“Former Sun CEO Jonathan Schwartz took the stand here today as a witness for the defense, and disputed Oracle’s claim that Java APIs were proprietary code from Sun. Google’s lawyer, Robert van Nest, asked Schwartz whether, during his tenure at Sun, Java APIs were considered proprietary or protected by Sun. ‘No,’ Schwartz said in explaining the nature of open software, ‘These are open APIs, and we wanted to bring in more people… We wanted to build the biggest tent and invite as many people as possible.'” Whoopsie for Oracle.
Obviously, if the APIs were truly “open” as Schwartz suggests, then there was no point in negotiating with Google over licensing fees. And yet they did negotiate. And the negotiations failed. And then Schwartz was replaced by adults without ponytails. Sounds like he has an ax to grind.
The negotiations were over the implementation and the TCK (which you need to pass to call it Java). Google wrote their own implementation and called it Dalvik (not Java).
Swartz welcomed Google aboard at the time. This is no turn-around or revisionist history.
Given the fact that Google actually borrowed source files from Sun’s implementation, it’s clear this was no clean-room implementation; in fact, Google was probably referring to the original sources as they wrote their code. Which means that Google’s code is a derivative work and, thus, a copyright violation. But thanks for playing, anyway.
9 source lines of code out of 15 million lines does not a derivative work make.
http://www.groklaw.net/pdf3/OraGoogle-Trial-GoogleOpeningStills.pdf
Google’s code is a derivative work in exactly the same way that BSD was a derivative of Unix. The last bits of AT&T code were removed and BSD was deemed not to infringe on Unix.
This is the same model that Google took with Java. They removed the last bits of Sun (Oracle) code and the result will soon be deemed not to infringe on Java.
I’d suggest a deeper look at those licenses. There was never a license requirement for just using the language or the APIs.
That’s for the court to decide.
thus making the concept of license that more ironic…
Edited 2012-04-27 22:26 UTC
The APIs were as open as, lets say, UNIX.
You were entitled to create Java like implementations of those APIs/VM, but to be able to state that your implementation was Java compatible with the Java trademark logo, you needed to certify your implementation.
Except that’s not what Oracle is arguing. They’re arguing that Google actually copied certain Java sources, and that the Google implementation is a derivative work.
No, they are arguing that implementing API based on published specification is copyright infringement.
Read the slides. They’re also saying that Google out-and-out incorporated the sources.
Hello Pot? This is kettle, you’re black!
Noted. So I should only trust serious looking guys with short conservative haircuts in Armani suits, because the ones with long hair who dress casually are undeveloped unreliable losers.
In Schwartz’s case, yeah. He ran Sun into the ground.
I’m sure that’s all because of his hair…
Suit is going to bite Oracle in the ass hard.
Judge possibly going to GPL their API.
Poetic justice.
Edited 2012-04-27 03:36 UTC
Actually I bet this does absolutely nothing to nobody. More likely oracle will win but it’ll be some amount that for Google won’t even be pocket change, say $30 million, and that will be because too many guys wrote REALLY stupid emails, and then things will go back to business as usual. nothing will change, Android will still be right there next to iOS, same old same old.
If anything this whole suit should be a wakeup call that we really really REALLY need to rethink email, because legally they are given the same weight as a letter and IRL they have about as much thought put into them as your grandma spamming LOLCats. I have seen tons of things in emails where i have thought “Did you even ATTEMPT to think before you hit the send key?” yet as far as a court of law in concerned they carry the same weight as someone sitting down with a sheet of paper bearing the company letterhead and writing a long proposal.
Time and time again we have seen emails bite a company right in the butt so either they need to push the users to abandon email completely or to use some format that isn’t backed up, say an internal instant messenger. because until we fix the laws so that stupid emails aren’t given the same weight as formal letters having email is a BAD idea if there is any chance your company may be sued down the road.
Wasn’t he pointing out to Oracle that they can sue Google for patents, at the time the negotiations took place?
First of all there is no evidence of that. Second of all, patents don’t cover APIs, and concerning APIs, Oracle is asserting copyright violation.
Lesson learned: Don’t use Java.
Browser: Mozilla/5.0 (Linux; U; Android 2.2.2; el-gr; LG-P990 Build/FRG83G) AppleWebKit/533.1 (KHTML, like Gecko) Version/4.0 Mobile Safari/533.1 MMS/LG-Android-MMS-V1.0/1.2
If you are only building apps, then it is no problem using Java.
If you are designing and implementing programming languages based on Java, then you should be careful.
How many does designs and builds programming languages based on Java? Not many. Most companies just use Java to write software, and there is no problems with that.
Correction:
This means no:
– Objective-C
– C#
– Java
– F#
– VB
– Delphi
– …
If you want language freedom, you should use ISO/ANSI languages, or complete open source ones.
I will not give up my Delphi. :p But I am trying to branch out into FreePascal.
I also use Python on occasion and am dabbling with Lua.
C# is both an ECMA (ECMA-334) and ISO (ISO/IEC 23270:2006).
Good try though.
Edited 2012-04-27 14:44 UTC
Really?
Please provide me the EMCA standard number for .NET 3.0 and .NET 4.0 CLR, along with C# 3.0 and C# 4.0.
I can wait, because you see, those standards are not updated since .NET 2.0.
Uh, c# is the language, the .net framework is just an set of libraries. They are related, but not the same thing. The language can be implemented separately from the .net runtime.
Oh, and you can stop waiting.
