Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Google Bowls a Googly - Possible invalidation of Oracle's patents (computerworlduk.com)
108 points by bluesmoon on Nov 12, 2010 | hide | past | favorite | 32 comments


Without getting into detail (or touching on the related copyright issues), I would sum this up as follows concerning the software patents at issue in this case:

Google is taking its cue, of all things, from the Supreme Court's Bilski decision, which upheld business method and software patents in general but which also gave lower courts a powerful tool with which to smash them if it liked - and that is to invalidate them as unpatentable "abstract ideas" if they did not satisfy the so-called machine-or-transformation test and if there were no other basis by which to remove them from the abstract-idea category.

Many here were deeply disappointed in the Bilski outcome but, take heart, that battle is not over yet. Even while rejecting the machine-or-transformation test as an exclusive test for saying that a particular business method meets the test for patentable subject matter, the Supreme Court nonetheless upheld that test as a useful indicator of whether something was patentable or not. If, for example, a software patent failed to meet that test, then a court could potentially conclude that it did not constitute patentable subject matter. What then? In all likelihood, a court would say, as did the Supreme Court in Bilski, that the challenged patent fails as constituting only an abstract idea, which is not patentable.

Courts have more and more been using the Bilski precedent as authority to strike down software and business method patents on abstract-idea grounds. Bilski pointed the way and many courts are following. (See, e.g., http://ipspotlight.com/2010/08/19/district-court-rules-inter...)

The Groklaw piece referred to in this article discusses this in more detail (see http://www.groklaw.net/article.php?story=20101111114933605). Moreover, in an August 13, 2010 piece (http://groklaw.net/article.php?story=20100813112425821), Groklaw reported that "Groklaw member Celtic_hackr went through [the Oracle software patents] and sees none that are tied to any specific machine."

With its latest filing, Google is revealing its strategy as one of taking on Oracle's software patents head on and asking the court to declare them invalid on the abstract-idea ground (among others). This is very aggressive and has potentially huge consequences.

Bottom line: Oracle can go fishing for opportunistic dollars but it will do so against Google only at a potentially devastating cost to its newly-acquired Sun assets. In the precatory words of Groklaw (in the item last cited above): "And may software patents crash into the ocean as an unintended consequence of this patent attack."


Before people celebrate, remember that Google's arguments are negotiations via the legal system, not a battle-to-the-death over principle.

If Oracle offers Google and its ecosystem a reasonable deal, in which Google publicly acknowledges that Oracle's patents are valid, the matter will be resolved -- and Oracle's patents will be stronger than ever, presumptively, because they were legally tested by a dispute with a deep-pocketed adversary who eventually agreed to license them.

Compare also: when first sued by the Author's Guild in a class action over book-scanning, Google originally argued they had a fair-use right to scan/index the books without prior permission. When a settlement deal favorable to Google's continued operations was proffered, Google stopped fighting for the general fair use principle, and instead has been defending the benefits of their settlement agreement. [1] What should others who want to scan/index books do? Now, Google suggests they should get explicit permission beforehand, using the precedents Google set. [2]

I don't blame Google for making necessary practical legal compromises for their mission, but I also don't put them up on a pedestal, as if they were a disinterested champion who will establish equal rights for everyone. They'll pursue their organizational interests, as with any other ambitious organization.

[1] http://www.publicknowledge.org/node/1828

[2] http://www.nytimes.com/2009/10/09/opinion/09brin.html?pagewa...

"If Google Books is successful, others will follow. And they will have an easier path: this agreement creates a books rights registry that will encourage rights holders to come forward and will provide a convenient way for other projects to obtain permissions. While new projects will not immediately have the same rights to orphan works, the agreement will be a beacon of compromise in case of a similar lawsuit, and it will serve as a precedent for orphan works legislation, which Google has always supported and will continue to support."


In general, I'm inclined to agree with you, certainly with any other company, and probably with Google..

But Page and Brin probably have the muscle to make a bad business decision and fight out the patent if they think they can win, and it's far reaching enough that they might go for it.


Count on Google to fight the good fight! Like they did when they gave up Net Neutrality to make a deal with Verizon.


Playing devil's advocate here, but their deal would secure net neutrality - for broadband. It's only mobile that gets fucked over. Now, anyone can see that mobile is the future, so this isn't much to hang your hat on, but it is something.


You're probably right, but let's see, it's not necessarily a rational self interest versus altruistic community spirit kind of question.

In the case of book scanning, Google didn't have any other way to scan books. But in the case of Android, the cost of buying off Oracle and the cost of replacing the Android app runtime with V8 or Go have to be weighed against each other. There's also the open question of the Android/Chrome OS duality. If Java becomes too expensive for them, they might just resolve this issue with a unified Google OS that doesn't use Java at all.


>they might just resolve this issue with a unified Google OS that doesn't use Java at all.

