Robert Pogson

One man, closing all the windows.

Google: Judge, Throw out the Copyright Claims of Oracle

  • Aug 04 / 2011
  • 23
technology

Google: Judge, Throw out the Copyright Claims of Oracle

Google has moved to dismiss Oracle’s claims that Google’s Android/Linux is a violation of copyright.

“The ability to copyright Java language application programming interface specifications presents a pure question of law, Google attorney Michael Kwun said…

aside from the Java language API specifications, Oracle’s copyright claim is based on minimal alleged copying, concerning only 12 files out of over 50 thousand in Android, and material of no quantitative significance. A reasonable jury could only conclude that any such similarities are de minimis and thus not actionable. Google is therefore entitled to summary judgment on the entirety of Oracle’s copyright claim.”

This was my opinion from Day One. You cannot claim copyright on a language. Google’s developers write in Java and cross-compile to Dalvik, using Google’s stuff, not Oracle’s. This will be an easy call for the judge. He could delay a bit waiting for Oracle’s response, but I can see him ruling from the bench on such a simple concept.

see CourthouseNews – Google Tries to Toss Android Copyright Suit

23 Comments

  1. oe

    Stallman is a strange cat no doubt, but so aren’t a lot of folks whom history later recognized as ahead of their times.

  2. oldman

    “and people respect him globally.”

    Unfortunately pog, that only shows that P.T. Barnum was right.

    “There’s a sucker born every minute”

  3. Contrarian

    “Stallman did work at MIT, which governments support”

    No arguement in regard to the fact that the toot gets a lot of government grant money although that is a lot less true today than it was in the late 1900s. But Stallman only got “funding from the government” in the same sense that the guy MIT paid to pain stripes on their parking lots got “funding from the government”. Stallman’s posturing about MIT associations just profanes the real achievements of the hundreds of thousands of genuine scientists and engineers who have successfully graduated from MIT after providing real contributions to technical progress.

    In technical terms he is just a loudmouthed wannabe with an absurd agenda that brings him some notoriety from those who cannot recognize his feet of clay.

  4. Robert Pogson

    So, it’s not baloney. Stallman did work at MIT, which governments support.

    “MIT’s extensive collaboration with the federal government on research projects has also led to several MIT leaders serving as Presidential scientific advisers since 1940.”

    see Wikipedia – MIT Collaborations

  5. Contrarian

    “Stallman did work at MIT, receiving funding from government…”

    Baloney. Stallman got a BS in physics from Harvard, and got admitted to the MIT grad school in their EE program. He got a part time job as a computer operator in the AI lab, never finished his first term, and dropped out of classes. He has done no research at MIT and published nothing of interest other than his bizarre manifestos.

    Some years later, he got a grant from the MacArthur Foundation, Lord knows how. His association with MIT was as an hourly employee back when he did all the GNU stuff. After the MacArthur grant, AI Lab let him park his carcass in their spaces, but they cut off the web site connections to the FSF pages when they found out he had sneaked in links on mit.edu.

  6. Robert Pogson

    The web is nothing like ARPANET which did not scale very well. It had a little over 200 nodes by 1981.

    The web took off because the cost of connecting a server to the web was mostly the cost of service by ISPs almost universally running Stallman‘s stuff:

    LAMP, based on GNU/Linux
    Internet Standards Consortium is also Free Software
    Linux, UNIX and Ciscos are all involved in routing the Internet but Linux is the lowest cost and is widely used.

    “The benefits of open source are really clear to all the executives I work with that report all the way up to [CEO] John Chambers,” Cisco engineer Roland Dreier told InternetNews.com. “There is no question that open source is a key part of a modern engineering organization. As budgets get tighter, I think that open source will be one of the last things we would retreat from, since we see such huge benefits from the participation that we have now.”

    see Cisco (Quietly) Adds to Linux Kernel

    Stallman did work at MIT, receiving funding from government… He conceived the idea of Free Software while working at MIT.

  7. oldman

    “Yet you enjoy, use and are productive because of those freedoms every time you connect to the Internet. In your opinion they are worthless to you. That doesn’t mean they are worthless to you.”

    Nope. The internet or rather the DARPAnNet was there practically at the beginning of my carreer. I am productive because of the groundwork that was laid at taxpayer expense as part of grants to scientific and eduational institutions, long before twerps like Stallman came along and whined about so called software freedom.

