back to article Oracle tells Supremes: Fair use? Pah! There's nothing fair about 'Google's copying'

Not to be outdone by Google in ominous warnings over the future of software, Oracle has declared to American Supreme Court justices that no company would make an "enormous investment" like it did in Java SE if rivals get a free pass to copy code simply because it is "popular" and "functional". The firm filed a brief yesterday …

  1. Anonymous Coward
    Anonymous Coward

    A plague on both of them

    but let Oracle's version be far, far worse than the one that Google gets.

    Oracle wants to... pah never mind. Just Die Oracle, die.

    1. Zolko
      Mushroom

      Re: A plague on both of them

      What, actually, is so bad about Oracle ? They're a commercial vendor, if you don't like them you can easily escape them. On the other hand, it's technically impossible to completely evade from Google. I have absolutely 0 (zero) interaction with Oracle, but even though I try (hard) to avoid Google it's impossible: from Google Docs that people share to Vaze & Maps when I drive with someone using it, apart from FaceBook (which I don't use either) there is probably no other company more evil than Google.

      They did to Java what they did to Linux: it's "sort-of "the same.

      I'd actually like to see them loose on this one.

      1. m4r35n357

        Re: A plague on both of them

        New here?

      2. iron Silver badge

        Re: A plague on both of them

        While Google are certainly evil and in most cases I would like to see them lose in this case there is far, far too much at stake in the precedent it will set. The very nature of programming and the use of libraries and apis could change.

        If Oracle win then Intel could charge everyone a license fee if they write code for x86. Plus another fee for AMD if it is 64 bit. Plus fees to IBM and Microsoft and... and... and...

        It could become impossible to write anything more than 'Hello World' without employing a lawyer to sort though all the licensing and rights issues.

        1. Zolko

          Re: A plague on both of them

          "It could become impossible to write anything more than 'Hello World' without employing a lawyer to sort though all the licensing and rights issues."

          sounds good for Open-Source / Free Software. Now I doubly would like to see Google loose: we'd all be writing open-source software for ARM processors. Might be a trouble for US software companies, but should I care ?

          1. JetSetJim Silver badge
            Mushroom

            Re: A plague on both of them

            > Now I doubly would like to see Google loose: we'd all be writing open-source software for ARM processors. Might be a trouble for US software companies, but should I care ?

            You would care if you wrote code that resulted in an API that broadly resembled someone elses who happened to have a lot more money than you, discovered it, and then sued you into oblivion

            1. Zolko
              Holmes

              Re: A plague on both of them

              "code that resulted in an API that broadly resembled someone elses"

              that's not what Oracle-vs-Google is about though, is-it ? In that case, Google actively tried hard to duplicate exactly the API that pre-existed and that they knew about, in order to be able to use compilers written for that pre-existing API, for free. I have little sympathy for that.

              1. sqlrob
                Pirate

                Re: A plague on both of them

                Pre-existing API? Like, say, SQL existed before the Oracle Db?

              2. cornetman Silver badge
                FAIL

                Re: A plague on both of them

                > In that case, Google actively tried hard to duplicate exactly the API that pre-existed and that they knew about, in order to be able to use compilers written for that pre-existing API, for free. I have little sympathy for that.

                You say that like it's a bad thing.

                Public APIs made specifically to establish a standard for others to follow.

                On both sides of the fence, either the client or the supplier.

              3. Anonymous Coward
                Anonymous Coward

                Re: A plague on both of them

                They didn't try hard. They just used that API. That was seen as perfectly acceptable. Sun (who owned Java) said that was fine for them to do. No-one blinked na eye. Oracle was calling for Java to be MORE open.

                Then Oracle bought the rights to Java and decided that Android had infringed its patents and copied large swathes of its code.

                One by one it was found that they didn't copy or the code was already in the public domain before being utilised in Java. There was a ruckus for a long time over some code of a few lines that checked the range of if a variable (rangecheck). The code was basically is it higher than n then x, is it the same as n then y, is it lower than n then z. Yep it was inserted into Dalvik by the original writer of it for Java. 9 lines of code.

                SO after appeal after appeal it ended up that the only thing left was the API. This isn't the implementation of the API but the actual API structure, functions and names. Something that has been utilised forever by companies to perform compatibility.

