Google asks Supreme Court to overrule disastrous ruling on API copyrights

[DarkSwordsman]

Main Article: https://arstechnica.com/tech-policy/2019/01/google-asks-supreme-court-to-overrule-disastrous-ruling-on-api-copyrights/

Quote

Then, in 2014, a court known as the Federal Circuit Appeals Court issued a bombshell ruling taking the opposite view. Oracle had sued Google, arguing that Google had violated Oracle's copyright by re-implementing APIs from the Java programming language. The case has been working its way through the courts ever since, with the Federal Circuit issuing a second controversial ruling in 2018.

 

Here's a snippet of the previous court case from Wikipedia:

Quote

Oracle had come into ownership of the Java programming language including its patents, documentation and libraries through the language's application programming interfaces (APIs), through its acquisition of Sun Microsystems. Oracle made this information freely available to developers to use, but licensed its standard implementation on various platforms including mobile devices. Google developed its Android operating system atop the Java language, including its APIs and a cleanroom version of the standard implementation, to build its own mobile device platform. While Sun had not taken action against Google prior to its acquisition, Oracle became concerned that the Android operating system was a competing product, and filed a lawsuit against Google, claiming both copyright and patent violations. Google claimed that it was unaware of any patent infringements and that its use of the freely available APIs was within fair use allowances. Oracle has sought upward of US$8.8 billion in damages due to the commercial success of the Android system.

 

It appears that Oracle has been dealing with this for quite some time. 

 

I personally think this ruling is a joke in it's current state, but the more I read the licensing and read about other information on the web, I think Java is in the right. From my understanding, here are five licesnses:

Java SE 6: https://www.oracle.com/technetwork/java/javase/terms/license/javase6rilicense-2395278.html

Java SE 7: https://www.oracle.com/technetwork/java/javase/terms/javase7-license-3976698.html

Java SE 8: https://www.oracle.com/technetwork/java/javase/overview/javase8speclicense-2158700.html
Java SE (general?): https://www.oracle.com/technetwork/java/javase/terms/license/javase-license.html
Java SE Binaries: https://www.oracle.com/technetwork/java/javase/terms/license/index.html

 

They're all pretty much the same. At the core, they generally say, "You can use this license to '...create and/or distribute an Independent Implementation of the Specification that: (a) fully implements the Specification including all its required interfaces and functionality...'". However, it also says explicitly, "No license is granted hereunder for any other purpose (including, for example,...distributing the Specification to third parties)" as well as, "You may neither: (a) grant or otherwise pass through to your licensees any licenses under Oracle's applicable intellectual property rights...".

 

I can't find any substantial evidence showing the full extent of what libraries they include, but here's what's detailed under their documentation: https://developer.android.com/reference/packages

 

Whether or not it's right and whether or not the licenses are valid in the first place, I think it's fairly clear that Android did violate Java's licenses. It is important to note, these licenses are under copyright; all licenses are. These licenses simply spell out what you can and can't do with something under copyright law. Essentially, the judges are ruling that APIs should be able to be claimed under copyright. Traditionally, this didn't seem to be the case for APIs since they are designed on purpose for other people to use them.

 

Thinking about it, I think APIs should be able to be covered under copyright, but for certain cases only. Let me give an example:

 

Instagram used to have a wonderful API. You could gather tons of information and use it in great ways. However, they nuked it a couple years ago and now it's two useless endpoints. However, however, if you dig through the network tab on chrome or firefox, you can see all the network calls it makes to their back end servers, or API. 

 

Now, what's the difference between the API they give me and the one I found? Intent. To me, their Public API is intended to be used by anyone for any reason, within reason. Their Private API, while easy to access and navigate, is intended not to be used publicly. I think if we apply this knowledge to APIs and implement it in a law someway, we can probably very clearly define who is in the wrong and right. Therefore, this is what I propose:

 

- Public APIs are APIs that are easily available and portrayed to be used by anyone and everyone.

- Private APIs are APIs that are not easily available (such as an invite basis or requiring more than a simple download) and can only be used by their developer or company and by other developers or companies with explicit written permission and within certain guidelines.

- Anything that is not an API would be source code and can not be modified, re-distributed, and branded as the same source code

- Anyone can create their own source code that is nearly identical to another, and two applications can serve the same purpose in a nearly identical way

- Methods or design that can be reverse-engineered without using another's source code can't be copyrighted.

