CEO Larry Page defends Google on the stand: “Declaring code is not code”

Jamie4443[/url]":22q4iyd3]I'm no programmer, so what are these "declarations" ?

It kinda sounds like Google used these and then modified the detailed code they're comprised of, so as to avoid having to get a license from Oracle.

But again I don't know what they are and I'm just trying to figure it out from the article.

This is over-simplified, but imagine you're taking instruction from someone to bake a cake. First, you agree that the instruction will be given in English, or Spanish, or over email, or via text, and that you'll use certain measurement systems (cups, tablespoons, milliliters, degrees Fahrenheit, etc). This is roughly what is meant by terms like "the Java language", "interfaces", "declarations", and these things were copied. They didn't reinvent their own tablespoon, they use the same definition. The process of actually making a cake, that's the implementation, and that code is kept separate and was not copied.

So you can imagine, Sun came up with their own measurements and way of communicating a recipe and started a commercial bakery where people can bring in recipes in their special format and they'll be made. Google started their own bakery, and accepts Sun's recipe format(or a subset of it), because there are already a lot of people who know how to write recipes in that format. What goes on in the kitchen to actually make the goods is totally different, but the two bakeries are compatible because they both accept the same instructions. Now Oracle buys Sun, and is angry that Google's bakery is dominating and accepts recipes in Sun's format. They think they deserve a cut for coming up with the format.

 

[WaveRunner]

fenris_uy[/url]":1eard5h4]

d4Njv[/url]":1eard5h4]

fenris_uy[/url]":1eard5h4]
As the appeals court ruled that the reimplementation of the APIs was (They ruled that APIs are covered, that would make Dalvik a derivative work of the Java APIs, since OpenJDK is covered by the GPL, that would make Dalvik a derivative work of code covered by the GPL).

The question of whether a codebase is a "derivative work" is separate from the issue of the interfaces are actually covered by the GPL. Consider for example the fact that some versions of the C standard library are licensed under GPL while others are under other licenses. All publish the same interfaces. If you write your own implementation of those interfaces, how can the author of the GPLed library argue your implementation to be a "derivative work"?

Isn't C an ISO standard?, and as such they have their API released under the licenses that ISO imposed unto them. If those licenses allow a GPL implementation, that doesn't means that other implementations are derivative work of that implementation, because the original API has another license that allows implementations of the C API. With the Java API you are either, a derivative work of the OpenJDK, or have a license from Sun/Oracle to create your own implemetation. Google is covered by neither. IBM has a license to create their own version.

I guess you could lift them... but that raises the question in that one would only implement the C library if one were to put it under a new license or they thought they could make a better one. Either way no one would start from someone else's work. It's hard enough to read ISO documents without using those headers as trusted landmarks.

 

[RndNum123]

o1knives[/url]":afrq6ixr]Unfortunately, I am expecting Oracle to win. Their argument is basically, we own the rights to the API and Google didn't pay us for it and they should have. Google's argument seems to be, yes we knew the API was copyrighted but we copied it anyway because we thought it was OK to do so. From an uninformed jury perspective (and not realizing what fallout of the precedents this case will set will be) Oracle's argument seems the best.

Do you know if it would be possible for the software industry to fight this "stupid legal definition of APIs" that will ruin the software industry by starting to include an Anti-Oracle clause into most open source projects, that simple states : This work is licensed under MIT license, except for Oracle and all affiliated corporations, they are not allowed to use this software in any way.

 

[RobDickinson]

Jamie4443[/url]":1ln4iy4m]I'm no programmer, so what are these "declarations" ?

Programs make use of a toolbox of functions created for them, they do through an application programming interface (api).

All this is is a document that tells everyone what tools ( methods) exist, what they need to be able work (size, time etc) , and how you finds them (which toolbox).

So rather than developing your own tools to do basic things ( and hard things like directly working with the pc hardware) you rely on these.

An analogy here is a jigsaw puzzle.

Sun made a puzzle, with a picture of a cottage on the front, out of cardboard. They gave this away free.

Because Sun/Oracle played hard with licensing this to Google, google went off and made an entirely different jigsaw puzzle with a picture of a robot on it, out of plywood.

But the shape of the space in between ( tessellations?) is the same.

Oracle is arguing Google owe them money for this.

