back to article No do-overs! Appeals court won’t hear $8.8bn Oracle v Google rehash

Over eight years of feuding between Oracle and Google over the use of Java code in Android may be nearing its end following a Tuesday court ruling. The US Federal Circuit Court of Appeals has declined [PDF] to re-hear the case in which it found Google to be in violation of Oracle’s copyright on Android API code. The Chocolate …

  1. JohnFen

    On the one hand

    On the one hand, I think this case was wrongly decided, in a way that causes harm to all software developers, and I hope that the ruling doesn't stand in the long term. On the other hand, this is a battle between two awful companies, and I'd love to see them continue to burn money to spite each other.

    1. a_yank_lurker

      Re: On the one hand

      What I would like to see is the verdict overturned with some harsh words by the Nine Seniles. Then for Chocolate Factory to sue Leisure Larry and his minions for a malicious suit to keep it going.

    2. bazza Silver badge

      Re: On the one hand

      Well it depends on your view as to how strong copyright law should, or shouldn't be. Personally speaking I believe copyright laws should be strict; a grey area on that matter won't serve anyone.

      Remember that the only thing that forces things like Linux to be open in the way it currently is is the strict interpretation of copyright laws. Any easing off of that interpretation damages things like GPL and software based on it. It's then a slippery slope for the GPL, whose only protection is copyright, and companies with deep pockets could make it more slippery.

      And never mind Linux. Any open source copyrighted code becomes fair game. If Google are allowed to use bits of Oracle's code without permission, why can't I use bits of your source code without your permission? Of course just good manners would prevent me from doing something like that, but what about a money grabbing corporate whose only incentive is profit?

      1. EricM

        Re: On the one hand

        > Remember that the only thing that forces things like Linux to be open in the way it currently is is the strict interpretation of copyright laws.

        Copyright for code: Yes. Copyright for APIs: Hell, no.

        A good part of the SCO wars was about the claim that infringing code implements the same API as Unix and therefore is owned by the copyright holder of Unix (SCO).

        Which at the time was eventually rejected by the courts.

        Now in this case a clean-room re-implementation of an existing API IS regarded as copyright violation. That's outright dangerous.

        To that end Linux and other O/S software that cleanly re-implements existing APIs might even be endangered by the Oracle/Google ruling...

        1. bazza Silver badge

          Re: On the one hand

          @EricM,

          Ah yes, SCO. But isn't there a subtle difference between the two cases? I thought SCO were claiming ownership of the API specification (corrections most welcome, memory slightly hazy); Linux didn't re-use any of the source files from Unix. It was a ground up fresh implementation, including header files, and anyway has a different system call interface. Whereas Google simply lifted parts of Oracle's code, admitting as such as part of this case.

          I'm fairly certain that had Google implemented the same API identically without simply copying bits of Oracle's source code files, it would never have got this far. People have been clean room reimplementing APIs for decades - IBM's BIOS being a prime example - without so much as a hint of legal difficulties.

          Otherwise I admit to grasping at straws to explain why one historic case concerning (at least in part) ownership of APIs hasn't prevented the Google/Oracle case getting this far.

          On general principles I don't like the idea of a giga-corp like Google getting away with a poorly defined fair-use decision, the consequences of which might have an unintended consequence.

          1. Doctor Syntax Silver badge

            Re: On the one hand

            "Whereas Google simply lifted parts of Oracle's code, admitting as such as part of this case."

            AIUI only a few lines of implementation were copied, the rest was simply the declarations.

            1. bazza Silver badge

              Re: On the one hand

              @Doctor Syntax,

              AIUI only a few lines of implementation were copied, the rest was simply the declarations.

              Yes, but the files containing the declarations were also copyrighted too I presume. However I don't know whether Oracle's submissions covered these or whether the court case considered them. Judging solely by the popular reaction "don't copyright APIs" I'm guessing that they did.

              As to whether or not that's right, a copyrighted file is a copyrighted file, no matter what it contains or defines. If I took someone's non-free paper specification manual for an API and published it in my own name, I can rightly expect to be contacted by a bunch of keen lawyers eager for redress.

              However it is pretty pointless making people jump through a clean-room rewrite of an API's declarations simply to avoid the copyright on them. If they want it they're going to do it (unless they're Google, or so it seems). But if that's the price of upholding strong copyright laws then perhaps that's a price worth paying.