- However, these methods or design can't be intentionally used in poor taste or in a way to damage another person or company. For example, a method on how to store data in a 3 dimensional array is fairly common knowledge. Figuring out how an encryption pattern works without referencing source code can'tbe copyrighted, especially since at that point it's a major security flaw. However, someone using that encryption pattern to break into a database or steal information (generally hack) is illegal.

 

This idea could certainly be ironed out, but I think those general guidelines make sense. Also, this essentially throws patenting out the window, which I think is the goal of open source, but it keeps things fair and equal.

 

By these rules, Oracle can't sue Google. In my eyes, if they built an API with the expectation that people can use it freely and make it so easy to obtain, they shouldn't have grounds to sue Google for making one of the world's most impactful operating systems just because it ended up being used everywhere.

 

If I go and build a node API module for people to use, I shouldn't incriminate someone for using my API to build their software and let other people use my API through their software to interface with other people's software.

[Fnige]

Hey look... Is that qualcomm

 

No... its their twin

[Guest]

I think creating a mass operating system on millions of phones worldwide is a little out of fair use. For instance maybe something like fair use on the YouTube. You can have a sample of the video in your video and make a little money, now if that whole video was, say, a reaction video and the sample video was a main proponent, the original author might get a little concerned if your video had nearly a billion views, making quite a lot of money.

[straight_stewie]

Personally, I don't think that any software which is shared in any way should be able to be sued over.

We're at a point here where companies are trying to claim that things that are equivalent to steering wheels in cars are their idea that no one else can use. And they know this, which is why we often don't see companies asking big companies to stop doing the thing, just to pay them unheard of amounts of money to be left alone.

Could you imagine a world where AT&T sues someone every time a Unix like OS is distributed?

[williamcll]

I don't understand, who is in the wrong here?

[Fnige]

39 minutes ago, williamcll said:

I don't understand, who is in the wrong here?

both... lets just pretend its both.

[LAwLz]

6 hours ago, williamcll said:

I don't understand, who is in the wrong here?

For the future of all software development, I sure hope it is Oracle that is in the wrong, but who is "wrong" just depends on who the court sides with.

 

Here is my attempt to explain the lawsuit in simple terms.

 

 

Basically, when you program you use what's called APIs to perform various functions.

For example if I want to print something to a console in C, I can write:

Quote

printf("Hello World");

"printf" is the API. However, that in itself is just a name for more complex code which actually carries out the function. 

 

What Google did was take the name of a bunch of Java functions, rewrite the "back-end" of it (which is to say, the code which actually executes when you write for example printf) but kept the API name and general structure of how developers access it the same as in Java.

 

So a function which looks like this in Oracle's Java:

SortNumber(1, 6, 2, LargestFirst);

 

would look like this in Android:

SortNumber(1, 6, 2, LargestFirst);

 

That is to say, identical.

How the two implementations actually went about sorting the numbers were completely rewritten by Google though.

The argument now is if copyright protects the name of APIs or if anyone is free to use the same names as long as the actual content (as in, the back-end code) is different. If it is found that the names of APIs are protected under copyright then all hell will probably break loose since A LOT of software uses the same names and often for the reason of interoperability, portability or helping developers move from one thing to another without having to relearn a bunch of stuff.

For example LinusTechTips has an API just called "api". Call that and you get some info. If this ruling passes, the first person who made an API called "api" could potentially sue LinusTechTips for infringing on their copyright.

 

It's obviously a bit more complicated, especially since Java has different versions with different licenses etc but that's the gist of it.

 

 

To me, this is like Youtuber1 trying to claim copyright on Youtuber2's video titled "Spiderman 2 Review", because Youtuber1 also has a video called that, even though the actual content is completely different. The videos both share the same title, maybe even structure, and people can view them for the same purpose, but does that mean Youtuber1 should have the exclusive right to that title?

[jagdtigger]

If this ruling stands that would be the wet dream of copyright trolls... Prepare for a gory power struggle.

[Trixanity]

It's always fun to have world changing rulings be carried out by people who have no idea about the subject they're to give their verdict on nor have any idea of the implications of either outcome.

[Commodus]

On 1/26/2019 at 11:46 PM, straight_stewie said:

Personally, I don't think that any software which is shared in any way should be able to be sued over.

We're at a point here where companies are trying to claim that things that are equivalent to steering wheels in cars are their idea that no one else can use. And they know this, which is why we often don't see companies asking big companies to stop doing the thing, just to pay them unheard of amounts of money to be left alone.