 

[Marcellus111]

jukes_[/url]":1l2tzbyc]Given the appeals court decision I don't see how google gets out of this without going higher about copyright and APIs. How is the jury supposed to find anything other than that the APIs are copied? At this point it seems like they are doing damage mitigation, as they must.
[Edit: added link]

This trial has two parts. One is to determine whether Google infringed Oracle's copyright. Then, if it is found they infringed, they will determine damages in the second part. The previous case at the Federal Circuit ended concluding that APIs are copyrightable. There is now not a dispute that Google copied the APIs or that APIs are copyrightable, but the current arguments are whether Oracle held a copyright on those copied APIs. If Google can convince the jury that there was no copyright, the case is over. If not, the second phase of damages begins, but Google can still argue at that point that, despite having infringed a copyright, the use was fair use and Oracle is not entitled to damages.

 

[tricorn]

fenris_uy[/url]":11nz9vxw]
They are unlicensed from the owner of the original copyright (Sun, then Oracle).

The Harmony project was created under a specification license from Sun. The terms of that license required passing the TCK in order to use the Java name and logo and patents. They never reached an agreement over the terms of getting access to the TCK, so they never passed it.

It was understood by all parties to the license that the copyrights Sun had did not apply to the API itself.

Similarly, when Sun applied the GPLv2 with Classpath exception (so called because Classpath was the name the FSF called their reimplementation of the Java platform, including the API) they also knew how the FSF interpreted the language of the license.

The intent of the licensing party should matter, regardless. What did they think was covered by the specification license or GPLv2+CE?

Declaring code is "code", it is just not "executable code". It makes no difference to copyright or fair use. Declarations don't "do anything" without the implementation, implementation does nothing without the declarations. The reason copying part or all of an API should have been found to be not infringing has nothing to do with whether they were declarations or executable code.

There's not even a good distinction as to what exactly "executable code" is. Code is what's written by a coder. Even some of the javadoc comment text is code, directions to the javadoc program that processes it.

Copyright law defines a "computer program" as "a set of statements or instructions to be used directly or indirectly in a computer in order to bring about a certain result." which is a pretty good working defintion of what "code" is.

 

[stickboy]

michaelzehr[/url]":3b4cuffh]
Is the court really saying that the first person who writes "double distance(Point A, Point B)" in some language is the only person that can ever create such a function? (method/macro/whatever)

A few people have answered "yes", and while IANAL, I'm going to instead claim that the answer is "not quite".

That's not how copyright works. The purpose of copyright is to specify rights regarding copying. If two entities independently create the same expression (whether it be code, a novel, artwork, whatever), then each of them owns copyright for their own work. (This is unlike the trademark and patent systems.)

If someone else sees that work and replicates it without permission from the copyright owner, then that's copyright infringement.

So regardless of this case, copyright on a "double distance(Point A, Point B)" function declaration is not based on a first-come-first-serve basis. (Furthermore, it's generally accepted and uncontroversial that code implementation is copyrighted, yet nobody will claim that only the first person to implement such a distance function can claim copyright.)

Finally, even though a previous court ruled that APIs are copyrightable (which I don't personally agree with), replicating APIs probably still would be considered fair use most of the time. The fair use claim in this particular case is murkier because Google didn't strictly need to use those APIs for interoperability. Even if Google loses this case, it's not necessarily the end of the world.

 

[WaveRunner]

sprockkets[/url]":1krad226]

WaveRunner[/url]":1krad226]

sprockkets[/url]":1krad226]

WaveRunner[/url]":1krad226]

Some Idiot[/url]":1krad226]
Here's the issue at hand, however: Oracle absolutely FUBARed their implementation of Java on phones. Google...did not. For all of Android's kinks, it largely 'just works'.

If Java was so damn FUBARd/broken how come Android continues to steal from Java APIs that are very recent additions done by Oracle...

Are you really that thick to think that you can steal something GPL licensed if you are referring to openjdk 8?

Still trolling I see...

Bull shit. You are trolling by saying that, instead of responding to your totally false premise.

You also typically get copyright stuff totally wrong as well, so goodbye to your shitty posts as well.

Only a moron would think a statement that Oracle is still innovating on Java and Google is the one playing catchup, would be read as a statement about licensing.

