Sun CEO explicitly endorsed Java's use in Android | Hacker News Reader

Sun CEO explicitly endorsed Java's use in Android

zdnet.com

105 pointsmrsebastian33 comments

I think it's a weak argument, actually I think it's a slippery slope if CEOs can't say anything without that implicitly granting all sorts of permissions. It's a valid defense that Schwartz obviously meant to welcome Android to being based on Java, assuming they didn't infringe on Suns intellectual property.

"I'm happy people is so excited about watching our movies" is not an endorsement of pirating said movie.

The crucial difference between this situation and the one that Google is in is that the activity the Sun CEO refered to was the allegedly infringing activity. "I'm happy people are watching our movies" wouldn't be a cause for invoking estoppel, but "I'm happy people are bittorrenting our movies" would be.

Also, in this case the crucial aspect is what the law is, not the quickly considered opinions of non-lawyers like us.

I am not a lawyer, but I completely disagree when estoppel is taken into consideration[1]. Sun publicly endorsed Google's implementation of Java and there is probably plenty of evidence it endorsed it privately as well.

There is a reason a CEO doesn't just spout off whatever he wants without consulting a lawyer. Especially on company letterhead which is what the Sun Blog amounts to these days. If it was a mistake it should have been retracted. But it wasn't, so maybe it wasn't a mistake in the eyes of the law.

And yes, if a film's rights holder gives explicit or even implicit endorsement of people watching the film through piracy, as did Michael Moore [2], that does protect you from him suing you for pirating it. You don't need a license contract to tell you it's OK.

[1] http://en.wikipedia.org/wiki/Estoppel

[2] http://www.youtube.com/watch?v=R7uvRCyVfa0

He endorsed google using Sun Java, not googles implementation of "Java".

This whole line of argument is FUD From pro-google people who are trying to pretend like its hypocritical to want to defend Java.

Android was using Sun Java before the licensing dispute caused them to go their own way.

He endorsed google using Sun Java... Android was using Sun Java before the licensing dispute caused them to go their own way.

This simply isn't true.

Google/Android never used Sun Java, the libraries or the Sun JVM.

Android always had their custom Dalvik VM, and it used the (clean room) Apache Harmony Java-compatible libraries.

I originally clicked reply to a similar message to say much the same thing as what you've said, however the original announcement that he was replying to gave no specifics, save one-

http://www.openhandsetalliance.com/press_110507.html

"Thanks to the availability of our Jbed™ Java™ VM on the Android platform, we offer immediate compatibility to the standard Java ME world to enable Java ME-based mobile services with the Android platform."

So it's a bit fuzzy. I'm on the fence as to what this means, but given that Android was actually based on Java prior to Google's purchase, and this was not long after that transaction, I have to think Schwartz really was referring to a J2ME implementation. Indeed he specifically talks about support for netbeans and the like, which of course is entirely untrue for today's Android development.

So I don't think this is the smoking gun that many think it is. Scwartz perhaps jumped the gun before formalling licensing it, but it isn't some grand endorsement of an Apachy Harmony derivative.

http://blog.yafla.com/Did_Jonathan_Schwartz_Endorse_Androids...

Estoppel is meant as a protection against the following situation. Sun tells Google, "Don't worry. Android doesn't infringe on our intellectual property." Google then relies on that information when making future plans for its operating system. Sun then comes back and says, "we're suing you for infringing on our intellectual property." Sun (now Oracle) would be reasonably estopped from collecting damages because they had earlier made assertions that Google reasonably relied upon in their decisions.*

Google's lawyers note that Sun offered to license Java to Google (http://www.networkworld.com/news/2011/072211-google-sun-offe...). Therefore, there is no reasonable way that Google could try to claim estoppel. Google was aware that Sun believed that Google needed a license. Therefore, there is no possibility that Google believed that Sun was saying that Google's Java implementation didn't infringe on Sun's IP.

There's a distinct difference between that scenario and the one that occurred. In the hypothetical, Sun has stated a fact "your implementation of Java doesn't infringe". Later, Sun changes that fact "your implementation of Java does infringe". With what Sun's CEO has said, one can create a dialogue that doesn't contradict itself. "We're happy you're using Java"; "We're still happy you're using Java, but you have to pay for its usage or code around our patents like we said before when we offered you a license for Java."

*Estoppel is meant to make sure that people can't trick you into situations (and often times requires some proof that the person being estopped had that motive). For example, let's say that Sun had gotten together in a board room and started saying, "we'll tell people that there are no patents on Java and they can all use it for no charge and then later, once they've all started using it due to our assurance that there are no patents on it, we'll change our mind and sue them for patent infringement." That isn't the case here.

"Now, I am not a lawyer, but [...]. Hopefully, the U.S. District Court will see it the same way and throw this case into the trash heap where it belongs."

Maybe I'm just a journalism snob, but that's some pretty poor reporting.