              1. Doctor Syntax Silver badge

                Re: On the one hand

                "Yes, but the files containing the declarations were also copyrighted too I presume."

                The point at issue isn't whether it's copyright, it's whether using the API is fair use. That's been the assumption in the past. If an API can't be incorporated into applications without infringement the library it supports is useless.

          2. Anonymous Coward
            Anonymous Coward

            Re: On the one hand

            My understanding was that this wasn't the basis for the successful appeal. The appeal court ruled plain and simple that the API itself could be copyrighted. The half dozen lines of lines of code which might and might not have come from the publicly-available test suite weren't what was at issue

          3. Peter Gathercole Silver badge

            Re: On the one hand @bazza

            The SCO case originaly hinged around SCO's assertion that IBM included parts of the source code obtained under IBM's UNIX source code license that they held for IX and AIX into the code they contributed into Linux (particularly LVM code).

            What became apparent is that the only code that was in Linux that came from a UNIX source tree was from ancient UNIX (Edition/Version 7) which SCO themselves had put into the public domain under a fairly unrestricted license. When SCO, with full access to the AIX source tree, were unable to demonstrate anything else more than a general resemblance in the TTY and other device switch (which were basically a series of C switch [case] statements which made no sense to write any other way), that part of the case collapsed.

            It then became muddied, because Novell successfully challenged SCO's claim to the the copyright holder of the UNIX source in the first place!

            Apart from the entertainment value, I'm so glad that those cases has finally been put to bed.

            In this case, I thought that Google had bee accused of directly copying the include files (and only these files) that essentially defined the API between the application and the runtime. I thought that Java had actually been published under a fairly permissive license by Sun (as they were very keen to get it adopted as a pervasive language), so I'm actually surprised that this case has come to this conclusion. But I suppose it's Oracle, so all reason goes out of the window as greed takes over.

            Not that Google's any better these days,

          4. EricM

            Re: On the one hand

            @ bazza

            > But isn't there a subtle difference between the two cases?

            There may be a lot of subtle and not-so-subtle differnces, sure. I'm a CS major, not a lawyer.

            However the latest appeals court ruling makes in my understanding - and I'm probably over-simplifying here - APIs copyrightable and therefore even clean-room re-implementations of any given API might/does violate copyright.

            From Wikipedia:

            > This led the court to conclude "that the overall structure of Oracle's API packages is creative, original, and resembles a taxonomy" (p. 14). It therefore reversed the district court on the central issue, holding that the "structure, sequence and organization" of an API is copyrightable.

            Maybe there are some other clauses that cause the myriad of other APIs in the industry to not be affected by this ruling. As stated: IANAL. And mybe my understanding of this ruling is incorrect.

            BUT if not, this ruling could be rather dangerous for nearly every software developer ...

      2. Doctor Syntax Silver badge

        Re: On the one hand

        "Personally speaking I believe copyright laws should be strict; a grey area on that matter won't serve anyone."

        The only thing worse than a grey area here is placing APIs under copyright.

        There is a grey area in copyright and that's the idea of fair usage. The assumption which has been fundamental to S/W development in general, not just open source, is that use of the code that defines an API is fair use. If the line is drawn on the other side of that then part of almost every body of code are infringing copies of someone else's code. Oracle are probably as vulnerable as anyone else here: inter alia who owns the copyright to SQL?

        This decision is dangerous to anyone in S/W development wherever the decision is binding. It's not too hard to envisage development driven out of the US altogether and any S/W sold into the US carrying a copyright surcharge.

      3. Pascal Monett Silver badge

        Re: "the only thing that forces things like Linux to be open"

        Is that they were presented like that from day 1, and thus they stay that way.

        I do not care what copyright law says, if someone says : this is free, then I can take it and nobody can complain.

        The law can dance any way it wishes, it cannot prevent free.

        There can, however, be conditions to free. Not paying does not necessarily mean you can do anything with it, as in free for private use means you cannot build a company around it. But that is in the contract when you download the code and you are supposed to know about it.

        1. Doctor Syntax Silver badge

          Re: "the only thing that forces things like Linux to be open"

          "I do not care what copyright law says, if someone says : this is free, then I can take it and nobody can complain."

          Actually, as someone else has pointed out, the GPL is a licence granted under copyright law with certain conditions attached. The BSD licence is similarly a grant under copyright law with rather fewer conditions. But both depend on copyright law to keep them free (for different values of "free"). So you should care what copyright law says.