That is why you are a troll. You read one sentence, disagreed with it and went on a rant that is totally irrelevant simply because you did not follow the thread of discussion or care what people actually have to say. Next time stop and think for a second better yet, stop and read.

 

[Cilvre]

RobDickinson[/url]":6ey1y21a]

Jamie4443[/url]":6ey1y21a]I'm no programmer, so what are these "declarations" ?

Programs make use of a toolbox of functions created for them, they do through an application programming interface (api).

All this is is a document that tells everyone what tools ( methods) exist, what they need to be able work (size, time etc) , and how you finds them (which toolbox).

So rather than developing your own tools to do basic things ( and hard things like directly working with the pc hardware) you rely on these.

An analogy here is a jigsaw puzzle.

Sun made a puzzle, with a picture of a cottage on the front, out of cardboard. They gave this away free.

Because Sun/Oracle played hard with licensing this to Google, google went off and made an entirely different jigsaw puzzle with a picture of a robot on it, out of plywood.

But the shape of the space in between ( tessellations?) is the same.

Sun is arguing Oracle owe them money for this.

should edit that last line, "Oracle is arguing Google owes them money for this."

 

[Chemical Tribe]

WaveRunner[/url]":8o0dusij]

sprockkets[/url]":8o0dusij]

WaveRunner[/url]":8o0dusij]

sprockkets[/url]":8o0dusij]

WaveRunner[/url]":8o0dusij]

Some Idiot[/url]":8o0dusij]
Here's the issue at hand, however: Oracle absolutely FUBARed their implementation of Java on phones. Google...did not. For all of Android's kinks, it largely 'just works'.

If Java was so damn FUBARd/broken how come Android continues to steal from Java APIs that are very recent additions done by Oracle...

Are you really that thick to think that you can steal something GPL licensed if you are referring to openjdk 8?

Still trolling I see...

Bull shit. You are trolling by saying that, instead of responding to your totally false premise.

You also typically get copyright stuff totally wrong as well, so goodbye to your shitty posts as well.

Only a moron would think a statement that Oracle is still innovating on Java and Google is the one playing catchup, would be read as a statement about licensing.

That is why you are a troll. You read one sentence, disagreed with it and went on a rant that is totally irrelevant simply because you did not follow the thread of discussion or care what people actually have to say. Next time stop and think for a second better yet, stop and read.

Your post was one fucking sentence! You can't even get that right and complain about me trolling?

Excuse me while Iaugh at what you did!

edit: Let's break this down since you rather bitch about trolling like you are deanrozz:

"If Java was so damn FUBARd/broken how come Android continues to steal from Java APIs that are very recent additions done by Oracle..."

First JavaME is crap, and no body "stole" from it, not Apache or Google.
Next, Google used apache harmony, yet Sun nor Oracle cared about it. That's ASL licensed for anyone to use.
And now with android N, Google switches from Haromony to the latest version, you know, *Oracle* openjdk 8, your part about "recent additions" that one would logically assume you are referring to.

I've done way more to try to understand you than you have, yet I'm the troll. GFY if that's how you want to be

 

[MKD]

ethd[/url]":3qwsfvjq]
Recipes aren't eligible for copyright, so if they did give me their recipe,

Well, until the CAFC brain farts again and declares that they are eligible for copyright protection.

 

[Nychold]

Jamie4443[/url]":3jh6t2si]I'm no programmer, so what are these "declarations" ?

It kinda sounds like Google used these and then modified the detailed code they're comprised of, so as to avoid having to get a license from Oracle.

But again I don't know what they are and I'm just trying to figure it out from the article.

Code declarations are like a job description. The code itself is like the employee who fulfills the job description.

 

[Joao Sousa]

Nychold[/url]":1sghajo3]

Jamie4443[/url]":1sghajo3]I'm no programmer, so what are these "declarations" ?

It kinda sounds like Google used these and then modified the detailed code they're comprised of, so as to avoid having to get a license from Oracle.

But again I don't know what they are and I'm just trying to figure it out from the article.

Code declarations are like a job description. The code itself is like the employee who fulfills the job description.

Very simplistic. It's much harder for programmers to write a good API then to write implementation code.