                DO you want to write SQL but ever version you write for you have to use Select X from Y for one, Get X in Y for another, Retrieve X through Y for another? Or not allow interoperability without licencing simple interfaces? DO yo want copyright trolls backed by lawyer firms buying up old companies that just happen to have an interface that has become 'standard' by thousands of software platforms across the world (and subsequently by millions of companies) and start getting them all to pay thousands of pounds for licencing forcing many FOSS projects offline?

                It's an API , it's part of the computing and hacking world since inception (before software patents were a thing). By getting it judged as a literary work similar to a book does not do the argument justice.

                Anyway, I'm off to print("something") while I still can.

            2. James 139

              Re: A plague on both of them

              Broadly resembled? Isn't Oracles argument basically "if it looks like a duck.."?

              Such that having an API with the same name and declaration means its identical, even if said functions did two totally different things.

              Next you know they will be claiming that source code file names are copyrightable.

              1. John Brown (no body) Silver badge

                Re: A plague on both of them

                "Next you know they will be claiming that source code file names are copyrightable."

                Like Oracle, just claim it's a "literary work" and you're cool!

          2. TVU

            Re: A plague on both of them

            "...but should I care ?"

            You should care because if Oracle wins this legal case then the implications for developers are horrendous.

            While I take a dim view of Google's corporate behaviour these days, they are in this instance the lesser of two evils and I sincerely hope that they prevail in this case. Many other major tech companies are also on Google's side in this matter, e.g. Microsoft, IBM and others, and I hope that their interventions help Google's cause.

          3. ExampleOne

            Re: A plague on both of them

            > sounds good for Open-Source / Free Software. Now I doubly would like to see Google loose: we'd all be writing open-source software for ARM processors. Might be a trouble for US software companies, but should I care?

            Except glibc and gcc would immediately fall foul of the copyright on the C API, whoever it is that owns it at this point... As would all implementations of C, C++, and any other language that derives an API from C as no one ever licensed it as no one thought it needed to be.

            Who does own the C API? Is it derived from any thing older? At least we can be confident the ownership of the FORTRAN API lies with IBM. Who owns the SQL API? Who owns... As all of this is post war, we can probably safely assume it is still all subject to copyright!

            Basically, the appeals court may have had a valid point in purely abstract terms deciding that defining an API does involve a creative spark, but they failed to consider the wider implications of effectively changing the law at this point.

            Who owns the copyright on the SQL API?

            1. james_smith Silver badge

              Re: A plague on both of them

              SQL and C are international standards, so I don't think licensing them is possible even for the originators (IBM and AT&T respectively).

              1. stiine Silver badge

                Re: A plague on both of them

                But if they were protected from their first publication, then it doesn't actually matter what happens later, i.e. becoming a standard. All that means is that, in the case of SQL and C, IBM and at&t will make a shitload more money.

                1. ExampleOne

                  Re: A plague on both of them

                  > But if they were protected from their first publication, then it doesn't actually matter what happens later, i.e. becoming a standard. All that means is that, in the case of SQL and C, IBM and at&t will make a shitload more money.

                  It is even worse, because almost everything now is derived from earlier works, with some new bits added, it means everything is infringing an older copyright. Hence who owns what is going to be an interesting question.

                  For a simple example, what percentage of the java API is actually owned by K&R, as they probably never signed the copyright no one thought existed over...

                  Even though I like popcorn, this is not one I really want to reach for the popcorn for!

              2. eldakka Silver badge

                Re: A plague on both of them

                SQL and C are international standards, so I don't think licensing them is possible even for the originators (IBM and AT&T respectively).

                If the standards body doesn't have written evidence of a transfer of copyright or a license to use said copyright from the owners of the copyright, then the standards body had no business including copyrighted material in the standard, and is itself in violation of copyright, as is everyone who then followed that standard.

                The thing is, prior to this stoush, no one thought you needed a copyright transfer or license to use or incorporate APIs into the standard in the first place. If Oracle wins, then they will all have been wrong and all of them will be exposed to copyright infringement claims from the original creators of the APIs (AT&T?).

                1. Richard Plinston Silver badge

                  Re: A plague on both of them

                  > original creators of the APIs (AT&T?).

                  AT&T -> Unix System Labs -> Novell -> Santa Cruz Operation (SCO) -> The SCO Group?