Reading the Wikipedia page on estoppel[1], it seems that Google's case is strengthened much more by Sun's long delay in pursuing legal action against Google than by a casual statement by the CEO. Of course, IANAL.

[1]http://en.wikipedia.org/wiki/Estoppel

Except he congratulates them on using Java, not dalvik. The assumption being that Google would not circumvent the conditions for using Sun's Java tech/IP, naturally.

Right. It does make the $6 billion damages claim that much harder to swallow, though.

Dalvik was already launched by the time of the post. So clearly he implied the Java inside Dalvik.

Amazon uses a custom JVM on Kindle devices and they paid a per-device licensing fee to Sun (and I assume they still do to Oracle). If I recall, this is actually how Sun used to handle their licensing; if you had a completely standards compliant implementation then you didn't have to pay licensing fees, but if you have any sort of custom VM that ran java (even if it was a strict subset or superset) you had to get them to agree to licensing terms.

Maybe I'm off base, but it seems like just because he made a statement that he was happy to see that they were going to use a java-based VM doesn't mean he didn't assume that they were going to pay licensing fees for it.

Nov 2007!

http://androidcommunity.com/forums/f8/sun-welcomes-android-w...

IMHO, this is a much stronger case:

Nothing in the licensing terms will prevent open source projects from creating and distributing their own compatible open source implementations of Java SE 6, using standard open source licenses

From JSR-270

http://www.jroller.com/scolebourne/entry/jcp_bonn_meeting_oc...

That doesn't help because Dalvik is not compatible (and to make it compatible would probably require making it too slow and bloated to run on phones).

It depends what "compatible" means, though, doesn't it?

If you include the Java bytecode-to-Dalvik compiler as part of the "system" it could be considered compatible.

I'm not sure that the Java bytecodoe-to-Dalvik compiler actually completely Java SE 6 compatible (eg that it is a superset), but I think that is irrelevant since their licensing actually prohibits supersets. This is why Sun was able to force Microsoft to discontinue their JVM; it had nonstandard features in it, and therefore their licensing wouldn't be valid.

but I think that is irrelevant since their licensing actually prohibits supersets

Android/Dalvik isn't a licensed Java implementation so the irrelevance is irrelevant ;)

The original statement referred to the license to use a Java compatible API, not to be a Java licencee.

It depends what "compatible" means, though, doesn't it?

It's fairly well defined by JSRs 270 or 216, and Android isn't even close. For example, the GUI libraries are quite different.

If you include the Java bytecode-to-Dalvik compiler as part of the "system" it could be considered compatible.

I don't think the compiler runs on the phone, so it doesn't count.

This article seems to try and make a link between the fact that the blog post is no longer available, and the Google/Oracle court case.

That isn't how it happened - Oracle took down the old Sun blogging site (along with all the blogs) at the same time. It was a typical Oracle move - wanting to control the dialog about their platform - but I don't think it was specifically about this case.

I think we people, who are using and depending on Java and Android shouldn't take these lawsuits too personal. It is just one point where the Oracle army (made of lawyers) found a possible hole in the defense of Google and tries to attack it. Who wins this battle doesn't actually concerns our lives at all (excluding the CEOs of Oracle and Google, of course).

This is extremely optimistic and advocates a total head-in-the-sand view (except ostriches don't really do that).

Companies seem to care very little if their little patent wars leave a little or a lot collateral damage.

GIF, anybody?

> Companies seem to care very little if their little patent wars leave a little or a lot collateral damage.

Totally true and never argued a different point of view. Just consider how killing Android would be a major blow to Google and how likely it is to land such a blow on a company like Google in a situation like this. What I meant was, that considering the context and the chances, it is much more likely that whoever wins, wins some millions, maybe even billions. That will hurt the loser, maybe hurt the companies overall value, but it will not kill Java or Android.

Btw. I think it is really interesting, how both sides in a discussion can think the other one is kind of naive and blind, because they don't really get the point of the other one. Your ostriches-argument is exactly what I thought when I wrote my comment (just exchange optimistic with pesimistic)!

Do you have any sources about how GIF patents ended up doing a lot of collateral damage? Wikipedia says that .png was invented in response but was that really damaging?

It's a trap!

Yes, SUN JAVA, not Google Dalvik. Android was using Sun licensed java in the early days, switched when google didn't like the licensing terms.

No. Android was released with Dalvik in November 2007, when Schwartz's blog post was written.

http://www.betaversion.org/~stefano/linotype/news/110/

Blog post: Nov. 5 Google released dalvik: Nov. 12

So technically Schwartz could have been blindsided by Dalvik.

I'm pretty sure Schwartz was aware of the fact that Google hadn't licence JavaME from them...

How can you be sure? There are other people pretty sure of the opposite

http://news.ycombinator.com/item?id=2809775

Because it would have been a huge deal, the kind of thing they'd make press releases out of. To suggest that the board members of Sun would have been unaware of such a major deal is farcical.

Wrong.

Reply on news.ycombinator.com