As a manager I don't bother looking at the implementation code, but I enforce strict rules on the API because it is the API that defines how all the code is structured, how classes talk to each other. You can rewrite implementation code easily, but a poorly structured program is extremely hard to fix.

 

[Abhi Beckert]

Nychold[/url]":1h7waudy]

Jamie4443[/url]":1h7waudy]I'm no programmer, so what are these "declarations" ?

It kinda sounds like Google used these and then modified the detailed code they're comprised of, so as to avoid having to get a license from Oracle.

But again I don't know what they are and I'm just trying to figure it out from the article.

Code declarations are like a job description. The code itself is like the employee who fulfills the job description.

And an API like Java is a list of job descriptions so complicated that it takes years of careful work by experts to create.

Since it's so much work, and since it is "creative" (there are many ways to name a function - for example strlen(string) or string.length() or string.byteCount() or string.characterCount()?), the courts have decided that APIs are eligible for copyright.

Wether or not an API is "code" is irrelevant. What matters is wether or not copyright applies, and the supreme court has been very clear on this: copyright does apply.

 

[Chemical Tribe]

Abhi Beckert[/url]":y4wzkzrn]

Nychold[/url]":y4wzkzrn]

Jamie4443[/url]":y4wzkzrn]I'm no programmer, so what are these "declarations" ?

It kinda sounds like Google used these and then modified the detailed code they're comprised of, so as to avoid having to get a license from Oracle.

But again I don't know what they are and I'm just trying to figure it out from the article.

Code declarations are like a job description. The code itself is like the employee who fulfills the job description.

And an API like Java is a list of job descriptions so complicated that it takes years of careful work by experts to create.

Since it's so much work to create, APIs are eligible for copyright.

Wether or not an API is "code" is irrelevant. What matters is wether or not copyright applies, and the supreme court has been very clear on this: copyright does apply.

1. Whether or not something is hard to implement isn't a test about whether or not something deserves copyright. An internal combustion engine is hard to make yet is NOT covered by copyright.

2. The supreme court made no such ruling. Current copyright law does NOT cover functions which most definitely an API is. The fact that the CAFC got it wrong while everyone else didn't is proof of that.

 

[NullTroll]

Abhi Beckert[/url]":1s3icn7j]

Nychold[/url]":1s3icn7j]

Jamie4443[/url]":1s3icn7j]I'm no programmer, so what are these "declarations" ?

It kinda sounds like Google used these and then modified the detailed code they're comprised of, so as to avoid having to get a license from Oracle.

But again I don't know what they are and I'm just trying to figure it out from the article.

Code declarations are like a job description. The code itself is like the employee who fulfills the job description.

And an API like Java is a list of job descriptions so complicated that it takes years of careful work by experts to create.

Since it's so much work to create, APIs are eligible for copyright.

Wether or not an API is "code" is irrelevant. What matters is wether or not copyright applies, and the supreme court has been very clear on this: copyright does apply.

The Supreme court has not in fact made this clear. They have been notably silent on this issue, and declined to hear Google's appeal of the CAFC, which decided the copyright issue outside of it's jurisdiction and disregarded 9th circuit precedent when issuing the order.

 

[tricorn]

Creating an API, given a set of functions you want to be able to do, is actually not that difficult. The implementation code was found to be not infringing, so obviously simply using a set of functions is not protected.

There are two things you need to do - decide which functions and state (values of variables) belong together and then what to name them.

Naming things is pretty easy. Tell me what a class represents, then give it a name. Tell me what a variable represents, then give that a name. Tell me what a method does, and the various variables I might want to use to affect that behavior, and I can name it and give it a signature (return value, types of arguments, errors that it can generate).

Getting the organization right is more difficult, but is also not very difficult, at least not for the types of things that Google actually used from the Java API. That's where the API is actually designed. A poor design will be more difficult to use, but for base classes like this a lot of it is pretty obvious, especially when it's all very similar to many other language APIs.

Declarations are checked by the compiler. Errors are not complex, and don't depend on complex runtime interactions. Implementation code is more difficult to debug, there are all sorts of special cases and interactions. Doing the implementation usually uncovers design errors, which gets refined as you do more and more of the implementation.