    3. TVU Silver badge

      Re: On the one hand

      "On the one hand, I think this case was wrongly decided, in a way that causes harm to all software developers, and I hope that the ruling doesn't stand in the long term. On the other hand, this is a battle between two awful companies, and I'd love to see them continue to burn money to spite each other"

      I agree though this time the choice is between bad (Google) and worse (Oracle) because of the obstacles that will be created for developers by an Oracle win. I hope that Google does go ahead with the Supreme Court final appeal and that they go on to win.

  2. Mark 85

    So what's next, some patent troll will sue for using one of the C languages? The has all the look and feel of Oracle being a patent troll.

    1. Anonymous Coward
      Anonymous Coward

      No Trolling

      Regardless of how one views the ruling, Oracle isn't trolling here. Oracle owns Java via the acquisition of Sun Microsystems.

      1. Fungus Bob

        Re: No Trolling

        And Sun was OK with Google copying the API's. Oracle is hardly in the right for coming along afterward and changing things. This is reminiscent of the company that bought the JPEG patent and tried to bill the world.

        1. JLV

          Re: No Trolling

          Without in any way supporting Oracle here, Sun had already pulled a bunch of shady tricks wrt “standard, open” Java for marketing purposes while asserting control by any means possible.

          Anyone remember Sun’s Java compatibility tests, where they basically decided who’d get annointed or not? This crap went on way past the MS J++ sliminess that might have justified it in the beginning.

          re copyright. code/implementation, by all means, depending on license. API - no way, that would have such a stifling effect on software innovation, would strengthen monopolies and cartels. and it fundamentally contradicts the stated purpose of _public_ ways to interact with software artifacts.

          Who would benefit?

        2. Anonymous Coward
          Anonymous Coward

          "Oracle is hardly in the right for coming along afterward and changing things"

          So if you buy a bankrupt company you're bound to the same behaviour that brought it in that situation, and you shoudn't change anything? When Jobs got back to Apple it should have not changed anything and keep on with the bad ideas of his predecessors? One of the first thing he made was to kill the licensed clones...

          Why Sun open sourced Star Office? Should it be allowed to bought it and "change things"?

          Sun was OK? Did it ever write that in an official document? Did it license Java to Google? No, it didn't, and Google didn't bother even to ask - even if inside Google someone warned about that. So whatever its pony-tailed previous CEO thought, it's really irrelevant.

          New owners, new rules, sorry. And one of the main reasons Oracle bought Sun was exactly to take control of Java.

          1. Fungus Bob

            Re: "Oracle is hardly in the right for coming along afterward and changing things"

            "New owners, new rules, sorry. And one of the main reasons Oracle bought Sun was exactly to take control of Java."

            As successors in interest, the new owners can't change past agreements. Sun, by not suing, allowed Google to do what they did. Oracle could have said no new copying, but they can't sue for past behavior that the previous owners were apparently OK with. Again, look up what Forgent Networks Inc. tried with the JPEG patent.

            1. Anonymous Coward
              Anonymous Coward

              Re: "Oracle is hardly in the right for coming along afterward and changing things"

              No. You are not forced to sue instantly, and the law allow you to sue later.

              "the new owners can't change past agreements."

              Sure, legally valid agreements. In this case, there were none.

              "they can't sue for past behavior that the previous owners were apparently OK with"

              Even you say "apparently" - and that's not really a valid, binding, legal agreement.

              Had Google a valid license, Oracle could have done nothing. Google was to greed to ask for one, maybe sure that Sun obsession to get "Java everywhere" would have ignored it - just Oracle has different plans, and owning Java, can now sue Google.

        3. the spectacularly refined chap

          Re: No Trolling

          And Sun was OK with Google copying the API's. Oracle is hardly in the right for coming along afterward and changing things. This is reminiscent of the company that bought the JPEG patent and tried to bill the world.

          I think you mean the GIF patent. JPEG has never been encumbered.

          1. Fungus Bob

            Re: No Trolling

            "I think you mean the GIF patent. JPEG has never been encumbered."

            Nope, JPEG. Look up Forgent Networks Inc.

            1. Anonymous Coward
              Anonymous Coward

              "Nope, JPEG. Look up Forgent Networks Inc."

              Forgent asserted that one of its patents was applicable to the JPEG format - which was later ruled out (still, they made money before when some companies agreed to pay).