  8. Richard Chapman

    [“You get the four freedoms just by obtaining the code accompanied by the licence.”]

    “As IMHO worthless as they are.”

    Yet you enjoy, use and are productive because of those freedoms every time you connect to the Internet. In your opinion they are worthless to you. That doesn’t mean they are worthless to you.

  9. oldman

    “You get the four freedoms just by obtaining the code accompanied by the licence.”

    As IMHO worthless as they are.

  10. Robert Pogson

    oiaohm wrote, “To use X function you must agree to GPL or rewrite that function completely same with LGPL”

    The GPL is a licence, not a contract. You don’t have to agree to anything to use code licensed under the GPL. You get the four freedoms just by obtaining the code accompanied by the licence.

  11. oiaohm

    The issue here is GPL also is part of using the called functions as well.

    The distinction between LGPL and GPL will still exist even if Google wins. To use X function you must agree to GPL or rewrite that function completely same with LGPL. Just removing the copyright off the header does not change the copyright on the functions.

    Yes in one way you could see the Copyright notice in an non code containing header as a custom to prevent coders from getting into trouble.

    The issue is removing a step from the cloning process officially. Currently printing out the header file and having a human type it back in is acceptable. Even if the files end up basically the same.

    Really I would liked it changed that the copyright in the header for headers like the one you pulled is for the backing code. So is legally free to be replaced when backing code is replaced. Again this makes cloning simpler.

    Some of the code in question is disassembler generated. Google wins cloning windows could be simple. Just run a disassembler over windows generate a stack of source code then processed to clean up over time.

    Google win can completely alter the model of reversing. Could be the end of closed source.

  12. saulgoode

    @oiaohm, you are correct in pointing out the flaw in the example I provided. Here is a different Android header that better illustrates the point of Google claiming copyright on nothing more than functional templates: http://android.git.kernel.org/?p=platform/dalvik.git;a=blob;f=vm/Ddm.h;h=01f5d18ba1cc1397279163ce3ed73a41b5be1b4b;hb=HEAD

    Bear in my mind that my question was largely rhetorical; I believe their copyright claims are potentially valid, assuming they are the copyright owner of the original functionality that is being abstracted in the header files.

    “Really if google wins this cloning Windows will become many times simpler. Clean room reversing requires a middle step ie documents of api function that the new abi is built from.”

    Should Google win this based on the information in header files being deemed unprotected by copyright, then not only will the protection of proprietary software from being cloned* will be impacted, but the effective distinction between the LGPL and the GPL will have been eradicated.

    * Such cloning has traditionally, and in my opinion correctly, been adjudicated under the Fair Use Doctrine, with the ultimate determination on infringement being fact specific to the nature of the copyrighted work and the market impact of the unauthorized product.

  13. oiaohm

    Opps bugger. after include should be cutils/atomic-inline.h in the include markings.

    Yes I forgot here did not have auto wrapping.

  14. oiaohm

    The 12 files are what started the mess.

    Since they are only 12 files if they are in breach google should be able to create new ones not in breach.

    Issue is that some of those 12 files were part generated by disassemble tools of Sun/Oracle provide Java runtime.

    Really if google wins this cloning Windows will become many times simpler. Clean room reversing requires a middle step ie documents of api function that the new abi is built from.

    saulgoode also you need to be more careful what you quote.
    #include
    Is a key line. If you go into that you will see it contains inline functions. Inline functions are covered by copyright. So that header file did need a copyright notice or warning of copyright notice to be clean code.

    Yes there are good example headers of google being two faced but the one you choose is a bad example.

  15. Robert Pogson

    Android/Linux is copyright. The notices can be placed anywhere and everywhere. That does not mean every part of the software is “a work”.

  16. saulgoode

    “One can copyright a particular writing of sqrt but function sqrt(x:real):real is not a “work” as defined in copyright law:”

    You claim that the declaration ‘function sqrt…’ is not a “work” yet your citations only suggest that it does not qualify as a “computer program”. I would submit that the declarations made in header files are indeed “works”; that they are “derivative works” of the “computer program” they recapitulate; that the making of derivative works of a copyrighted computer program is the exclusive right of the holder of the copyrights under U.S. law (Title 17 §106); and that those derivative works similarly benefit from the copyright protection afforded the computer program.

