Oracle vs. Google and Java Patents

6 Comments

• 
  Florin November 17th, 2010 at 07:56

  Evil side of both companies: agree. But still, I think Google is less evil than Oracle.

  Did you look at Google’s counterclaims? There are also some interesting comments on that document on groklaw. Essentially, Google says three things: first, they did not breach the copyright and did not infringe on patents, second, Oracle’s patents are not valid anyway and if there is copyright breach in Android’s code it’s not in code written by Google, and third, Oracle should lose the case even if it is right in its statements, because neither Sun nor Oracle have previously enforced the patents in Java, Oracle has defended a complete opening of the Java platform in the past, but changed its mind later on and by doing so has misled people into believing they can use the Java platform openly, even after it was acquired by Oracle.

  Furthermore, by not being specific in their claims, and not putting their fingers on the Android code which violates copyright or infringes patents, although the Android code is freely available for quite some time, Oracle has put itself at risk to have the case dismissed for different reasons (there is a precedent).

  Google has to get only one statement accepted by the court on each count, and the entire case is dismissed. I think Oracle’s position is weak. But then again, American justice is weird, so we’ll never know until it’s over.

  Reply
• 
  Florin November 27th, 2010 at 13:09

  Just remembered one more thing: at the time MS was sued by Sun, MS was clearly in breach of licensing terms, since they provided a Java VM which was not 100% compatible with Sun’s VM. As opposed to Google, who doesn’t provide _any_ Java VM. So I don’t think there’s double standards being applied when people defend Google although they didn’t defend MS.

  Reply
  • 
    Radu November 27th, 2010 at 13:13

    Dalvik is a JVM… Furthermore Google has infringed the patents granted by not respecting the copyright notice you see in the post. Dalvik is a subset of the JVM which doesn’t allow code written for another JVM to be executed. No more cross-platform implementation, although that’s the strongest attribute of Java.

    The Java platform is open (since 2007, with some minor proprietary exceptions – see here), only that when one creates a JVM has to respect the standards.

    Regarding the code example, I have tweeted a picture a few weeks ago which is now an admitted proof to the case. So, do you still think that Google is not evil regarding this?

    Reply
    • 
      Florin November 28th, 2010 at 19:38

      Dalvik is a VM indeed, but IMO it’s definitely not a JVM. Its bytecode format is different. Even its architecture is different (stack machine vs. register-based – says wikipedia). The fact that there is a translator from Java source code to Dalvik bytecode does not make it a JVM, IMO. The only thing that would make it a JVM was proof that it can run Java bytecode, which it cannot. There are lots of other translators, for other languages, too.

      Since Dalvik is not a JVM (MO), what JVM-related copyright notices has Google infringed? If you disagree, could you please explain what makes Dalvik a JVM in your opinion?

      I agree, Dalvik took over lots of concepts and tried to stay as compatible as possible with an actual Java platform regarding the provided APIs. But these APIs are not copyrighted in themselves. IMO, it would be a quite damaging thing to allow APIs, not implementations, to be copyrighted or even worse, protected by patents. Can you imagine what this would mean for ReactOS or Samba? Or even Wine? I think even gcj would be in danger of being discontinued.

      Particularly about the code in the picture you have posted on twitter, here’s a comment on the issue (IMO a pertinent one): http://carlodaffara.conecta.it/?p=549. The code in question is not Google’s code, but was taken over from Harmony, which is licensed under the Apache license, which took it from OpenJDK, which is licensed under the GPL. Which supports Google’s claims that if there is any copyright infringement, it isn’t in Google’s code.

      Additionally, you are a coder. Try to code the same logic that’s expressed in the code you list. Then look if your code looks significantly different. I think you’ll find that there are no larger differences between any two of the code samples you posted plus your version. The code clearly isn’t identical to the bit. A claim of copyright infringement on that code is IMO almost similar to a claim on copyright infringement on two C versions of “Hello World”.

      I think Google has definitely not created Android for fun, but for its potential to make money. But I don’t think this is evil in itself. IMO, Google’s evilness is to be found in the way it uses the information it collects – there are lots of cases emerging where Google was evil in this regard. But still, what Google does for the open source movement is IMO laudable. Not so what Oracle does – it contributes very little, and now puts thousands of open source projects at risk of being shut down or loosing user/sponsor confidence. Besides, Oracle’s openly admitted policy on open source is to prey on it, not to contribute to building a viable services-based business model based on open source (like IBM does, for instance). Like if Google was a gardener, whereas Oracle was just a gatherer. Which automatically makes Oracle more evil for me – in the open source world, it consumes resources in an unsustainable way.
    • 
      Radu

      November 28th, 2010 at 21:02

      I corrected myself in the same comment by saying that Dalvik is a subset of a JVM (it includes classes from Java SE 1.5 – at least collections and threads from what I’ve read until now). I shouldn’t have written that Dalvik is a JVM. Regarding the code in the picture I posted on Twitter, check this out.

      Anyway, as I have said in the post, let’s let the court decide on the validity of the case and let’s not jump to rushed conclusions.

      And Oracle does contribute to the FOSS movement. More details here.

      I am not on any company’s side, I just want people to stop being so radical when it comes to conflicts like this.
    • 
      Florin December 12th, 2010 at 19:43

      Just one more comment: if you look at the open source projects Oracle supports they are of two sorts: first, the most of them, open source projects which Oracle got through acquiring other companies (Glassfish, MySQL etc.), second, projects pushing a technology in which Oracle has a specific commercial interest (Apache MyFaces Trinidad – JDeveloper didn’t get off as a commercial product, so they open sourced a tiny piece of code related to it, in the hope they’ll get more developers using it).

      BTW, are you aware that the ASF has quit the JCP? Did you look up the ASF’s motivation? I think it’s interesting to read, not necessarily in relation to the Oracle vs. Google lawsuit.

      OTOH, I like Java, and I like the ecosystem built around it even more. What Oracle does makes me think I’ll have to switch to something new – something that will most probably come from Google. So I’m obviously biased.

    [language]
    code lines
    [/language]
	

    [bash]
    /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
    [/bash]
	

    The <code>$USER</code> variable holds the current logged in username.