The only reason to use an existing API, but not copy the implementation, is to be compatible with existing code and programmer knowledge. Even more obvious that it was done for compatibility is that they rejected the parts they thought weren't important and WROTE THEIR OWN FROM SCRATCH.

The package hierarchy is even more obviously copied only to be compatible with existing code. It isn't necessary. They could have dumped all those classes into one package "system", made all references to packages under "java" and "javax" be equivalent to "system", and everything would work.

The CAFC decided that not only the names but the design were expression, not a functional idea. One of the hallmarks of an idea is that it can be transformed into many different forms that still express the exact same idea, such as documentation of the entire Java API or the compiled class files. The documentation files (generated in part from the same declarations that were found to be infringing) were found by a jury to be not infringing.

The GPLv2+CE explicitly allows the compiled classes to be distributed under any license you want, when combined with one or more independent modules. Since they also contain a complete copy of the Java API, that alone should mean that the API has also been licensed that way, e.g. under the Apache license.

 

[Joao Sousa]

tricorn[/url]":1bhmfjhq]Creating an API, given a set of functions you want to be able to do, is actually not that difficult.

Right.....

Anyway, it is irrelevant since that issue has been decided. This case is now only about fair use.

 

[tricorn]

Whether designing an API and naming the elements is difficult or not has bearing on fair use and damages, and has nothing to do with whether they are protected by copyright (the only real issue already decided).

Code that defines an API doesn't need to be very complex in order to be protected by copyright. For example, the GNU readline functions.

The API isn't very big, even though the implementation details are quite complex. However, you can create a very simple implementation that conforms to the API, yet very little other than return a line of text without allowing it to be edited, keep a history, or anything complex.

 

[Joao Sousa]

tricorn[/url]":2mhkw73x]Whether designing an API and naming the elements is difficult or not has bearing on fair use

I forgot you were such a legal "expert".

 

[raptormissle]

DarthSlack[/url]":zc5792q8]

Why do you keep saying this? Dalvik was based on Harmony, not Java, and Harmony uses the Apache
license.

Are you implying that Dalvik was based on the Harmony VM? Because if you are it's incorrect. Dalvik is a register based VM while the Harmony and Java VM's are stack based.

 

[Zlotnick]

davnel[/url]":63ue8rw3]Back in the early '80s, when the PC had just been introduced by IBM, several companies wanted to reproduce IBM's work and sell their own version of the PC. Problem was that IBM had published the source for their BIOS code, and copyrighted it to death.

Compaq's answer to that was to contract out a "clean room" reproduction of the calls and functions of the BIOS by Phoenix Technologies, but by people who had never seen the original source. They did this by having one team read the BIOS code and produce a specification for it, then gave that to another team that had never seen the IBM BIOS code who cold-wrote software to implement that specification.

When all was said and done, Phoenix had a viable, unique BIOS that did not violate IBM's copyrights and that passed several court tests. That specification we would call an API today. What's the difference? I don't see any way to copyright a specification or an API, or the contents of a phone book. BAH!!! IMHO more money-grubbing by Oracle, as usual! Someone should spend 57 cents, buy the company and squash it. SCO, too.

The difference is that Phoenix produced the "API", not IBM.

The Java API was copyrighted either implicitly or explicitly by Sun, then the copyright was acquired by Oracle. It's not mentioned in the article, but the API copyright infringement would happen if Alphabet (nee Google) published substantial portions of the Java API as written by Sun. Then, Oracle could receive damages the value of which depend on whether they filed for copyright or just put a (c) at the bottom of the page.

That's not the meat of the article, however. We have a lawyer for Oracle trying to make the case that using the API to create an implementation is a copyright violation, which it plainly isn't. A quick look at Beyond Compare will tell you if the source is the same, or even if a few minor changes were made to hide the origin. I'm sure Google had the resources to write their own implementation of Java and was risk-averse enough to not blatantly steal source code, so I don't like Oracle's chances with this line of reasoning.

Proving actual damages on the API copyright violation will be nearly impossible since Sun didn't sell the API itself except maybe as an appendix in a Java programming book. They gave it away on their web site that wasn't supported by advertising, so the API copyright damages claim is very thin.

The first commenter has it right -- Sun gave away their muffin recipe and is now upset that other people are making muffins with it. If Google published the recipe almost exactly, then there will be some damages awarded.