I don't think that would be feasible. That would mean deserting all their Android developers, telling them that their investment in Android skills is now worthless. Nobody would develop for Google's replacement OS.


I also think it is unlikely but if Oracle wins outright and demands excessive royalties Google might be able to justify the move. And it's not exactly rendering all skills worthless if you're forced to use the very same API from JavaScript or Go instead of Java. I would think it is much more likely that they would use V8 since it is already on every Android device and everyone already knows JavaScript.


For those in the US, a googly is a way of delivering (pitching) a ball in Cricket. The American equivalent terminology would be 'throwing a curveball'


Just being pedantic, curveball is closer to swing. Perhaps the closet too googly would be a splitter. That said, from the perspective of the language/idiom, curveball would be appropriate usage in the context ...


And for those of us NOT in the US a splitter is...


A split-finger fastball, looks like a fastball then suddenly drops down before it crosses the plate.


Its now frequently a third pitch (in addition to the "normal" fastballs mixed usually with a change-up or curve ball and sometimes a slider). Generally a higher risk/reward pitch than any of the aforementioned due to challenges in controlling its motion.


Thank you, I kept expecting Googly to be some new product they were unveiling related to side-step the fight in some way or fight it.


So you're saying that anyone outside US knows about this Cricket thing.


Cricket is probably the #2 sport in the world.. http://www.mostpopularsports.net/

Of course, I assumed that American football would be right up there : Turns out, it's a minority thing, by a long way.


Take that list with a grain of salt considering they left out rugby completely.


They seem to have called it "Football", i.e. lumped in with American Football.


Have you heard of India?

(I've had a few random conversations with Indians while travelling recently, and they knew more about the players and results in domestic leagues in my own country than I do)


I live in Turkey where there's no such thing as Cricket. No one knows about it.

If he starts with "For those in the US", he's assuming that the rest of the world knows about it. Rest of the world != Commonwealth countries.


No:

(in america => does not know cricket) <=> (knows cricket => not in america)


Ironically the first international cricket match was between the USA and Canada in 1844 - http://en.wikipedia.org/wiki/United_States_v_Canada_%281844%...


By offering 20 defenses, Google is showing it intends to fight the case just as aggressively as Oracle. In another sense, they're not revealing much at all, but rather raising every conceivable objection now to keep their options open (as a case progresses, introducing a new argument if an existing one turns out to be flawed becomes problematic).

The most interesting thing - though I don't know enough to guess how far they will get with this one, is paragraph 3 under 'Defenses': '3. Each of the Patents-in-Suit is invalid under 35 U.S.C. § 101 because one or more claims are directed to abstract ideas or other non-statutory subject matter.' This particular argument would, I believe, invalidate most software patents if sustained at trial. I am not a lawyer, and especially not a patent lawyer, just a pre-law student with a hunch.

It will be interesting to watch this case unfold. Judge Alsup wants to retire from the bench and return to private practice, but he has a very good reputation for handling complex litigation cases (and explaining the legal issues clearly in his opinions); this one looks to be time-consuming, but would also boost his reputation significantly.

Edit: snap ;-)


Exhibit J is interesting. So much of this code looks like boilerplate to me that I don't think it says much about whether it was copied. Given standard naming conventions, the public interface, and the simplicity of the code, a lot of developers would create almost identical implementations. I mean, the code doesn't actually DO hardly anything.


The mXZY naming isn't standard. This piece of code is clearly copied in some shape or form. Google's defence also doesn't claim that it's not copied as I understand it.


Some aspects of it look copied to me - the class's 10 fields are in identical order. Some parameters to the function look lifted straight from code generated by disassembling bytecode (flag flag1 set set1). However, these may be explainable by some other means.

http://www.wolframalpha.com/input/?i=permutations+of+10+item...


They were probably listed in that order in the documentation, and the javadocs are generated from the source.

It'd be more suspect if they were in a different order, especially if they were ordered as if they were in a hashtable that was iterated over.


Interesting but unconvincing theory. I have never seen public API documentation that lists private fields before. This certainly makes no mention of private fields: http://www.docjar.com/docs/api/sun/security/provider/certpat...


"better search" is Evolution. If Google is able to create precedent invalidating the software patents system, that would be a Revolution (not that a Revolution with all its mess is always, or even sometimes, a best thing though :)


I think we all know how this will end: a joint-licensing agreement that allows Oracle to go after or threaten the next VM vendor.


I know this is entirely tangential, but did anyone else notice how ugly Google's code is? Sun's is neat and clean; Google's is a mess.

http://groklaw.net/pdf2/OraGoogle-36-10.pdf


While it is Oracle vs. Google, Oracle is buddying up to Apple at the same time, this is one of the best viral marketing campaigns ever. Attach/Attack current big names and be very visible.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: