM (Dalvik), not about suing every product in the face of the earth that uses standard JREs, there is no legal basis to do that and shows that you either a) don't grasp what's going on, or b) you think everyone here is an idiot.
Let me make this clear: your post is simply garbage, this has nothing to do with suing people or products that use standard JDKs such as Sun's, OpenJDK or IcedTea... Go ahead and uninstall all Java, OpenOffice, MySQL, Alfresco and other SUN software, but you're just being a child. It's like going to uninstall all .NET, IIS, SQL applications because Microsoft is not putting resources into dynamic languages such as IronRuby anymore, it's daft and irrelevant to what this is about.
And to think that they haven't even won the suit yet! Jesus, you're such an M$ astroturfer.
In what way would using proprietry software protect me from patent litigation ? It is very clear that end users can be sued along with vendors, that's why Novell & MS were selling patent insurance.
Why else would this happen : Microsoft Buys Patent Insurance
http://dotnet.sys-con.com/node/1275104
Buy you're half right, what's a guy to do ?
.. is to make people think about "Commercial Open Source" and the attached risk. I admit not being an expert on the different GPL versions.
Indeed SUN has behaved quite well in the past, but Oracle is turning really nasty and what I wrote is the worst-case scenario.
You are seduced by "Free Open Source Software", integrate it into your business and then Larry Ellison shows up to Charge You Retroactively. That is ,at least ethically, plain fraud. MS tells you upfront that they want hefty amounts of dollars you know about. Larry sends the bill aftwards. And if you don't like it, he has the power to shut down your business completely.
I have learned that GPLv2 contains some defences against this; maybe TheReg could write an article on the subject of OS licenses and Patent Protection.
A nasty Commercial Open Source Supplier could actually collaborate with a third party, integrate their patents and release their cOS wares. After some strong adoption the Third Party would sue the pants off the users.
Such collusion is easy to perform if the programmers have the explicit order to use a specific algorithm.
I was a great fan of OpenOffice for a long time, but that's it. Larry destroyed the trust into SUN/Oracle and Open Source completely. I always thought Richard Stallman were a bit too extreme, but at the moment I have much more sympathy for him than for Larry Ellison.
Code can infringe a patent whether it's closed or open.
The unfair advantage of closed-source code is that it may hide patent infringement well enough that the patent owner never notices. It's far harder to take a collection of binary blobs and decompile them to see what algorithms are being used, than to read commented open source code in a high-level language.
And yes, the way to fix this is to scrap software patents. The right protection for software is copyright (and/or copyleft). Literary authors have known for a long time that although plagiarism may be grounds for a lawsuit, mere imitation is just the sincerest form of flattery. It should be the same for software.
Wow, what a truckload of FUD...
Dalvik is not GPL, therefore is outside Oracle's patent grant. GPL is perpetual, Oracle can't just remove GPL code and patent grant they gave to OpenJDK or OpenOffice or MySQL. All forks of GPL'd projects are off limits for Oracle, they could sue but they can't win. They can sue Dalvik because it is NOT GPL, IT IS ASL, therefore outside of GPL patent grant that goes with OpenJDK, but Google could get Oracle's patents invalidated with prior art. Oracle may lose anyways.
I see you are FUDing against GPL, but there is no facts in your post. Those companies you listed mostly don't have patents, and ones that have are giving the patent grant because they are distributing under GPL. Oracle doesn't need to buy anyone to sue BSD licensed projects(or GPL'd for that matter). They just need a patent. Once they distribute GPL'd project (like ones they bought with Sun), they can't sue anymore against GPL'd forks. That is not true for BSD, since it has no patent grant and distributors have no obligations to users, they can sue whenever they want.
I hate Oracle, but I use things they distribute under GPL. I just strip their logo. It is safe to use. I would be scared to use something they release under BSD because they could come after me every day.
So going around and spreading FUD like you do is doing disservice to everyone. DIcks at Oracle can't hurt Sun's open source projects by suing, they can just shut them down and force us to fork. If they sue, they lose on point that they already waived right to sue.
