Copyright on Computer Programs: Solicitor General Argues that APIs are Unquestionably Copyright Eligible

by Dennis Crouch

In recent years, much attention has focused on whether the output of computer software engineers is properly the subject of patent rights. Now, however, an important case is pending before the U.S. Supreme Court regarding whether computer programs are protectable under copyright.  Here, the particular issues involve copyright protection over program interface (API) function calls that allow programs to communicate with one another.

Google v. Oracle (awaiting writ of certiorari).

When Google developed the API-toolkit for Android, it wanted to use Java-like functionality, but didn’t want to pay the license fee. So, rather than copying the Java code, the company had its engineers re-code the functionality.  Because copyright doesn’t cover functionality, this approach works to avoid copyright infringement. The one caveat was that Google did not want to force developers learn a whole new toolkit of functional calls and so the company copied the set of more than 6,000 function calls.  This approach allows Google to free-ride off of the popularity of Java.  As I wrote earlier:

As an example, Google used the Java method header “java.lang.Math.max(a,b)”.  When called, the “max” function returns the greater of the two inputs.

In considering the case, the Federal Circuit ruled that the Java API taxonomy was copyrightable — rejecting the idea/expression merger doctrine since there are many other ways that functionally equivalent method-calls could have been constructed besides those found in Java.  The court wrote: “merger cannot bar copyright protection for any lines of declaring source code unless Sun/Oracle had only one way, or a limited number of ways, to write them.”

The petition for writ of certiori to the Supreme Court asks the following question:

Whether Section 102(b) [of the Copyright Act] precludes copyright protection for original software code that defines and organizes a set of functions that are useful in writing computer programs.

In the most recent filing in the case, the Solicitor General has suggested that the court not take the case – because it was correctly decided by the Federal Circuit.  For the SG, computer programs are unquestionably copyrightable, including the API function calls at issue here.  Rather than being a question of copyrightability, the SG suggest that Google’s best argument is fair use — although the SG does not offer an opinion of whether that is a winning argument.

[Read the New SG Brief: SGBriefGoogleOracle]

Petitioner contends, however, that even if the declaring code is an “original work[] of authorship” under Section 102(a), it is not entitled to copyright protection because it constitutes a “method of operation” or “system” within the meaning of Section 102(b). That argument is incorrect. . . . Section 102(b) is not a limitation on what kinds of expressive works may be protected by a copyright. Rather, it is a limitation on how broadly the copyright extends. Although a book on how to build a bicycle may be eligible for copyright protection, that copyright does not include any exclusive right to practice the bicycle-building method that the book explains; nor can the author prevent another person from writing a better book with a clearer explanation of the same process.

In years past, the Supreme Court has often followed the recommendation of the SG in deciding whether to grant petitions for writ of certiorari.  However, this particular brief does not wrestle with the copyright issues in a straight way, but rather appears to argue in favor of a politically chosen conclusion. In my mind, this suggests that the court should give less weight to the brief than may have been expected apriori.

158 thoughts on “Copyright on Computer Programs: Solicitor General Argues that APIs are Unquestionably Copyright Eligible

  1. 18

    Software sometimes has “undocumented” interfaces to application programs. These interfaces are not necessarily secret. The fact that they are labeled as “undocumented” only means that users can not rely on them to be supported in future releases, and, perhaps more importantly, cannot complain about malfunction of such interfaces.

    In contrast, a true application program interface is not just a technical thing, but implies a stable presence. Which is most relevant if the interface ise used by others. Precluding other to use these interfaces by means of copyright hence seems inconsistent. is the only issue an attempt to make some money? Remember that APIs are not naturally valuable. Some vendors are even prepared to subsidize competortir for using their interfaces, since that attributes to their “network value”.

  2. 17

    It is ruight that computer programs are copyright protected, since programs are often read by programmers (=human beings). Computers can execute badly written programs, but programmers cannot read them. That makes computer programs genuine copyrighted subject-matter.

    Typically in copyright, the dichotomy is between the story and the way it is told. The story itself is not copyright protected, it is only the way it is told, and others are allowed to tell the same story in treir own words.

    For APIs, the issue is not how hey are described, they are “the story” themselves. Porhibiting APIs from being used by others prohibits competition: APIs typically are used to allow third-party software. Prohibiting competition is bad for the consumer – no explanation needed.