              It's wholly different from this situation, because Oracle now fully owns the whole Java.

      2. JohnFen

        Re: No Trolling

        Oracle owning Java has no bearing on whether Oracle is trolling or not.

    2. Nick Kew

      Um, wouldn't being a "patent troll" imply use of patents somewhere in the case?

      And since both parties here are actual technology companies with actual technology businesses, they don't fit the definition of "patent troll", however much you may want them to.

      1. TheVogon

        "Um, wouldn't being a "patent troll" imply use of patents somewhere in the case?"

        And being a troll would require Oracle to be an NPE. Whereas Oracle do in fact sell Java based products.

        1. TVU Silver badge

          "And being a troll would require Oracle to be an NPE. Whereas Oracle do in fact sell Java based products"

          True, but this most recent case is still an act of speculative and malicious money grabbing by the notoriously anti-open source Oracle which is why I hope that they go on to lose at the Supreme Court.

    3. Doctor Syntax Silver badge

      "The has all the look and feel of Oracle being a patent troll."

      Yet another of those things we have to repeat over and over again. Patents and copyrights are not the same thing.

    4. Anonymous Coward
      Anonymous Coward

      "some patent troll will sue for using one of the C languages?"

      C/C++ are ISO/IEC standards, so that is impossible. Derived languages like C# may be copyrighted.

      Google decided to make a 1:1 copy of Java - without asking for a license - because it would have allowed it to reuse a lot of existing libraries and tooling to hasten Android deployment and reduce costs.

      It looks asking Sun for a license would have cost them far less... but that isn't the Google way.

      It's interesting to note that OS/2 supported Win 3.x API because IBM had a license to do so.

      Probably BIOS/DOS calls couldn't be copyrighted because they were just some registers value and INT calls- more an ABI than a API.

      1. Doctor Syntax Silver badge

        Re: "some patent troll will sue for using one of the C languages?"

        "C/C++ are ISO/IEC standards, so that is impossible."

        The issues would be more with the standard library definitions. Just because it's a standard doesn't mean it's free. Standards can and do incorporate proprietary technology but this has to be covered by FRAND licences although this normally applies to patentable material rather than copyright.

        APIs in standards could be contributed by a 3rd party. If not they're going to be copyright of the standards body and that doesn't guarantee they're royalty free.

        1. Anonymous Coward
          Anonymous Coward

          "The issues would be more with the standard library definitions. "

          AFAIK, the C++ standard library underwent ISO standardization - I never said because they are ISO standards they are "free" nor there are no patents or the like, just it's impossible or very unlikely that a patent troll could sue - because the standardization process should have checked what patents are involved.

      2. 2Nick3

        Re: "some patent troll will sue for using one of the C languages?"

        "It's interesting to note that OS/2 supported Win 3.x API because IBM had a license to do so."

        And OS/2 with WinOS2 was more expensive than OS/2 with "Bring your own Windows" because of that - IBM transferred the licensing cost to the customer.

  3. whitepines

    This goes interesting places. Maybe IBM can enforce copyright on everything that looks vaguely like a PC or C style API, sending the US back to the 1960s in computing technology! Or only commercial compilers will be legal, all open source banned?

    Hope the supreme court picks this one up and clears that nonsense out of the wings.

    1. Anonymous Coward
      Anonymous Coward

      From what I understand, it's pretty unlikely that the Supreme Court would intervene here. It's way too specific to be of any interest to them.

      1. Doctor Syntax Silver badge

        "It's way too specific to be of any interest to them."

        The bad news is that they might see it that way but it's far from specific. The precedent it sets is frighteningly wide ranging.

    2. B83

      Is it IBM compatible

      Haha that comment made me laugh, maybe too early in the morning for me since I've not had my coffee yet!

      What popped to mind was, I can remember the days when people actually said "Is it IBM compatible" and then they totally lost the plot and Windows made its move.

      IBM still exist but could have been a very different beast if their strategy had been different in the 80's.

      I'll go have my coffee now, cheers

      1. Pascal Monett Silver badge

        Re: "if their strategy had been different in the 80's"

        If that were the case, IBM might have filed a patent on their PC and we wouldn't have the Internet today, or smartphones, or much of anything else in computer tech because IBM would have stifled the market by forcing everyone to work with it at its prices.

        Companies would have a few PCs, like in accounting, design and HR, but that would likely be all.