 

[tricorn]

So apparently you think that the difficulty of designing an API has no bearing on fair use or damages? You don't think that Oracle will try to claim that their APIs were so critical and Google had no choice but to copy them because it was so hard to do that Oracle deserves 100% of the credit for Android's success?

That it wasn't compatibility but simply need, that Google couldn't have created their own API, so it wasn't fair use?

 

[tricorn]

Zlotnick[/url]":2bgk0fnx]
The difference is that Phoenix produced the "API", not IBM.

The Java API was copyrighted either implicitly or explicitly by Sun, then the copyright was acquired by Oracle.

The API that Phoenix used is the same API that IBM created. The API is the functional specication that's passed from the dirty side to the clean side when doing clean room reverse engineering and reimplementation.

According to the CAFC, the Java API was copyrighted expression, according to most precedent, the Java API is a functional idea and only the source code was copyrighted.

 

[fenris_uy]

sprockkets[/url]":1froybko]

Dude you've been shown 20 times how wrong you are about this, yet continue spewing bull shit, so at this point you are confirmed to be trolling and should be ignored.

Did they complian with the GPL? Because it's weird that their lawyers aren't claiming that. Also it's weird because they are moving to openjdk N to avoid the claims that Oracle is presenting now.

 

[mpontes]

Maybe I'm being naive here, but how is it possible for this case to be going in this direction? It seems to me that all Google's lawyers have to do here is to point at this product called "Oracle Linux" that Oracle happens to happily sell subscriptions for, go through the hundreds of software packages containing reimplementations of other companies' copyrighted APIs and ask Oracle how much they paid in licenses to said companies. By then, Oracle will basically be forced to admit that reimplementations of others' APIs is common industry practice and they won't have a leg to stand on.

How is this not over yet? "Declaring code?" Is Google actively trying to sabotage themselves?

 

[tricorn]

GPLv2+CE does not require source code distribution if you distribute the library combined into an "executable" under the Apache license (as allowed by the Classpath exception).

The only GPL violation, assuming that the implementation of the same API makes it a derivative work, would be distributing source code (if they chose to distribute it) with the Apache license instead of the GPLv2+CE. The Apache license and GPLv2+CE are very similar with regard to what is allowed.

There's also evidence that ASF had a license to develop Harmony, and Sun had a written agreement to actually allow it to be certified and released under the Apache license, which would make Google's use of it under the terms of the Apache license "authorized copying".

 

[elozada]

joemullin[/url]":1kaj7isk]

arkiel[/url]":1kaj7isk]Is the jury allowed to declare Oracle the winner and then give them $1 in damages and no attorney fees?

The jury can't weigh in on attorneys' fees, but the answer to the rest of your question is: yes!

I don't think one dollar is a particularly likely result, but Oracle winning the liability case and then essentially losing its damages case is absolutely a possibility.

Such a result would suggest the jury felt "instructed" into holding Google liable, due to the constraints of copyright law as it was explained to them, but still felt Google basically did the right thing.

Any win for Oracle harms the software industry as a whole though, even if they get just 1$ in damages

 

[ip_what]

tricorn[/url]":7q40m6j6]So apparently you think that the difficulty of designing an API has no bearing on fair use or damages? You don't think that Oracle will try to claim that their APIs were so critical and Google had no choice but to copy them because it was so hard to do that Oracle deserves 100% of the credit for Android's success?

That it wasn't compatibility but simply need, that Google couldn't have created their own API, so it wasn't fair use?

Strictly speaking, difficulty doesn't have any bearing on copyright in the United States. There's a handful of cases dealing with copyrightability of databases in which the US court system has firmly directed "sweat of the brow" as a factor in determining copyrightability. (Europe is different). What matters is creativity, not difficulty.

 

[tricorn]

The issue of copyrightability is not being examined. The reason Google used part of the Java API instead of writing their own is going to be part of the fair use and damages that the jury decides on. How difficult it would have been to write their own similar APIs is part of that calculation, and of how important/significant it was for damages. Should they get a percent of Google's profits based on percent of infringing material? How do you calculate how important a bunch of scattered lines of code were?

If there were alternatives to Google infringing on the APIs, but it's shown that they could have used OpenJDK or written their own and been just as successful, and they didn't because they thought, as most people did, that the API was free to reimplement, that is significant both for fair use and for damages.