    Google may have a strong case for employing the Fair Use defense, but I disagree with their contention that header files are not protected by copyright. (And if Google believes that header files are not protected by copyright, why are they putting copyright notices in theirs? http://android.git.kernel.org/?p=platform/dalvik.git;a=blob;f=vm/Atomic.h;h=6c3a66f6bc2f1df4f57a0f0aa2ef08f4d075d219;hb=HEAD)

  17. Contrarian

    “You cannot copyright a class.”

    And at what law school did you learn that, #pogson?

    I think that you are wrong on that point. Even so, I spent more time than I should have looking into all this and one thing that I found was

    http://www.technewsworld.com/story/72871.html?wlc=1312510692

    It seems to suggest that Android does run java programs and hence gets into hot water copyright-wise as well as patent-wise.

  18. Robert Pogson

    Contrarian wrote, “Can the Android/Dalvik configuration run programs written in java that use the java system classes?

    No. Dalvik runs Dalvik programmes, not Java programmes. The developers work in Java and the programme is translated to Dalvik. The SDK even runs Dalvik so the developer can see the end product.

    You cannot copyright a class. Those are names and specifications/calling parameters, like header files in C. They have no “creative” content, like a phone list. A telephone directory can be copyrighted as a collection but no one copied Oracle’s stuff. One can copyright a particular writing of sqrt but function sqrt(x:real):real is not a “work” as defined in copyright law:

    “computer program” means a set of instructions or statements, expressed, fixed, embodied or stored in any manner, that is to be used directly or indirectly in a computer in order to bring about a specific result;

    “(h) in the case of a computer program that can be reproduced in the ordinary course of its use, other than by a reproduction during its execution in conjunction with a machine, device or computer, to rent out the computer program, and”

    There is no way a header or even a header file can be construed as a computer programme. It might define a data structure but it does not define an algorithm.

  19. Contrarian

    Well, I don’t believe that you are an Android developer, but maybe someone else is. Let me rephrase my statement since you seem to be confusing the issue.

    Can the Android/Dalvik configuration run programs written in java that use the java system classes?

    If so, the Dalvik engine has copied those classes exactly and may be subject to copyright claims. If not, then there is no issue that I can see although Oracle seems to see one and is pressing it in court. IP law being what it is, there is more than a simple issue at bar.

    There is a larger issue of patent infringement as well. Oracle/Sun does cannot hold a copyright on concepts, but they may very well hold a valid patent on enough of them to matter.

  20. Robert Pogson

    Contrarian wrote, “”What I think may be the case is that the overall structure of the java class libraries, i.e. the “java.xxx” methods, are what make java a “unique expression fixed in a medium” in order to lay a copyright claim. Android doesn’t use those. It’s basically a Dalvik machine that interprets its own language different from Java. The only reason Dalvik is in Android is to eliminate the need to have a Java VM.

    Dalvik is not a copy of a Java VM. It’s its own machine. Java is “write once, run everywhere” so it is not device dependent. You cannot copy device independence… That’s a concept that goes on to the 1950s or before and Oracle and SUN had nothing to do with that idea. They certainly do not hold copyright on a concept.

  21. Contrarian

    “I can see him ruling from the bench on such a simple concept.”

    There is always the chance that you are right, #pogson, same as a broken clock might be a couple of times per day. Of course Oracle has some pretty expensive and highly trained legal talent on their side who think their dog can hunt, too. So it has to go to the judge.

    Do you think that Oracle is so far off base and their lawyers are so stupid as to have overlooked the clear sense of “such a simple concept”? What I think may be the case is that the overall structure of the java class libraries, i.e. the “java.xxx” methods, are what make java a “unique expression fixed in a medium” in order to lay a copyright claim.

    If you took a FOSS program, for example, and ran the source through a word processor to just change all the variable names and even if you went into various subroutine functions and changed the sequences of code execution, you could not just re-compile and somehow avoid a copyright infringement claim. Or so I was told by some expensive lawyers a couple of years ago. It is not just the individual source code statements that are protected by the copyright, but also the effective arrangement of those statements determine the overall expression.

    Now I don’t know “beans” about java myself. :-) But I am inclined to think that in order to run a java program under Dalvik, if that is the term, one would have to mimmick the java system classes exactly, else the java program would not work correctly and the totality of the java system classes are what determines the copyrighted material.

Leave a comment