                  1. Michael Wojcik Silver badge

                    Re: A plague on both of them

                    No. The UNIX copyrights were not transferred to TSG; Novell retained them. That was a major outcome of SCO Wars. So it's Novell -> Attachmate Group -> Micro Focus.

                    Also, I suspect courts would be dubious of the arguments that copyrights on much of the C standard library were included in the UNIX copyrights. Some portions of the library would have appeared in headers included with Version 4 UNIX (the first one with a C kernel), and presumably even with Versions 2 and 3 (some utilities in C); and it's likely that those headers had copyright banners; but then K&R1 came out in 1978 and it asserted copyright for those APIs. Now, the copyright for K&R1 (and for The C Reference Manual, which preceded it) was owned by AT&T; but the various versions of the ISO C standard, in their bibliographies, claim it's still owned by AT&T, which implies it was not transferred with the UNIX copyrights.

                    And those copyrights would have had the library functions in K&R form, which is substantially different from modern ISO C form. And copyright is a matter of the particular expression, not the ideas behind it. That would rather weaken any copyright assertion on APIs based on ISO C.

                    Then we have K&R2 and the standards. Starting perhaps with the drafts produced by ANSI X3J11, and certainly with K&R2 (1988) and ANSI X3.159-1989, and then with the successive releases of ISO/IEC 9899 (1999-2018, to date) we have new copyright-asserting documents for that API. This is the API in the form that's been widely (if haphazardly) adopted by libraries for a number of other languages.

                    Meanwhile, the FSF asserts copyrights over the standard headers for the standard C library implementation in glibc, though most of the text in there has previously appeared in other documents with senior claims. Microsoft similarly asserts copyright for the standard headers that appear in the Windows SDK. I didn't bother checking other vendors; I assume they do the same.

                    Copyright on the standard C library API could likely be contested by multiple parties, including AT&T, Ritchie's estate and Kernighan (if they hold the copyright for K&R2), and ISO.

                    1. Anonymous Coward
                      Anonymous Coward

                      Re: A plague on both of them

                      Great summary! Though I'm not sure the courts still hold that

                      > copyright is a matter of the particular expression, not the ideas behind it.

                      after the "Blurred Lines" rulings. Pretty sure all the old players would feel obliged to test it in court.

                      1. Michael Wojcik Silver badge

                        Re: A plague on both of them

                        Interesting point. Of course, we never know what the courts will do until they do it.

        2. James 139

          Re: A plague on both of them

          "It could become impossible to write anything more than 'Hello World' without employing a lawyer to sort though all the licensing and rights issues."

          I am not sure it goes quite that far though.

          Oracles argument is that Google has taken their APIs and re-implmented them in a way that prevents Oracle from monetizing it, ie they have taken Oracles "work" and provided an alternative that is a copy, a bit like the difference between a genuine Rolex and a knock off "Bolex".

          Writing software that uses an API would be unaffected, however an alternative version of the API would, think "printf" in GCC libC and Win32 C runtime libraries.

          The people that should be really worried by an Oracle win are projects like WINE or ReactOS, where they are directly providing the same APIs but with their own implementations.

      3. Wade Burchette Silver badge

        Re: A plague on both of them

        You want the short list or the long list? FYI, the short list is probably as big as the Bible. Here is just one example, one of thousands: Oracle makes their employees sign binding arbitration clauses and when the arbitration rules against Oracle, they sue the arbitrator. Source. If ever there was a company that had a factory just for punching puppies and kicking kittens, it would be Oracle.

  2. WolfFan Silver badge

    Please, Supremes, please

    Rule that they both lose. Please. And fine them 10% of their gross, each, for wasting the court’s time.

  3. Anonymous Coward
    Anonymous Coward

    Lifes not fair, life's just life...

    "Oracle has declared to American Supreme Court justices that no company would make an "enormous investment" like it did in Java SE if rivals get a free pass to copy code simply because it is "popular" and "functional"."

    While I don't dispute Oracle's statement - it in no way strengthens or proves Oracles argument that Google has infringed.

    Please oh please won't somebody protect poor little Oracle?

    All Oracle did was buy somebody else's code, try to change the rules to screw people over Java licences but it got screwed in the process. Supposedly.

    It's all so unfair...

    1. EricM
      Joke

      "Creative choices", yeah, sure ...

      Actually Oracle bought SUNs server biz.

      Java was an addition they never knew how to use - until they came up with using it as tactical weapon in court to unfairly attack competitors.