 

[ProfessorGuy]

WaveRunner[/url]":galh1es1]And Linux is not UNIX. It is an implementation of the POSIX standards.

I worked with P.J. ("Bill") Plauger on POSIX when he was chair of the POSIX committee. The entire purpose of the project was to EXACTLY DUPLICATE UNIX'S API. So whoever has UNIX now has the same claim to linux as Oracle has on Google's system.

If Oracle wins, a shitstorm is a-comin'.

 

[Cloudgazer]

tricorn[/url]":19i85y9x]

fenris_uy[/url]":19i85y9x]
They are unlicensed from the owner of the original copyright (Sun, then Oracle).

The Harmony project was created under a specification license from Sun. The terms of that license required passing the TCK in order to use the Java name and logo and patents. They never reached an agreement over the terms of getting access to the TCK, so they never passed it.

It was understood by all parties to the license that the copyrights Sun had did not apply to the API itself.

Well, apparently not all parties.

http://meincmagazine.com/wp-content/uploa ... DACTED.pdf

But what I was missing is the fact that the copyright on the API is real and hard to ignore.
Simply by implementing a class with the same signature of another, in >another namespace and simply by looking at available javadocs could be >considered copyright infringement, even if the implementation is clean room.
So, we are, in fact, infringing on the spec lead copyright if we distribute >something that has not passed the TCK and *we know that*.

 

[AlexisR200X]

"Do you know whether or not that declaring code and the structure, sequence, and organization of those 37 API packages—do you know whether that's in each and every one of those 700,000 phones that you told your shareholders are 'lit up every day?'" said Bicks, raising his voice again.

This right here irks me to no end. Its overly dramatic and redundant, "structure, sequence and organization"... Really? Give me a fucking break throwing out extra synonyms doesn't make the issue any bigger. Its merely manipulative and devoid of substance. Honestly I really hate the stupid asshole that decided that even the declaring code is worthy of copyright protection. What is next, math functions?

edit: Thanks to ip_what For his enlightening reply/explanation. Honestly I didn't know it was a customary thing. Legal jargon is not my forte but its roots are quite fascinating.

 

[Cloudgazer]

ProfessorGuy[/url]":uzjf8gxi]

WaveRunner[/url]":uzjf8gxi]And Linux is not UNIX. It is an implementation of the POSIX standards.

I worked with P.J. ("Bill") Plauger on POSIX when he was chair of the POSIX committee. The entire purpose of the project was to EXACTLY DUPLICATE UNIX'S API. So whoever has UNIX now has the same claim to linux as Oracle has on Google's system.

The owners of the UNIX specification are the Open Group.

http://www.opengroup.org/austin/papers/ ... x_faq.html

Good news! They're also heavily involved in the POSIX standardization process, and naturally enough the licensing is all above board.

 

[WaveRunner]

ProfessorGuy[/url]":3kr5ktvq]

WaveRunner[/url]":3kr5ktvq]And Linux is not UNIX. It is an implementation of the POSIX standards.

I worked with P.J. ("Bill") Plauger on POSIX when he was chair of the POSIX committee. The entire purpose of the project was to EXACTLY DUPLICATE UNIX'S API. So whoever has UNIX now has the same claim to linux as Oracle has on Google's system.

If Oracle wins, a shitstorm is a-comin'.

Oh cool.. I guess he will also be sued to high hell and will have to shut down. Maybe you should ask him about that what his plans are. But I do love his product, even though right now I'm suffering with an odd bug from it. though I do blame MS lexer or codegen that's really causing the issue.

 

[Marlor_AU]

elozada[/url]":2rjkxbvb]

joemullin[/url]":2rjkxbvb]

arkiel[/url]":2rjkxbvb]Is the jury allowed to declare Oracle the winner and then give them $1 in damages and no attorney fees?

The jury can't weigh in on attorneys' fees, but the answer to the rest of your question is: yes!

I don't think one dollar is a particularly likely result, but Oracle winning the liability case and then essentially losing its damages case is absolutely a possibility.

Such a result would suggest the jury felt "instructed" into holding Google liable, due to the constraints of copyright law as it was explained to them, but still felt Google basically did the right thing.