        Maybe Apple would evolved very differently, and the war between PC and Mac would have given Apple a far greater share of the market, but we're still talking closed design here.

        It is the fact that IBM never put a patent on its toy that birthed the computing world we have today.

        And gave many of us our careers.

        So thanks, IBM, for not having had today's patent lawyers on hand back then !

        1. John Brown (no body) Silver badge

          Re: "if their strategy had been different in the 80's"

          "or much of anything else in computer tech because IBM would have stifled the market by forcing everyone to work with it at its prices."

          One of the OS options on the original IBM PC was CP/M 86 because there was already a PC ecosystem out there. IBM were johnny come latelys to that game, and even then it was sort of "sneaked past" the higher level suits. If they had patented and fought to keep what they invented, then the existing market would have worked around them and maybe we'd have a more diverse hardware and OS market. That could be both good and bad depending on your point of view, but it would certainly be a different world in terms of computers.

          1. Michael Wojcik Silver badge

            Re: "if their strategy had been different in the 80's"

            If they had patented and fought to keep what they invented, then the existing market would have worked around them and maybe we'd have a more diverse hardware and OS market.

            Indeed. In the early years of the IBM PC, there were a lot of 8- and 16-bit systems available, and some 32-bit ones (like the Fortune 16/32). IBM's marketing power and existing customer base, plus the market for PC clones,1 led to the eventual dominance of the IBM PC.2 But it could have gone much differently.

            Even after the IBM PC had caught on, there were certainly moments when it looked like there might be serious competition. Pretty much everything else outperformed it in one way or another. Had Apple come out with a Mac II-style separate-display Mac sooner, or released a cheaper Lisa after the Mac rather than before it... who knows? A cheaper, more open DEC Rainbow might have had a chance. Xerox might not have screwed up marketing the Alto so very, very badly. And so on.

            1It's important to remember that the IBM PC was not the only architecture that got cloned. Apple had Franklin, for example; it was Apple v. Franklin that led to the use of clean-room development for cloning the IBM PC BIOS. (And that's apropos this article, since it was a software copyright decision.) And there were open architectures such as S-100, which started as the Altair bus but was widely used by other manufacturers.

            2Lynn Wheeler has argued (in alt.folklore.computers, I think, and no, I'm not going to search for a citation) that the PC's ability to serve as a 3270 terminal - first through third-party cards like the DCA IRMA, then with IBM's hybrid 3270 PC, and eventually with 3270 and TN3270 emulators over LAN connections - gave it a big boost for business adoption. Mid-level managers could justify getting a PC to replace their dedicated 3270 terminal as a way to be "more productive", and it became a status symbol. Having made inroads with the 3270 users, it would have become a standard piece of business equipment at the managerial level; that reduced demand for the typing pool, which gradually disappeared, making PCs necessary for all employees who created documents. I don't know of any methodologically-sound research supporting this, but it fits my anecdotal experience of the era.

        2. Michael Wojcik Silver badge

          Re: "if their strategy had been different in the 80's"

          It is the fact that IBM never put a patent on its toy that birthed the computing world we have today.

          According to various sources, such as this one, IBM received nine patents for technology in the original IBM PC. It wasn't that much trouble to dig some of them up; see e.g. US4442428 and US4528626.

        3. B83
          Thumb Up

          Re: "if their strategy had been different in the 80's"

          Yes this is me replying to a comment made nearly two years ago. Home working during Corona Crisis and I am bored i.e. I much prefer the office environment. After making a comment on a ASP.NET article I discovered you can review the comments you historically made, re-looked at this article and saw this comment that made me think, if only I had become an author!

          What a fantastic book your point would make. If IBM had put patents on there stuff and had those kind of patent lawyers where would we be with tech at this precise moment (2020).

          How would the smart phone evolve, if at all, Tablets, home computing, the rise of India as an major IT outsourcing area, gaming, the list is massive.

      2. Claptrap314 Silver badge

        Re: Is it IBM compatible

        I doubt I will ever forget the discussion that a buddy and I had with a professor in 1992 about IBM compatibility.

        "I just want the computer to be IBM compatible."

        "Then don't buy IBM."

        .....

        Very, very fond memory.

    3. Doctor Syntax Silver badge

      "Maybe IBM can enforce copyright on everything that looks vaguely like a PC or C style API"

      What rights do IBM have over C. That came from Bell Labs. SQL, however..... How would you like that, Oracle?