    It is a mistake to consider copyright a “catch all” law to cover any information good. Besides, this violates the freedom of speech. Copyright does not violate the freedom of speech since it only protects the substance not the contents, and that limitation should be respected.

  3. 16

    Can you believe it? — another public rights case divides the Supreme Court. 6-3 Wellness Int. Network v. Sharif, May 26, 2015. link to scotusblog.com

    The majority held that one can waive their right to a trial in an Article III court just like one can waive their right to a trial by jury.

    Roberts, Scalia and Thomas disagreed, and strongly so. The first two believed that that an action to have Sharif’s trust included in his estate was not a Stern claim as it did not involve an action against a third party.

    Thomas, agreed with this, but. writing separately, Thomas forcefully argued that non Article III “judges” did not have the power to issue judgments, issue writs or the like. He questioned Congress’s authority to empower bankruptcy judges to act like they were judges when they were not. Regardless of whether Sharif had waived his right to object, Thomas did not find congress had the authority to assign to Article I tribunals the authority to decide cases and issue judgments within the exclusive authority of the courts when core private rights are involved.

    IPRs, meet the Supreme Court.

        1. 16.1.1.1

          Nice non sequitur Ned.

          It is not a matter of why such are written, as it (clearly) is the weight that you give to them.

        2. 16.1.1.2

          …but to answer your question directly, they DO serve a good function of highlighting what is NOT the law.

          Sort of like 3, the new 4, is still not 5.

    1. 16.2

      Tell me Mr. Heller, are patent appeals within the original jurisdiction of the Supreme Court? What does the answer in the negative portend to the Congressional power, duly appointed in the Constitution, to prevent the Supreme Court from hearing that subject matter (as long as some Article III court has jurisdiction to satisfy Marbury)?

      I do not think that you have thought all the way through this issue, good sir. I think that you stop at your altar of worship of the Royal Nine, and critical thinking goes out the window, with you embracing the notion that O’Malley explicitly warns against: the judiciary violating the separation of powers and invading the sanctioned domain of statutory law. How much of the disarray caused by the Court can truly (in an inte11ectually honest manner) be traced to a real constitutional issue and how much is merely the Court sticking its finger into wax noses?

      1. 16.2.1

        Anon, regarding the Supreme Court and its jurisdiction, it has original jurisdiction in certain cases and appellate jurisdiction in others. Congress some time ago removed the right of litigants dissatisfied with a judgment in a lower court of any right to actually appeal to the Supreme Court. Rather they can petition, and it is the Supreme Court that decides whether to take the case.

        Now you ask me whether or not Congress has the power to abrogate all review by the Supreme Court of patent cases. The answer is no. The constitutional power regarding cases and controversies remains in the Supreme Court such that any attempt by Congress to totally abrogate Supreme Court jurisdiction over cases and controversies for which it does not have original jurisdiction but only appellate jurisdiction would be unconstitutional.

        1. 16.2.1.1

          The petition mechanism is a non sequitur Ned.

          Try to answer my post directly and with a modicum of inte11ectual honesty (even as you can see the necessary logical conclusion – and contemplate how the O’Malley article still calls out your Royal Nine over-fealty).

        2. 16.2.1.2

          Your second paragraph misapplies Marbury, as it is Article III review – NOT Supreme review that must be maintained for the checks and balances of that decision.

  4. 14

    OT but related by party: on May 19 the government filed its amicus brief (requested by the Court) in Google v. Vederi, where the petition for cert is pending (conference scheduled June 18 2015). See: link to scotusblog.com

    The issue (as framed by Google in their cert petition):

    Issue: Whether, when an applicant for a patent amends a claim to overcome the Patent and Trademark Office’s earlier disallowance of the claim, a court should (i) presume that the amendment narrowed the claim and strictly construe the amended claim language against the applicant, as this Court has held; or (ii) presume that the claim scope remained the same and require that any narrowing be clear and unmistakable, as the Federal Circuit has held

    In a nutshell, the government’s argument can be summarized as follows: This Court has not adopted any categorical rule that all ambiguities in amended claim language must be resolved by giving the amended claim its narrowest plausible construction. Nor would such a rule be a sensible approach to the interpretation of amended claims. Rather, the Court has held only that an amended claim must be interpreted narrowly enough to exclude any subject matter that the amendment was intended to disclaim, see, e.g., Exhibit Supply Co. v. Ace Patents Corp., 315 U.S. 126, 137 (1942), and the Federal Circuit has followed that guidance. (page 11 of the brief)

    Previous discussion of this case and the junky claims being asserted by Vederi may be found here: link to patentlyo.com

  5. 13

    As I wrote earlier, the idea that the name of the function isn’t functional is ridiculous. Just as the word “Copy” is merged because it describes the function, the phrase Math.max should be merged because it describes the function. It’s not expressive of anything other than the functionality.