      "creative choices" in API naming... can't make that shit up.

      The only thing "creative" I see here is the Oracle legal argument....

      1. Anonymous Coward
        Anonymous Coward

        Re: "Creative choices", yeah, sure ...

        Joke Alert notwithstanding, Oracle bought Sun for Java and was convinced to buy the rest. (Being a huge Solaris/Sparc fan, that day of infamy still saddens me.)

      2. Anonymous Coward
        Anonymous Coward

        Re: "Creative choices", yeah, sure ...

        Oracle had been a heavy user of Java for a long time. And they clearly knew beforehand what Google did, and what Sun's position on it was, because it was no secret. Pretending *they* came up with the idea is ludicrous. Sun simply couldn't do to Google what they could do to Microsoft a decade earlier.

    2. big_D Silver badge

      Re: Lifes not fair, life's just life...

      "Oracle has declared to American Supreme Court justices that no company would make an "enormous investment" like it did in Java SE if rivals get a free pass to copy code simply because it is "popular" and "functional"."

      Hmm, let me see, Linux Foundation, Red Hat, SUSE, Ubuntu, Greenbone Networks, Mozilla, I could name others, but there are lots of companies that invest heavily in open platforms.

      1. stiine Silver badge

        Re: Lifes not fair, life's just life...

        re; big_D.

        Maybe this is why RedHat was purchased by IBM.

    3. Anonymous Coward
      Anonymous Coward

      Re: Lifes not fair, life's just life...

      This 10-year old article from the beginning of the lawsuit sheds a different light. Sun was pissed at Google, too. Google did hire former Java devs, so getting to the same API can hardly be a coincidence. And they hated the GPL in the same way that Microsoft or Oracle do, because they wanted a commercial, closed-source product, none of that FOSS hippie stuff.

      It's super hard to see Google as an innocent actor here. They did what they did because they knew Sun was in no position to retaliate.

      https://www.cnet.com/news/why-oracle-not-sun-sued-google-over-java/

      1. james_smith Silver badge

        Re: Lifes not fair, life's just life...

        Wrong. Sun released a lot of code under open source licenses and funded a number of projects that they never stood to benefit financially from (TCL/Tk for example). They were too much of a technical company really, investing way higher percentages of income into R&D to the annoyance of short term investors.

      2. Michael Wojcik Silver badge

        Re: Lifes not fair, life's just life...

        It's super hard to see Google as an innocent actor here. They did what they did because they knew Sun was in no position to retaliate.

        And that's irrelevant to the question before the court, which is whether APIs are protected by copyright.

  4. Anonymous Coward
    Anonymous Coward

    Big Red

    When I first started working in oil and gas exploration it was with Halliburton - commonly referred to as Big Red. (Back then, Halliburton was relatively unknown outside that industry but was, in fact, a multi-national employing over 100,000 people worldwide. It got better known following contracts with it's broader engineering side, all acquired through buy-outs, and the fact that it's CEO became US VP).

    Anyway, they were known as Big Red before Oracle was a gleam in anyone's eye so why hasn't Halliburton started sueing for copyright of their name? </sarcasm>

    1. WolfFan Silver badge

      Re: Big Red

      I'm fairly sure that this guy https://en.wikipedia.org/wiki/Hellboy was known as Big Red well before Oracle, too. Perhaps Marvel should join in that lawsuit.

      1. Michael Wojcik Silver badge

        Re: Big Red

        To me it will always be the home ranch of the Pratt & Ferris Cattle Company.

        (1886, friends. Spare me your weaksauce twentieth-century citations.)

  5. Anonymous Coward
    Anonymous Coward

    I'm confused by the statement that "Google could have just use the open-source Java license" - I seem to remember that Oracle was only willing to license Java ME, while the Android people needed Java SE capabilities for their smartphone OS.

    1. Blank Reg

      that's right, the Java SE license did not allow it to be used in mobile devices. but Google didn't need SE, Java Me could do everything that they needed, they just didn't want to fork over the few pennies per device in licensing fees.

  6. Frederic Bloggs

    OpenJDK

    Well? What happens to OpenJDK if Oracle wins?

    Hmmm???

    1. Dabbb
      Mushroom

      Re: OpenJDK

      Oracle, as a main sponsor and developer of OpenJDK, will continue to maintain it.

      Yeah, I know, mind blown.