      1. Anonymous Coward
        Anonymous Coward

        SQL, however..... How would you like that, Oracle?

        https://www.iso.org/standard/53681.html

        If IBM never asserted copyright or patents, and it became an ISO standard... Oracle is perfectly fine.

        1. Doctor Syntax Silver badge

          Re: SQL, however..... How would you like that, Oracle?

          It's been revised by https://www.iso.org/standard/63555.html. Go and look at that page. You can't just download the standard, you have to buy it, CHF 178. So the copy of the standard costs money. As I haven't bought it I don't know what the status of IBM's contribution is, whether it's assigned to ISO or licensed to them. But the fact that it's a standard doesn't mean it's not in copyright.

          1. Anonymous Coward
            Anonymous Coward

            Re: SQL, however..... How would you like that, Oracle?

            ISO documents are copyrighted - which doesn't mean the standard itself is copyrighted. It could be, or not. There could be patents, or not. Anyway, usually the aim of a standard is to broaden adoption, and licenses and royalties get in the way. Still, it may happen.

  4. ratfox

    8 billion dollars?

    Holy shit. This doesn't touch the Tobacco Master Settlement Agreement at $206 billions (yes), but as a payment of one company to another, it might be a record. Yet it is only 1% of Google's market cap, and about 4% of Oracle's.

    1. Doctor Syntax Silver badge

      Re: 8 billion dollars?

      "Yet it is only 1% of Google's market cap, and about 4% of Oracle's."

      So Google Alphabet is about 4 times the value of Oracle at market valuation.

      In that case here's an idea for Alphabet. Get together with one of the usual activist investor suspects and buy Oracle. Break it up and sell the bits off at a profit. Preferably in the course of this go back to court as the complainant in the original case and ask for a ruling to overturn the decision on the grounds that copying an API is fair use and a precedent to the contrary would impose overwhelming damage on the whole S/W industry.

      As the ultimate insult I wonder what CA would pay for the database business.

      1. Anonymous Coward
        Anonymous Coward

        "with one of the usual activist investor"

        They are not idiots. They want money, not silly revenges...

        As for copying APIs, it's a damage only for lazy companies unable to innovate and desperately need to piggyback on someone else's work.

        1. Doctor Syntax Silver badge

          Re: "with one of the usual activist investor"

          "As for copying APIs, it's a damage only for lazy companies unable to innovate and desperately need to piggyback on someone else's work."

          Do you write S/W? Maybe not. But if you do go and look again at your source code. Does it have includes or the like according to language? Where does the included material come from? Some of it might be your own but a lot of it will be the APIs of standard libraries that are part of the system S/W. This is how the S/W industry works. It uses standard APIs, even if, like Google did here, it reimplements the API (with, in this case, the exception of a few lines). This is the danger of this precedent - it drives a coach and horses through a basic working assumption of the entire development process.

          What work did Google save by using the Java API? Id didn't save the work of implementing a Java byte-code interpreter nor a compiler. It didn't save the work of implementing anything other than a trivial amount of the library behind the API. What's lazy about implementing all that? It didn't really save much work of devising its own API as all these APIs, C, C++, Java & the rest all follow much the same lines.

          What it did save was the need for huge numbers of non-Google, repeat non-Google, independent application developers having to learn a new API.

          But to return to my tongue in cheek proposal, one interpretation of the SCO suit was that IBM, being so much bigger than SCO, could just have taken them over to end the suit and that provoking this was the intention. Put that together with the fact that someone pointed out the disparity in size between Oracle and Alphabet and the possibility that an asset stripper might be able to make money breaking up something with as many lines of business as Oracle and you have an intriguing possibility - Alphabet could possibility save itself $8.8bn and make a bit of extra money as well.

        2. Michael Wojcik Silver badge

          Re: "with one of the usual activist investor"

          As for copying APIs, it's a damage only for lazy companies unable to innovate and desperately need to piggyback on someone else's work.

          It's a pity you posted anonymously. This is sufficiently ignorant and foolish that we know we could disregard pretty much anything else you post.

          Duplicating APIs is necessary for interpolation and emulation, to name but two of the important cases. Many operations take their names and parameterizations from outside sources (mathematical functions, descriptions of relevant algorithms, real-world names...), so corresponding APIs are "natural" and not substantially creative.

          The CAFC panel fucked up (or, more cynically, did what they're there for - their record is not encouraging). SCOTUS should reverse.