    The fed cir is completely ridiculous when they say that “copy” is merged but “math.max” isn’t. “Make.same.as” would also describe the copy function, but that didn’t stop it from being a merger.

    The holding in this case, as with so many others by the fed cir, is backwards – working from a result they want and trying to justify it, rather than simply applying the law.

    I will say this though – the commented-out dialogue might be sufficient to be expressive and subject to copyright. I doubt it though, usually the comments only say “This function takes inputs a,b, and does xyz to them” which is again a mere statement of functionality; statements of fact aren’t expressive either.

    This was a terrible ruling that has a decent chance of standing now because the Supreme Court knows they aren’t knowledgeable in the area and the SG says no.

    1. 13.1

      Random, “Stand?” You gotta be kidding. The Supreme Court is going to be all over this one.

  6. 12

    Re: TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23 (2001), this Supreme Court case was brought up below as an example of U.S. courts rejecting the proposition that if something has “aspects of functionality”, then only patent protection should apply. In TrafFix, however, the Supreme Court did find that the functionality of the design at issue precluded trademark protection.

    From Wikipedia: [TrafFix] was a United States Supreme Court decision in the area of trademark law, holding that a functional design could not be trademarked, and that a patented design was presumed to be functional….

    The Court, in a unanimous opinion by Justice Anthony Kennedy, held that there can be no trademark protection for something that is functional because that would work as a detriment to competitors based on something other than reputation, which is the key consideration in trademark law….

    Here, Justice Kennedy said, the design was clearly functional, and the plaintiff could not carry the burden of proving otherwise because the very characteristic that is sought to be protected by trademark is the one whose functionality was previously sought to be covered by patent.

    Question: does Oracle recycle (copy) it’s proprietary Java headers when it revises its own code for the associated functions? Or does Oracle “express” itself with new “expressive” (as opposed to informative) headers?

    In a relatively fresh albeit non-precedential case (Moldex-Metric, Inc. v. McKeon Products, Inc.) discussed here: link to us.practicallaw.com, the Ninth Circuit Court of Appeals rejected a defendant’s reliance on TrafFix in favor of an ealier Supreme Court case, Qualitex Co. v. Jacobson Prods. Co., 514 U.S. 159, 165 (1995).

    From the link: Qualitex’s essentiality standard requires that, for a trademark or trade dress’s color to be considered functional, the color must: be essential to the use or purpose of the article; affect the cost or quality of the article; or put competitors at a significant non-reputation-related disadvantage if granted exclusive trademark use.

    Note the last prong of the test is another reminder that the Supreme Court always keeps an eye on the public’s paramount interest in the proper use of IP laws.

    Strictly as an aside here, an issue that confronts plaintiffs trying to protect data and information with patents or copyrights is that all information worth protecting has “aspects of functionality” (e.g., all information is informative). Any judicial test relying on the existence of mere “functionality” to protect otherwise unprotectable information is therefore deeply flawed from the get-go.

    1. 12.1

      Your views – given your admissions against interests in knowing and understanding controlling law regarding writing having patentable weight – in pertinent conditions – are deeply flawed from the get-go.

      (you are also conflating different aspects in the Traffix case)

      1. 12.1.1

        Your views – given your adm

        *click*

        Billy’s already whiffing and it’s only Monday.

  7. 11

    Of background interest to the present cert petition, Lotus Development Corporation v. Borland International, Inc., 516 U.S. 233 (1996) arguably established that copyright does not extend to the text or layout of a program’s menus. [The Sup. Ct. 4 to 4 tie vote left the 1st Cir. decision intact.]
    Per a Wikipedia summary, Borland’s spreadsheet product, Quattro Pro, had a compatibility mode in which its menu imitated that of Lotus 1-2-3, a competing product. None of the source code or machine code that generated the menus was copied, but the names of the commands and the organization of those commands into a hierarchy were virtually identical. Quattro Pro also contained a feature called “Key Reader”, which allowed it to execute Lotus 1-2-3 keyboard macros. In order to support this feature, Quattro Pro’s code contained a copy of Lotus’s menu hierarchy in which each command was represented by its first letter instead of its entire name.