Could you imagine a world where AT&T sues someone every time a Unix like OS is distributed?

While I'm not saying Oracle is in the clear here, this is a lousy way of thinking about it.  Unless it's unreservedly open source, you have a legal right to enforce your terms.

 

Think of it this way: imagine if you wrote a book and shared it with everyone so long as they didn't alter it or pass it off as their own.  Now imagine that you find out someone has not only falsely taken credit for it, but inserted an erotic fanfic chapter.  Do you just shrug it off and accept that your book is no longer under your control, or take action against that person for violating the terms they agreed to when they received your book?  You'd probably opt for the latter.

 

I'm not arguing against freely sharing things without imposing terms, but that's up to the creator.

[straight_stewie]

4 hours ago, Commodus said:

Think of it this way: imagine if you wrote a book and shared it with everyone so long as they didn't alter it or pass it off as their own.  Now imagine that you find out someone has not only falsely taken credit for it, but inserted an erotic fanfic chapter.  Do you just shrug it off and accept that your book is no longer under your control, or take action against that person for violating the terms they agreed to when they received your book?  You'd probably opt for the latter. 

The problem with software, as opposed to natural language, is that there is very often only a few ways to do something, especially as you move out of user space, or as you move to more general definitions of "doing things".

The problem that we are seeing is that companies like Microsoft have *successfully sued companies essentially for not using Windows. The most famous one is Microsoft v. TomTom. Where Microsoft alleged that TomTom had violated over 8 of Microsofts patents, essentially by using Linux in it's GPS systems.

As a result, TomTom *successfully counter sued Microsoft for Microsofts attempt to create a device that "generates a maneuver for a turn at an intersection", "route generation in a vehicle navigation system", and "Quick selection of destinations in an automobile navigation system".

After having success with Microsoft, TomTom sued Garmin, because Garmin used "an algorithm designed to decide which streets are relevant to a users route". (or in laymans terms, because Garmin used a route finding algorithm to, well, find routes).

Seriously. We came very near to it being illegal for anyone other than TomTom to make GPS equipment, on the basis that ideas like route finding or file selection are wholly theirs.

 

But this takes me back to what I said before:

Quote

We're at a point here where companies are trying to claim that things that are equivalent to steering wheels in cars are their idea that no one else can use. And they know this, which is why we often don't see companies asking big companies to stop doing the thing, just to pay them unheard of amounts of money to be left alone.

And I'm not alone in this belief. Basically anyone who believes in variants of the GPL license is saying the same thing I am. I think the GPL license goes a little too far, but whatever. My views are more in line with Donald Knuth's, who also believes that while closed source software is acceptable, it's dangerous to grant patents for software.

In case you don't know who Knuth is, he is the man that popularized Big O notation, invented Tex, and invented LR parsers (of which nearly all modern compilers are based on). Imagine if Knuth had patented and enforced those patents on his parser: the face of programming today would surely look very different, if it would exist at all.

[peanuts104]

I don't trust the US Supreme Court to rule on this in a manner that will end well.  I doubt any of the justices serving even know what coding is or could possibly understand the ubiquity and ease of access to all this information.

[Commodus]

3 hours ago, straight_stewie said:

The problem with software, as opposed to natural language, is that there is very often only a few ways to do something, especially as you move out of user space, or as you move to more general definitions of "doing things".

The problem that we are seeing is that companies like Microsoft have *successfully sued companies essentially for not using Windows. The most famous one is Microsoft v. TomTom. Where Microsoft alleged that TomTom had violated over 8 of Microsofts patents, essentially by using Linux in it's GPS systems.

As a result, TomTom *successfully counter sued Microsoft for Microsofts attempt to create a device that "generates a maneuver for a turn at an intersection", "route generation in a vehicle navigation system", and "Quick selection of destinations in an automobile navigation system".

After having success with Microsoft, TomTom sued Garmin, because Garmin used "an algorithm designed to decide which streets are relevant to a users route". (or in laymans terms, because Garmin used a route finding algorithm to, well, find routes).

Seriously. We came very near to it being illegal for anyone other than TomTom to make GPS equipment, on the basis that ideas like route finding or file selection are wholly theirs.