          1. Anonymous Coward
            Anonymous Coward

            "Duplicating APIs is necessary for interpolation and emulation"

            Not true. Emulation can be done at the ABI level. Don't know what you mean for "interpolation".

            While its true sin(x) may not pass the creativity test, many other APIs can.

            Duplicating API is just a simple way for obtaining unlawful advantages once someone has arrived before you.

            I would like to know how you feel if you were selling a successful library, and someone with more money than you publish a drop-in replacement at lower prices undercutting you to push you out of the market.

            1. Charles 9

              Re: "Duplicating APIs is necessary for interpolation and emulation"

              You either live with it or sell on brand and reputation. There's a term for what you describe: the aftermarket.

  5. TheVogon

    Yep I guess Larry will be going yacht shopping again.

    Now we can grab the popcorn waiting for the GDPR fines to roll in. On top of the existing EU ones.

  6. fredfs

    This is a perfect example

    of the wastefulness of the existing patent/copyright economy. We now have an economy largely based on *not* letting people use things that improve their lives. From patents, EULAs, and dongles to DRM, DMCA, ebooks, five-lobed screw heads, and geo-locking, the lucrative business today is to obstruct, not to facilitate. Even though the world is full of problems still requiring solutions, millions are instead directed toward legions of patent and copyright lawyers to prevent people from using known solutions. This is a waste of law-school resources, a waste of court time, and a waste of useful ideas.

    Limited patents and copyrights can serve some good, but this particular copyright fight is not about making sure the developers and artists get paid -- this fight is only about making sure that some multi-billionaires can become even more multi-billionaired.

    1. Anonymous Coward
      Anonymous Coward

      Re: This is a perfect example

      "When legislation can be bought and sold, the first thing to be bought are legislators." - (Too many cites)

      1. Charles 9

        Re: This is a perfect example

        EVERY man has his price.

        - Even more too many to cite.

        If you can talk to a legislator, you can bribe or blackmail him, and there are ways around any law, such as hiring siblings and spouses as lobbyists (now go ahead and try to stop them talking to each other, especially if they have children).

        1. Tom 38
          Joke

          Re: This is a perfect example

          now go ahead and try to stop them talking to each other, especially if they have children

          IMHO, that does stop a lot of the talking to each other.

          1. Charles 9

            Re: This is a perfect example

            Uh unn. Because if a child gets in trouble, neither wants to be the sole one to blame, especially if the law's involved. Makes for very bad history should divorce and child custody hearings occur.

    2. Anonymous Coward
      Anonymous Coward

      "this fight is only about making sure that some multi-billionaires can become even more"

      It's exactly that: Google could have asked a licenses, but didn't want to pay a dime while reaping billions from Android... and it really didn't lack the money to ask for a license.

  7. aberglas

    *nix is toast

    If APIs are copyright. It all comes from System V. Who owns that now, still AT&T, or maybe SCO will rise again. But a very, very dangerous principal.

    Incidentally, Oracle bought Sun for $7.8bn, so got their money back.

    1. Flocke Kroes Silver badge

      Re: *nix is toast

      The SCO group never owned Unix but that did not stop the bankruptcy trustee selling TSG's (non-existent) rights to Unix to some dupe. Novell was bought by Attachmate which was bought by Micro Focus.

      Micro Focus is selling SUSE to EQT Partners which will split it out into a new subsidiary Blitz 18-679 GmbH. I have no idea if Unix is moving with SUSE.

      1. Michael Wojcik Silver badge

        Re: *nix is toast

        I have no idea if Unix is moving with SUSE.

        AFAIK, the UNIX copyright never belonged to the SUSE business unit. I don't see any reason why they'd be transferred with it.

        The UNIX copyright is largely of historical (and, possibly, defensive) interest now. When Novell's ownership was confirmed in 2007, Novell said they weren't interested in pursuing infringement claims against anyone; I don't see any signs that ever changed, before or after the Attachmate and then MF acquisitions.

        Of course, the UNIX trademark is owned by The Open Group, just to confuse things. (Originally it was licensed to OG, but at some point apparently they acquired it outright.)

  8. karlkarl Silver badge

    So can they carry on using Java or do they now need to approach Oracle for a license?

    Incidentally they could have just avoided this whole mess going with the industry standard of C or C++ whilst having a simplified build system and cleaner API. Anyone making serious software will be using the NDK anyway. I am not so sure why Google pissed about with Java.