Any win for Oracle harms the software industry as a whole though, even if they get just 1$ in damages

A win basically means that interoperability is prohibited. In order to achieve interoperability, you need to reimplement APIs, or file formats, or packet formats.

If any of these things are copyrightable, then interoperability (and therefore competition) goes straight out the window.

 

[berrardo]

This is not Farnsworth versus RCA. This is two FAT FAT companies in a stupid grudge match. The CEOs of both are obnoxious fatcats. What do I care about the outcome? A pox on both their houses.

ArsTechseems to think this is the Scopes Trial, and that we should all get terribly excited about it. Meanwhile some -worthwhile- news is certainly getting pushed to the side.

 

[Marlor_AU]

berrardo[/url]"yif7e9a]This is not Farnsworth versus RCA. This is two FAT FAT companies in a stupid grudge match. The CEOs of both are obnoxious fatcats. What do I care about the outcome? A pox on both their houses.

It's important because if APIs are deemed copyrightable, then this will be used as a precedent next time some established player wants to shut down a startup which is re-implementing their APIs for interoperability.

Google can afford to lose. But the industry as a whole can't afford Google to lose.

 

[mikefox]

Onyx Spartan II[/url]":2fp51d47]

Jamie4443[/url]":2fp51d47]I'm no programmer, so what are these "declarations" ?

It kinda sounds like Google used these and then modified the detailed code they're comprised of, so as to avoid having to get a license from Oracle.

But again I don't know what they are and I'm just trying to figure it out from the article.

To use a crude car analogy: an API is sort of like a document stating that a car will have a steering wheel, 4 wheels, an engine, seats, air conditioning, etc. It's an overarching document defining the function of something (functionality is not normally subject to copyright, IIRC). The implementation is what determines whether the car has a V4 or a V6, seats for four people or two, towing capacity, what the steering wheel and column looks like, what the dashboard looks like, how the air conditioning works, and other details like that. But the overall functionality (i.e., transporting people) remains the same across different implementations.

Would you let me take a slightly different angle on this?

I would suggest that it is a better analogy to describe an API as the idea of how the layout of the controls are set up. So brake throttle and clutch are foot pedals, steering is controlled by a wheel and so on. How the car interprets these commands and uses them to propel the car would be the body of the code. So in this analogy, if Oracle was to win then every car manufacturer would either need to license this basic control layout from the original designer or come up with their own designs, in other words a loss of standardisation.

Honestly I don't know what Oracle think they will get out of this, if they win sure they might get a massive payout from Google, but you know they will then move away from Java altogether and that then means there are less programmers who will be familiar with implementing Java and less likely to use it in other projects. That will possibly lead to a mass exodus from the platform in the long run, not necessarily a bad thing IMO, Java is a mess.

 

[Marlor_AU]

mikefox[/url]":1x1t2lxe]That will possibly lead to a mass exodus from the platform in the long run, not necessarily a bad thing IMO, Java is a mess.

I really don't get the hate for Java.

On the server side, it's brilliant. It has some amazing library support from Apache, and drives the backend of so many large enterprises.

I think many people think Java is all about applets, when it is really all about the server side.

 

[ruddy]

NullTroll[/url]":157pzbd9]

Abhi Beckert[/url]":157pzbd9]

Nychold[/url]":157pzbd9]

Jamie4443[/url]":157pzbd9]I'm no programmer, so what are these "declarations" ?

It kinda sounds like Google used these and then modified the detailed code they're comprised of, so as to avoid having to get a license from Oracle.

But again I don't know what they are and I'm just trying to figure it out from the article.

Code declarations are like a job description. The code itself is like the employee who fulfills the job description.

And an API like Java is a list of job descriptions so complicated that it takes years of careful work by experts to create.

Since it's so much work to create, APIs are eligible for copyright.

Wether or not an API is "code" is irrelevant. What matters is wether or not copyright applies, and the supreme court has been very clear on this: copyright does apply.

The Supreme court has not in fact made this clear. They have been notably silent on this issue, and declined to hear Google's appeal of the CAFC, which decided the copyright issue outside of it's jurisdiction and disregarded 9th circuit precedent when issuing the order.

That's standard Google propaganda, but not what happened. The CAFC followed 9th precedent, and Google didn't like the result.