 

But this takes me back to what I said before:

And I'm not alone in this belief. Basically anyone who believes in variants of the GPL license is saying the same thing I am. I think the GPL license goes a little too far, but whatever. My views are more in line with Donald Knuth's, who also believes that while closed source software is acceptable, it's dangerous to grant patents for software.

In case you don't know who Knuth is, he is the man that popularized Big O notation, invented Tex, and invented LR parsers (of which nearly all modern compilers are based on). Imagine if Knuth had patented and enforced those patents on his parser: the face of programming today would surely look very different, if it would exist at all.

Those are very fair points.  I'm mainly responding to the notion that sharing anything somehow invalidates your ability to chase after abuse.  If TomTom had shared its know-how to Garmin but found later that Garmin had violated the terms for receiving that material, then TomTom would've been in the right.

[straight_stewie]

52 minutes ago, Commodus said:

I'm mainly responding to the notion that sharing anything somehow invalidates your ability to chase after abuse.  If TomTom had shared its know-how to Garmin but found later that Garmin had violated the terms for receiving that material, then TomTom would've been in the right.

My general stance is that if someone isn't literally copying protected code, they should be in the clear.

Or in other words, it should not be illegal for me to make a program that does X, because a bigger company also makes a program that does X. But it should be illegal for me to make copy pasta from the bigger companies code base.

This makes intellectual property protection possible through copyright, but not through patent. Which is an important distinction to make, since a patent protects an idea, while a copyright protects an expression of an idea.

[Kroon]

On 1/27/2019 at 5:20 AM, RorzNZ said:

I think creating a mass operating system on millions of phones worldwide is a little out of fair use. For instance maybe something like fair use on the YouTube. You can have a sample of the video in your video and make a little money, now if that whole video was, say, a reaction video and the sample video was a main proponent, the original author might get a little concerned if your video had nearly a billion views, making quite a lot of money.

 

Your do not reflect the use case.  They are providing the API.  You can download same API for Windows for free, so should Microsoft also pay?  Only difference is that it's included in Androis so the user don't have to download it.  

 

If this ruling stands US users of Android would probably be forced to download the API as they activate the phone.  In EU the court have already ruled against Oracle so phones sold there would not be affected.

[Guest]

2 minutes ago, Kroon said:

 

Your do not reflect the use case.  They are providing the API.  You can download same API for Windows for free, so should Microsoft also pay?  Only difference is that it's included in Androis so the user don't have to download it.  

 

If this ruling stands US users of Android would probably be forced to download the API as they activate the phone.  In EU the court have already ruled against Oracle so phones sold there would not be affected.

Android is built on part of the API, its not so much product integration, but rather a component. 

[Kroon]

4 minutes ago, RorzNZ said:

Android is built on part of the API, its not so much product integration, but rather a component. 

 

Yes they have an SDK and some third party Android builds (Like Samsiung) have core applications using JAVA. However the Andriod core can be compiled and used without JAVA.  Some TV-Boxes that come with Android installed don't have JAVA support, you need to download JAVA yourself if you want to run and third party apps. So it would be really easy for Google just to make the JAVA part a download, just as Windows, and they would void this ruling. Rulings like this only make things harder for the end user.

[Guest]

3 hours ago, Kroon said:

 

Yes they have an SDK and some third party Android builds (Like Samsiung) have core applications using JAVA. However the Andriod core can be compiled and used without JAVA.  Some TV-Boxes that come with Android installed don't have JAVA support, you need to download JAVA yourself if you want to run and third party apps. So it would be really easy for Google just to make the JAVA part a download, just as Windows, and they would void this ruling. Rulings like this only make things harder for the end user.

Seeing as it’s really easy they probably should have done that then, just cost themselves money in lawyers - probably an oversight!

[LAwLz]

18 hours ago, Commodus said:

While I'm not saying Oracle is in the clear here, this is a lousy way of thinking about it.  Unless it's unreservedly open source, you have a legal right to enforce your terms.

 

Think of it this way: imagine if you wrote a book and shared it with everyone so long as they didn't alter it or pass it off as their own.  Now imagine that you find out someone has not only falsely taken credit for it, but inserted an erotic fanfic chapter.  Do you just shrug it off and accept that your book is no longer under your control, or take action against that person for violating the terms they agreed to when they received your book?  You'd probably opt for the latter. 

 

I'm not arguing against freely sharing things without imposing terms, but that's up to the creator.

That is not what is happening though. Google does not use any of the code written by Oracle/Sun in Android.

I am all for Oracle being able to enforce their terms of service on their own code, but what they are trying to do here in enforce their terms of Google's completely rewritten libraries because they have the same name.

 

Like I said before, it's like if a Youtuber started copyright striking other channels because they had videos with the same titles on them, despite the content being different.

 

 

7 hours ago, Kroon said:

Your do not reflect the use case.  They are providing the API.  You can download same API for Windows for free, so should Microsoft also pay?  Only difference is that it's included in Androis so the user don't have to download it.  

 

If this ruling stands US users of Android would probably be forced to download the API as they activate the phone.  In EU the court have already ruled against Oracle so phones sold there would not be affected.

4 hours ago, RorzNZ said:

Seeing as it’s really easy they probably should have done that then, just cost themselves money in lawyers - probably an oversight!

That is not quite what is going on there. The thing is that Google does not use any code written by Oracle/Sun, and that's why Oracle is pissed off.

Android does not come with Java. It comes with a library written y Google, which functions the same as the standard Java libraries.

[LAwLz]

23 hours ago, LAwLz said:

For the future of all software development, I sure hope it is Oracle that is in the wrong, but who is "wrong" just depends on who the court sides with. 

 

It's obviously a bit more complicated, especially since Java has different versions with different licenses etc but that's the gist of it. 

To elaborate a bit more on this, and to give some sympathy for Oracle, I guess I should explain their side as well.

First thing first, Sun was the company that developed Java. Sun was later bought by Oracle. So for all intents and purposes, this lawsuit is Oracle/Sun vs Google.

 

When Android was first being developed, Eric Schmidt from Google were in talks with Jonathan Schwartz (Sun's president) about licensing the Java libraries for use in Android. Sun wanted something like 50 million dollars for the license, which Google were willing to pay. However, Sun also wanted to share control over Android. Sun's argument was that they needed to control in order to make sure that Android's APIs were consistent with their own. The entire philosophy of Java is that one Java program written on one platform will work on all other platforms with Java support. Sun feared (rightfully so) that handing full control over Android's APIs over to Google would result in Java programs which only functioned on Android.

 

Since Google did not want to share control over Android with Sun, and Sun did not want Google to have full control over Java APIs on Android, they did not reach an agreement.

As a result, Google went "alright, if you won't let us use your libraries we'll write our own that looks and functions the same as yours!", and so they did.

 

So in summary, the conversation went:

Google: "Hey, we wanna use Java in our new operating system".

Sun: "Sure! But we want to share control over the operating system with you".

Google: "Sorry but you're not getting any control over Android"

Sun: "Then you're not getting Java".

Google: "Fine, then we'll just make a copy of it that we control".

 

 

It is important to stress that Google has no idea of how the code for the various APIs looks or functions in the back-end. All they know is that in Java, if you write for example

Quote

"FindShortestPath(Statue of Liberty, Empire State Building);

it returns a description of how to drive to get there (just a fictional example, that is not a real Java function).

How the program actually figures out the shortest path is completely unknown. They just know that FindShortestPath, and then followed by the name of two destinations will give you the shortest path back.

 

 

The trial is about if you can own the name of APIs and their structure (like "FindShortestPath" followed by the name of two destinations).

Copyright can be a bit tricky because you can not copyright a functional part of something. What that means is kind of up in the air. There was a trial before where someone tried to copyright the look of a cheerleader outfit.

The only suing said that the function of a cheerleader outfit was to clothed the person and absorb sweat, therefore the lookw as copyright-able.

Others argued that the function of a cheerleader outfit was to identify that person as a cheerleader, and as a result the look was functional (it has to look that way to fulfill its functional purpose).

Not sure how that trial ended.

 

The arguments being made right now is that the name and structure of APIs are functional, and therefore not eligible for copyright protection.

[RejZoR]

If Oracle owns the rights to certain things, they are entitled to be appropriately compensated for it if used by competitors. You can't just say "ups, we didn't know, but now it's in billions of devices and oh well, lets call supreme court to overrule it". If this goes through, everyone will play dumb, implement something and then say "hey, Google did it too". Copyright troll or not, that's not how things should be done.

[Kroon]

I have been reading the documents from Federal Circuit Appeals Court and I'm quite surprised.  When Android their Dalvik VM to run JAVA the arguments would actually make sense.  But Dalvik where removed  about 10 years ago, think it was with KitKat.

 

1 hour ago, LAwLz said:

That is not quite what is going on there. The thing is that Google does not use any code written by Oracle/Sun, and that's why Oracle is pissed off.

Android does not come with Java. It comes with a library written y Google, which functions the same as the standard Java libraries.

 

As you say there is not a single JAVA code in the Core Android, apps that are included with some phones are developed in JAVA but that is not Google fault.  Some light versions av Andriod even comes without the the ART. Android "USB sticks" often uses this version where all JIT functionality are removed. Some devices simple don't have the computing power but most of the time you can install the ART yourself.

 

If you download Android NDK (Native Development Kit) you will see that whole base of the operating system are in C and C++.  Top of that you will find applications made in Kotlin.  Koltin are now days the default language for Andoid  apps, I wounder if that can be the problem with Oracle.

 

I need to read more of the court documents to be sure but it feels like that Google are blamed for some changes that other companies do to make their own Android.

[LAwLz]

24 minutes ago, RejZoR said:

If Oracle owns the rights to certain things, they are entitled to be appropriately compensated for it if used by competitors. You can't just say "ups, we didn't know, but now it's in billions of devices and oh well, lets call supreme court to overrule it". If this goes through, everyone will play dumb, implement something and then say "hey, Google did it too". Copyright troll or not, that's not how things should be done. 

Well the debate is whether or not they actually own the right to certain things they claim they own.

And as for the whole "hey, Google did it too" issue, the thing is that everyone is already doing it and have been doing for decades. If Oracle wins this, they could potentially go after a ton of others like Microsoft. But at the same time, the people behind things like C could go after Oracle. It would become a wild western where everyone can sue everyone.

Everyone has kind of assumed API names and structured weren't protected by copyright, so everyone has been copying each other and as a result of Google pissing in Sun's coffee it is now being tested in a court, for real.

 

 

2 minutes ago, Kroon said:

I have been reading the documents from Federal Circuit Appeals Court and I'm quite surprised.  When Android their Dalvik VM to run JAVA the arguments would actually make sense.  But Dalvik where removed  about 10 years ago, think it was with KitKat. 

Not even Dalvik ran Java on Oracle developed Java.

In fact, Android does not run Java at all. It runs a completely rewritten set of APIs which is compatible with Java. This goes both for Dalvik as well as ART. That's why you often see people say Android runs on a "Java dialect", "fork of Java" or "Java-like".

[DarkSwordsman]

On 1/28/2019 at 5:15 AM, LAwLz said:

For the future of all software development, I sure hope it is Oracle that is in the wrong, but who is "wrong" just depends on who the court sides with.

 

Here is my attempt to explain the lawsuit in simple terms.

...

Actually, the way I understand it, Android does supplement the actual Java libraries on Android. The APIs they integrated are the original APIs. Modifying the APIs is a deeper part of the license that they didn't seem to violate.

 

What Oracle is suing for is the fact that those APIs can be accessed on Android. In their licensing it's a little vague on if that is inherently violating their licenses.

 

Also, to clarify, printf isn't an API itself but a method of a bunch of larger classes/modules in an API. For example, Java.lang is an API to use the basic Java language. Java.io is an API of methods used to interact with hardware devices connected to the operating system.

 

To me, Oracle shouldn't be suing Google, but they should be suing all the device manufacturers that implemented Android.

 

As mentioned before, Google's implementation of Java doesn't necessarily violate their license agreement. They provide the same features and functionality that anyone could download or implement into Android themselves. Also, Android being an open source software means that Google isn't selling it. It's a software that anyone can access and use.

[DarkSwordsman]

On 1/29/2019 at 5:01 AM, RejZoR said:

If Oracle owns the rights to certain things, they are entitled to be appropriately compensated for it if used by competitors. You can't just say "ups, we didn't know, but now it's in billions of devices and oh well, lets call supreme court to overrule it". If this goes through, everyone will play dumb, implement something and then say "hey, Google did it too". Copyright troll or not, that's not how things should be done.

Look at it this way. It's very clear in their licensing that Java can be used internally to a company to develop software. It's also clear that the software they make can be used internally, as well as given out for free to other companies/people, or even sold if it is given out as a binary (exe of sorts). 

 

Think about it this way. What is Minecraft made on? Java. Minecraft costs money. What is Android made on? C, but it has Java distributed with it (core libraries from Oracle) for applications to be supported on it. Android is free.

 

A little bit contradicting if you ask me.