    Let this be a lesson to all of us. No matter how horrible C and C++ are, there is no other correct choice and probably never will be :)

  9. Hans 1
    Boffin

    Oracle vs Google facts

    Exactly 9 lines of code were lifted and Oracle agreed to zero damages for that - so any comment@rd pulling the "But Google copied code so should pay $8bn" BS can henceforth remain silent.

    The rest, the whole rest, is "imitating an API" of GPL'd code with linking exception. If Oracle pull this one through, then not only will forking become illegal, but designing emulators or re-implementing API's that someone else designed will become illegal, think samba, wine, large chunks of Powershell, MS Exchange, MS SQL Server, Oracle database ... This is beyond silly, and I am no fan of either company ...

    Bibliography:

    https://en.wikipedia.org/wiki/OpenJDK

    https://en.wikipedia.org/wiki/GPL_linking_exception

    https://www.xda-developers.com/update-on-the-oracle-versus-google-trial/

    https://www.computerworld.com/article/2504709/technology-law-regulation/oracle-agrees-to--zero--damages-in-google-lawsuit--eyes-appeal.html

    1. Anonymous Coward
      Anonymous Coward

      Re: Oracle vs Google facts

      You just forgot that Google started using Apache Harmony, which was denied a TCK license by Sun.

      It doesn't matter how the OpenJDK is licensed *now*, it matters what license requirements were at the time Google decided to make its own copy of Java.

      For the matter MS had to cease its own copy of Java because it was violating Sun license, and was not compatible with "standard" Java.

      AFAIK Samba doesn't copy Windows SMB APIs - it just implement the protocols.

      Emulator that provides a compatible ABI may not infringe copyright.

      1. Michael Wojcik Silver badge

        Re: Oracle vs Google facts

        For the matter MS had to cease its own copy of Java because it was violating Sun license, and was not compatible with "standard" Java.

        They did not "[have] to cease". They just had to rename it. And that was a trademark violation, not a copyright one.

  10. Anonymous Coward
    Anonymous Coward

    Meanwhile..

    "Public updates for Oracle Java SE 8 released after January 2019 will not be available for business, commercial or production use without a commercial license."

    It's clear that Oracle have no interest in developers or commercial development using Java. They clearly just want to licence it to death and kill it - even though we were all involved (in the early day) on APIs to improve the platform.

    1. Hans 1
      Facepalm

      Oracle Java SE 8

      OpenJDK is the Java reference platform, who still uses Oracle Java SE ?

      1. Anonymous Coward
        Anonymous Coward

        Most Windows users, since they were used to use Sun Java installer.

        1. Hans 1
          WTF?

          Most Windows users, since they were used to use Sun Java installer.

          Hmmm, ok, so, if understand correctly, you are implying that Windows users who set up commercial or production systems, basically what I refer to as Windows Cleaners and Surface Experts, are not intellectually capable of deploying an MSI or ZIP file ?

          Now, do not get me wrong, I also think Windows users are by definition useless at administrating computers, but come on, they are not that thick ... or am I too kind with the bunch ?

          Sorry for the delay, was away ...

  11. Herring`

    Not a good week for Google then

    This and that issue where if you google to find out what Trump has said or done, it returns stories that tell you what he's said or done. Bastards.

  12. Maelstorm Bronze badge

    Interoperability

    There's been some cases dealing with similar issues that have already been decided. The cases are as follows:

    https://en.wikipedia.org/wiki/Chamberlain_Group,_Inc._v._Skylink_Technologies,_Inc.

    https://en.wikipedia.org/wiki/Lexmark_International,_Inc._v._Static_Control_Components,_Inc.

    A copyrighted file is a copyrighted file regardless of what it contains. However, APIs are key for interoperability and should not be copyrightable. As someone else said, if SCOTUS rules that APIs are copyrightable, then all software development will be driven out of the country which will bring the country down. There are literally hundreds of thousands of independent software developers out there whose livelihoods are being threatened by this decision.

  13. Charles 9

    Two questions here.

    First, does a court of appeals actually possess the right to refuse to hear a case? I thought only SCOTUS had that right.

    Second, was this the court in and of itself, not just three of the appellate judges (meaning it's essentially an en banc decision and cannot be reheard in the same court because it's already been "reheard")?

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

Other stories you might like