Hacker News Reader

Comment by Estragon | Hacker News Reader

I wish Google well in this conflict, but can someone explain to me why Oracle's former position on open-sourcing the TCKs is relevant to the case? I agree that the somersault makes them "look bad" as the OP says, I just don't understand its specific legal implications.

View on HN

>>These registrations appear to relate to versions of certain Sun Java materials that were released as open-sourced software in 2006 and 2007.

I'm not a lawyer (and even Groklaw goes over my head most of the time), but it seems like this is the crux. If the lawsuit is based entirely around items that are open-sourced, then there's no lawsuit. But there's no real way to tell, since Oracle hasn't specified what exactly is being infringed.

And from a comment below:

>>Oracle wanted Java more open when it worked better for them, now it doesn't. They are allowed to change their minds, aren't they?

Yes, Oracle is perfectly welcome to change it's mind and say "We don't want this open source anymore." However, they can't (I believe) un-open-source things that have already been released. They can remove the license, take down the Java.com downloads, and continue developing Java under a proprietary license, but the whole point of OS is that the community could take their last OS release and fork it themselves.

This also means that a company can't open-source software or development tools, wait until they're implemented in various systems, then turn around and "un-open-source" them and sue for infringement.

Google is using a clean-room re-implementation of the JRE, not a fork of the JRE so calling 'OpenSource' doesn't really apply here. Oracle wouldn't have a leg to stand on (iirc) if Google had just forked the JRE.

It's not about un-open sourcing anything. Apache, Oracle _wanted_ Sun to open source the Java testing kit etc. but Sun never did and placed restrictions on usage on mobiles.

So Google is claiming that Oracle is acting hypocritically depending on which side of the aisle it is, but that does not carry any legal weight.

The registrations you talk about in the beginning are to do with copyright claims and have nothing to do with the patent claims.

Actually, that does carry legal weight. It's called the doctrine of unclean hands; those who wish equitable relief must themselves act equitably. Acting hypocritically in this case may fall under that doctrine.

Patents are a creation of equity? Would have never thought!

I agree. Should we pretend that companies don't act selfishly? Oracle wanted Java more open when it worked better for them, now it doesn't. They are allowed to change their minds, aren't they?

I still hope Google wins though.

      Should we pretend that companies don't act selfishly?

As I've said in another comment, corporations are recognized by the law to have rights and responsibilities like real people. For example, corporations can exercise human rights.

Corporations can even be convicted of criminal offences, and it would be stupid to not hold them to the same standards as real people.

I don't know were this trend is coming from (that companies shouldn't be accountable for anything other than making money) but it doesn't do any good.

True as far as you went, but most people are not charged with a fiduciary responsibility to other people. This does tend to take some wiggle room out of corporate behavior, particularly where shareholder lawsuits are a possibility.

I'll make no distinction. People act selfishly too, and there is no law against it. I might vote against welfare, then fall on hard times and plea for welfare. Am I breaking the law? Should the welfare office say "Well, sorry, you used to be against welfare so we can't grant your request?"

But being a hypocrite is not a crime. The original question was: what legal relevance does that have? I have the same question, although I suspect the answer is "none." Although anything that makes Oracle look bad might help Google by biasing a judge, even if only a little.

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

Hypocrisy isn't a crime, but it can invalidate a claim if you publicly state one interpretation of law and later state another.

IANAL, but I'm pretty sure estoppel only applies to legal facts and promises. Merely stating that you think things should be a certain way and then changing your mind when something else would be more beneficial is not grounds for estoppel. Unless Oracle successfully sued for openness or told somebody that it would not do this, I don't see how they could be estopped.

I gathered it was related to the Unclean Hands (http://en.wikipedia.org/wiki/Unclean_hands) claim - but I am just a coder.

Edit: Because "The defendant has the burden of proof" that was included as (part of?) the proof they would need to provide

Well, that's a question I'd loved to see answered. Simply put, I don't know if you can be 'more open source', and then simply change your mind on such a thing.

Perhaps this situation isn't really reflective of what I'm talking about, but I don't think that you can un-GPL code that's already been opened that way or, even if you can, I don't think you can restrict those who are using the version you originally licensed as GPL in any way other than how you licensed it.

It's a complicated scenario, for sure, but I'd like to hear whether or not the sort of bait-and-switch I described is something that can be done, or if there are any precedents either way.

It's not about un-GPLing anything. The GPLed code was given a patent grant _only_ if it was compatible with existing Java. And the code to test the compatibility(TCK) was never GPL'ed and had restrictions.

Engadget has good breakdown:

Google uses a subset of the Apache Harmony Java implementation in Android. Sun open-sourced Java Standard Edition under the GPL in 2006 and 2007, but didn't include a patent or copyright license with the code. In order to get that license, developers have to demonstrate compatibility with the Java specification. The only way to demonstrate compatibility with the Java specification is to use Sun's Technology Compatibility Kit, or TCK, and Sun / Oracle and Apache have been bickering about the license for the Java TCK, or JCK, for years. (That's putting it lightly, actually. It's been more like a war.) The only license Sun ever offered Apache for the JCK included significant "field of use" restrictions, including a restriction on mobile phones. Because of these restrictions, Apache's never taken a JCK license to test Harmony. Oracle used to be on Apache's side in demanding Sun loosen up the JCK licensing restrictions, but that changed as soon as it bought Sun out. Google thinks this is very bad, and that Oracle and Sun are just big bullies who don't want Java to be open, even though being open is super amazing. That's where Google stops, right at the part where unchecked corporate greed threatens to destroy a open-source project and your heart swells with sympathy. It's a good place to stop! But the logical and unstated endpoint to this narrative isn't quite as good: Because Apache doesn't have a license to test Harmony with the JCK, it doesn't have a license for Sun's Java patents and copyrights either. Part of the reason Apache wants a JCK license is to assure its users they have the necessary IP rights. Google knew all this and used parts of Harmony in Android anyway.

In the relief section, Google is asking among other things for ...

     f. A declaration that the Oracle’s claims are barred by the doctrine of 
        unclean hands

They are also probably going to have to explain why they didn't even bother to make Dalvik Java-compatible.

Reply on news.ycombinator.com