Fortunately, I can tell you are wrong :-)
OpenOffice.org is released under LGPLv3. That license is written as set of additional permissions on top of GPLv3. So GPLv3 applies, and then you add that set exceptions you linked is called LGPL. Here is GPLv3.
http://www.gnu.org/licenses/gpl.html
GPLv3 is know to have best patent clause of all free software licenses. Check section 11, it is very broad patent grant, even broader than GPLv2. It apply even on patents that Oracle might acquire in the future, even if they stop developing and distributing OpenOffice.org. All those patents are licensed to all users of OOo code, even to forks or unrelated projects that just borow some of the code. (Those projects have to be LGPLv3 or GPLv3, of course. )
So there is no chance in hell Oracle can do anything against OpenOffice.org
Be happy, OpenOffice.org is best protected of all ex-Sun projects, license wise.
It sounds to me that you are Microsoft evangelist.
With all the respect I have for Mr Gosling, Sun sued Microsoft for 1 billion on Java and got it, so patents are patents, although some are stupid ones. The problem is what's granted as a patent. I don't agree with Oracle playing this game but I don't know all the details to be able to dispute that they may have a case.
Sun sued Microsoft not over using their patents, but abusing them intentionally so as to be anti-competitive.
Sun had allowed Microsoft to use Java code, so that people could install and run java code on their windows PC's and allow them to use the java language. (more or less)
Microsoft purposefully broke and changed the way they released Java on Wndows, so that it was not compatible with other versions of Java, and thus killing off Java as competition as a universal platform to code applications on.
Because this was done so obviously and willfully the judge found in Sun's favor, though this was less a case of IP victory but rather a case of Sun winning over MS in an Anti-competition suit.
Actually, it was a license dispute (adjudicated a violation), if I recall correctly. Sun argued that the Java redistribution license didn't permit the sorts of modifications that Microsoft made. Microsoft argued that it did. The Court found in Sun's favor and awarded Sun a (large, in Sun's estimation, small in Microsoft's) sum of cash plus a permanent injunction against specific behaviors by Microsoft within the context of redistributing Java or its (legal) derivatives. But that was a lot of pints ago, so my memory may not be correct.
"Microsoft purposefully broke and changed the way they released Java on Wndows, so that it was not compatible with other versions of Java, and thus killing off Java as competition as a universal platform to code applications on."
Google purposefully broke and changed the way they released J2ME on mobiles, so that it was not compatible with other versions of J2ME, and thus killing off J2ME as competition as a universal platform to code applications on.
There, fixed that for you.
Did Google have a licensing agreement with SUN when they developed their Dalvik implementation ?
Or does SUN/Oracle have the right to determine that certain devices can only run J2ME, irrespective of the user having signed any agreement ?
If yes, Java is a dangerous piece of crap that will submit every user to the mercy of Oracle. Maybe that's the truth - I don't know.
I know that I will not touch any "FOSS" from Oracle in the future.
While it may be possible to design around what you refer to as the "mixed mode" patent the workaround may have a sufficiently large overhead that JIT compilation is not worthwhile on Dalvik. See a a blog posting of mine, http://shape-of-code.coding-guidelines.com/2010/08/15/oraclegoogle-java-patents-lawsuit/, for more details.
My interpretation of the "Mixed Mode" Patent is that it describes an undeniable, irrefutable, unavoidable "process and system" that is "prior-arted" by exactly every programmable computer that has ever come before. At its core, every physical (hardware) execution unit interprets instructions of various types. So do software execution units. The granting of that patent was beyond inexcusable. It represents a complete disconnect between the USPTO (and examiner) and the history and practice of software development. It's as obvious to an "experienced practitioner" as the stupidity of the Evans and Sutherland "Cursor XOR Patent" from the first half of the 1970s or so. E&S was granted a patent on using the boolean exclusive-or operation to put a cursor on a screen and then remove it without disturbing the underlying image. I don't believe in software patents, but I might entertain a counter argument, were it not for the imbicilities that continually come from the USPTO.
Issues patents for complete crap. Just look at
http://www.freepatentsonline.com/crazy.html
I laughed my head off at the animal ear protectors.
The thing that got me was that while I was looking through them, the "Ads by Google" at the top of the page dropped a Rambus ad on me. "Rambus, A company of inventors - Watch the 5 minute video", I came perilously close to asphyxiation.
The only thing I'm trying to work out is whether this means that Rambus have paid to sponsor phrases like "patent troll", "idiotic patents" and "obvious prior art" or whether something at Google has a wicked sense of humour......
... just a potential headline of the future. The number of large software firms in the US is shrinking to only the largest players with the most expensive lawyers.The others have been helped out the door by patent law suits.
It looks more like the US patent system is like one of those ' take a ticket and wait for your number ' systems, but these are tickets to go and sue.
Insignia solutions, whom I worked for many years ago, already produced emulation tech that included "mixed-mode" type execution to run x86 code on UNIX/Apple (SoftWindows I think it was called). They used the same tech to come up with a Java VM called Jeode that used dynamic compilation - there's some prior art!
OR how about my, now ancient, Acorn Electron which could execute BBC Basic, obviously interpreted, and embedded assembly language as well. It even did to a bit or pre-compiling of the BASIC so that maybe counts as well!
>>"OR how about my, now ancient, Acorn Electron which could execute BBC Basic, obviously interpreted, and embedded assembly language as well. It even did to a bit or pre-compiling of the BASIC so that maybe counts as well!"
That was a first thought of mine, thought I'm not sure that really counts if the assembly was only converted at run-time.
However, given the frequency of calls in BBC basic to assembly OS routines, a program being run would be jumping between interpreting code, doing things as a result of interpretation, and making explicit calls to existing routines, either system ones or ones pre-written by the user.
Also, I'd guess that in pretty much any old-school home computer where it was possible to do so, people have been doing the dirty and jumping into machine code hidden in REM statements or POKEd into suitable memory locations whenever they felt the need to.
In reality, the patent would appear to be one of the usual software ones - pretending against all common sense that if a computer can be made to automatically do something pretty mundane that programmers have been doing by hand since interpreters were first written that it somehow miraculously counts as a meaningful 'invention'.
A half-decent patent system would include a facility for punitive damages for companies attempting to get frivolous patents when they should know better, and seriously punitive damages (and/or personal civil or criminal penalties) for people who try and use such patents to stop other people doing business.
"We got sued by IBM for violating the 'RISC patent' — a patent that essentially said 'if you make something simpler, it'll go faster'. Seemed like a blindingly obvious notion that shouldn't have been patentable, but we got sued, and lost. The penalty was huge. Nearly put us out of business."
Oracle should put IBM out of business. Doing the same thing to IBM, as IBM did to Sun.
They can't do anything against JBoss. However, Red Hat have some BPM patents that WebLogic infringes, and btw, those are licensesd freely to FLOSS community, of which WebLogic in not part of. Stop FUDing.
Ok, maybe someone can enlighten me.
Whats so Goofy about Goslings rejected patent?
Dont know, but somehow having a master switch near the door to switch off everything except e.g. lawn lighting (and optionally electric heating in winter) seems like a reasonable idea.
Left the light on upstairs? Just flip the selector and switch it off without doing a lot of running around...
Charles Nutter seems a very fitting name
There's only one way out of this.
The US government must pass emergency legislation to invalidate *all* software patents *now*, and annul any pending claims for unpaid royalties as far as article I section 9 of the US Constitution allows.
Flames, because something good is bound to rise from the ashes of the "IP bonfire".
I certainly agree with A J Stiles here but lets be honest America's legal system is running the show for big business out there and whether we vilify Larry or Eric, (because yes they ultimately push the go button on these decisions), we should reserve a lot of vitriol for the lawyers that push this business approach to them.
In the end no-one wins except the lawyers.
As far as I know, it is a valid 'defense' to a patent infringement lawsuit, in the US as in the UK, that the patent is invalid / not patentable
so the actual lawsuit is a perfectly predictable way to question whether the patent should ever have been granted, there would be no need for a proactive lawsuit from a prospective defendant against the patent office to declare a patent invalid
Now there are questions of general patentability of software, the appropriate penalty for patent bullies, the way the law generally is a bullies' playground, etcetera; these are not however justifiable criticisms of Google's actions in waiting until it was sued to raise the question of whether the patent should ever have been granted
@fionasboots - small world. Yes, SoftPC / SoftWindows had mixed mode 'hotspot'-style interpretation -> compilation in 1990 and possibly earlier, in any event, quite a few years before Java...