      Downvote is on the right, don't miss it.

      1. Michael Wojcik Silver badge

        Re: OpenJDK

        Right up until they don't.

        That said, copyright shouldn't be an issue for anything in OpenJDK now, and for anything Oracle add to it, because they're the copyright holder. None of the OpenJDK distributions (AOJ, Azul, Amazon, etc) are copying the API; they're getting it from Oracle.

        What Oracle could do is add new APIs to Oracle Java and not add them to OpenJDK. Then, if they win this case, no one else could add those APIs to OpenJDK without licensing the copyright from Oracle.

        That's not a huge danger, because new APIs are new; no one's locked into them yet. And Oracle hasn't been particularly outstanding at maintaining Java anyway. The community of vendors and volunteers around OpenJDK likely have the resources to do a better job of implementing whatever new features might have significant market appeal.

  7. HildyJ Silver badge
    Facepalm

    What is an API and why it matters

    Say you have an Application Program and you want to Interface with the file system to Open a file. Somebody somewhere wrote the first application to use the word "Open". Is that copyrightable?

    1. Michael Wojcik Silver badge

      Re: What is an API and why it matters

      Copyright, in the US, applies to any artistic work in a permanent medium. It applies immediately - you don't have to ask for it.

      However, copyright protection would not attach to just the word "open" by itself. It fails multiple tests: it's not creative (it's a common English word), it's too short, it's functional rather than expressive.

      Oracle's case rests in part on a claim that the API in question, as a whole, is large enough to be expressive and non-trivial. Lower courts disagreed, but CAFC (which is tied with FISC for the federal government's most useless court) overturned. Now we'll have to see what SCOTUS does.

  8. Marketing Hack Silver badge

    Java APIs are a "literary work"?

    Well, I admit they are better reading than the "50 Shades of Grey" novels, but really.....

    1. Jay Lenovo
      Pint

      Re: Java APIs are a "literary work"?

      So when will Programmers deservedly earn their Nobel Prize for Software Literature?

  9. inquisitive2014

    Copy Left

    I don't have a dog in this fight but I am concerned that we are missing the point about Open Source Software. One of the key tenets of Open Source Software is the idea of "copyleft". As I understand it this principle says that you can use the Open Source Software in your original work but you must recontribute any changes back to the Open Source community. It is my understanding that Google declined to accept an Open Source licence for this but rather took the code and used it in their products which was proprietary and sold.

    Copyleft is worth fighting for

  10. ZenCoder
    Linux

    FYI Android OS is Open Source

    Android OS is covered by the Apache License 2.0, except for the linux kernel which is GPLv2, and there are plenty of custom roms https://en.wikipedia.org/wiki/List_of_custom_Android_distributions

    The apps that run on top of Android written by Google are another matter. I am not asserting anything other than the kernel and core OS are open source in a real, meaningful, practical way. If you already knew then then I am sorry I wasted your time.

  11. Raedwald Bretwalda
    Stop

    Anyone who things an API is not a creative work (in the broadest sense, although I believe "creative work" is a technical term in coyright law) ... has never designed a complicated API.

    1. Michael Wojcik Silver badge

      An API is at best a creative work that's heavily constrained by functional factors. That's substantially different from other expressive forms protected by copyright.

  12. Michael Wojcik Silver badge

    The amicus briefs are more substantial

    Frankly, Oracle's own filings are pretty feeble, in my view. (I am not even approximately a lawyer.) They're mostly bluster.

    The various amicus briefs that have been filed - and there are a lot - seem to have much more substance. There was one filed yesterday by "INTERDISCIPLINARY RESEARCH TEAM ON PROGRAMMER CREATIVITY", which is a group of five academics in various disciplines, supporting Oracle; I don't agree with their arguments, but they're at least based on evidence and reasoning.

    The arguments in the Interdisciplinary brief mostly hinge on trying to develop a better legal test for "creativity", and the authors generally appeal to 1) a multitude of options from which 2) an intellectual selection is made. Personally, I don't think that's a particularly good test (when I make a sandwich, I use my intellect to select among a multitude of options, but I don't believe the result is particularly creative); I don't think the Java API, or APIs in general, meet it to a satisfactory degree (because of the functional constraints on APIs); and I don't think a creativity test should serve as the guiding principle for whether copyright should attach in the case of APIs, because of the chilling effect on industry.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2020