I responded to the parent who suggested only using standardized programming languages, and gave a list of those he felt were not. c# was on that list, and I explained that it was and ISO and EMCA standard. C# IS A STANDARD, even if the standard has not been updated, and even if you feel the need to nitpick.
Edited 2012-04-27 18:04 UTC
I stand corrected, C# v1.0 and v2.0 are standardized, nothing beyond that.
Legally, you are only allowed to implement up to C# v2.0,
there is no legal protection beyond that version, other than Microsoft stating in some web site that they will play nice with the open source community.
Plus .NET and C# are intertwined, for certain language features, you do require a specific CLR version. It is not as you could target CLR v1.0 with C# v4.0, for example.
So what is the point to use a language, which legally is frozen in time?
Edited 2012-04-27 21:34 UTC
Not really what we were discussing. The syntax and keywords and structure of the language may be at version 2.0, but that doesn’t stop your version of the cli from having whatever extensions you want, if you are implementing it. To be standards compliant, you just must be able to compile and run any arbitary C# 2.0 program, but if the cli exposes more objects, then the language can use them. Just like Mono exposes gtk+ without violating the standard.
Edited 2012-04-27 23:21 UTC
We agree that specific versions of C# require certain versions of the CLR. I am sure we would agree that CLR 2.0 only support up to C# 3.0 (.NET 3.5 era). We also agree that CLR 4 is not completely covered by the ECMA or ISO standards as they have not been updated. So, you could technically argue that sweeping patent grants that Microsoft has offered may not 100% protect Mono users of C# 4.0 features that require CLR 4.
So, let’s look at F#…
1) F# has been released by Microsoft under an Apache 2.0 license
2) Apache 2.0 includes an explicit patent grant
3) Microsoft explicitly released F# for OS X and Linux
4) F# 2.0+ REQUIRES at least CLR 4 to run
5) On OS X and Windows, F# REQUIRES Mono
6) Mono 2.10 and above BUNDLES the Microsoft F# code
Visit the website http://www.tryfsharp.org sometime…
1) tryfsharp.org is registered to Microsoft (it is an official Microsoft site)
2) http://www.tryfsharp.org/Resources/GetMono.aspx is pretty self-explanatory
3) Microsoft RECOMMENDS that you get F# as part of Mono 2.10 or higher (because Mono distributes it)
4) Mono 2.10 and higher DEFAULT to CLR 4
5) The tryfsharp.org site itself uses Moonlight (the Mono implementation of Silverlight)
So, Microsoft USES Mono themselves. They have also released a product (F# for Mac or Linux) that not only RECOMMENDS using the Open Source implementation of CLR 4 (Mono) but in fact REQUIRES it.
What court of law do you think would let Microsoft successfully sue somebody for using the Mono implementation of CLR 4 on that basis? This is even before debating the explicit protections that Microsoft offers (many).
Now, CLR 4 does not run F# code directly. CLR 4 runs CIL code. The F# compiler simply generates CIL. So, from above, Microsoft is quite fine with you running your CIL on the Mono version of CLR 4.
What does the C# 4.0 compiler generate? That’s right, it spits out CIL. And you can run that CIL on the same version 4 CLR that you run your F# code on (that Microsoft is TELLING you to use).
As another little tidbit…Microsoft has released a few games and applications for iOS (iPhone and iPad). Guess what they used to write them? They used MonoTouch (by Xamarin). This is a COMMERCIAL product based on Mono. So, Microsoft is actually PAYING to use Mono themselves.
The last big Apache 2.0 code drop from Microsoft was ASP.NET MVC 4 (including WebAPI and the Razor view engine). This was just a couple of weeks ago. The big news is that they are accepting EXTERNAL contributors now. Microsoft invited Miguel de Icaza (the founder of the Mono project) to be the first non-Microsoft contributor to the project. Mono is already shipping this new Microsoft code in the Mono repository on Github.
I find it kind of quaint how some people continue to insist that using Mono is unsafe. Using Ruby is probably more of a patent litigation risk than Mono is (and I see no reason why anybody should be concerned about Ruby).
As a consultant I use any tool my customers require, one of them is .NET, since the beta days. So I know it quite well.
Still, it is a company driven language.
As the standard does not cover any recent version, nor does it cover all the libraries usually required by any standard enterprise application.
Since the early days many companies got stuck by relying in such languages. But no matter, guys like us get to earn lots of cash in porting projects.
is that Mono has lost it traction on Linux to Java precisely for the fear that API could be patented/copyrighted.
Given that MS at least somehow formally asserted they won’t do that, .Net seems a saver choice on free systems now…
So did Sun, time changes things, but of course…
Wait for the verdict.
Actually, Sun did not.
1) They actively pursued Java licensing
2) They insisted on passing the TCK to call it Java
3) You were not allowed to “superset” the language
4) If you did not pass the TCK, they did not provide a patent grant
The kind of suit that Oracle has brought against Google would not have been possIble if Sun had offered the same patent grants and binding promises not to sue that Microsoft has.
I fear that if Oracle wins, the repercussions will affect more than just Java.
It depends how Oracle wins.
1) The court decides APIs are copyrightable. This would be a disaster but there is a tonne of that precedent says this will not happen (look at the history of UNIX for example)
2) Patents. Terrible but nothing new (plus patents are basically out of the case now)
3) Other license/copyright issues – probably will not have much applicability outside of Java
Personally, I do not expect a very sizable settlement or much interesting precedent to emerge out of this case.
Oracle is correct, it is their property, they bought it from Sun. Just because this ex-exec is sad and bitter he only got one golden parachute doesn’t mean his words are true.