    1. 11.1

      Thanks for the cite, Paul.

      The Sup. Ct. 4 to 4 tie vote left the 1st Cir. decision intact.

      To be clear, they refused cert on the 1st Circuit decision (Stevens recusing). Seems to me the Roberts court would be at least as likely to take the case and ten times more likely to affirm the holding.

      Borland CEO Philippe Kahn took the case to the software development community arguing that Lotus’s position would stifle innovation and damage the future of software development. The vast majority of the software development community supported Borland’s position.

      “They vast majority of the software development community is just a bunch of copyists who don’t understand law or technology!” <— sound familiar?

      1. 11.1.1

        RE: Lotus Development Corporation v. Borland International, Inc.

        Nope, I checked: “The Supreme Court affirmed the First Circuit without opinion in an equally divided, per curiam decision.”
        It was not just a cert denial.

        1. 11.1.1.1

          Thanks for fact checking, Paul. Copyrightable APIs would be terrible for the industry.

          There is a feeling (not without merit) that Google is getting an “unfair” free-ride off Java’s popularity (e.g., Dennis says as much in his article). Is there a case in which 102(b) was used to deny copyright protection when there was also a similar feeling that the defendant was getting an “unfair” free-ride?

          1. 11.1.1.1.1

            Genghis: Is there a case in which 102(b) was used to deny copyright protection when there was also a similar feeling that the defendant was getting an “unfair” free-ride?

            The difficulty with this question, Genghis, is that in pretty much every case where 102(b) was used to deny copyright (or patent) protection, there is someone somewhere who feels that the decision is “unfair” and somebody is “free-riding”.

            I also note that Dennis did not say that the so-called “free-riding” here was “unfair”. I don’t know what his opinion is on that issue.

            I do know that people free-ride off of all kinds of stuff that I produce and there’s nothing unfair about it. Likewise, I free-ride off other people’s work and there’s nothing unfair about that either. Welcome to the planet where you don’t get 100% control over everything just because you worked on it. It’s actually a pretty great set-up.

            1. 11.1.1.1.1.1

              people free-ride off of all kinds of stuff that I produce

              Delusions of grandeur from the one who is a legend in his own mind…

              Your “short script” retreads do not count as “producing.”

              That being said, the patent system is all about copying others – after the limited time of protection has run its due course.

              1. 11.1.1.1.1.1.1

                Delusions of grandeur

                None whatsoever. And I wasn’t referring to my writing here which is dedicated to the public in any event. Free rides for everybody.

        2. 11.1.1.2

          PM: It was not just a cert denial.

          Indeed it wasn’t. “Affirmed per curium.” Says so right there in the “opinion.” 😉 link to law.resource.org

          Somehow the line in the wiki summary “Lotus’s petition for a rehearing by the full court was denied” made it through the gray matter but the “affirmed per curium” part didn’t. My bad.

    2. 11.2

      Thanks for the informative post, Paul. I think the parallel between the text and the layout of a menu and an API is quite compelling. I would have supported Borland’s position in that case just as I support Google’s position in the current case.

    3. 11.3

      Thanks Paul.

      Copyright is intended to protect particular works. It is not intended to protect functional interfaces. Thus an alphabet is functional with respect to a language.

      Computer instructions are functional with respect to computers. The names of computer instructions are functional with respect to assemblers and compilers. Thus one controlling the names of computer instructions controls effective access to programs for computers and therefore to particular computers.

      Lotus v. Borland was a high watermark of IBM’s attempt to protect interfaces with copyright. The commands were used functionally by the program itself to operate the program. Manual operation of the program had a menu of commands that were accessed by pressing the 1st letter of the command. Thus if one wanted to recompile Lotus programs into Borland’s equivalent, Borland had to use the command structure of Lotus. Similarly the 1st letters of the commands for manual operation. Both for compatibility purposes.

      Thus controlling the command structure was controlling not only the command structure, but the underlying program, and was preventing competition in equivalent software.

      If such a command structure or API is copyrightable, then the idea that copyright cannot be extended to functional systems is an idea that has lost all content and meaning.

      1. 11.3.1

        Thanks for the explanations Ned and Warhawke. Do you think one could also argue that copying the API is de-minimis?

        1. 11.3.1.2

          Genghis, not in this case. If an API is copyrightable, there is no excuse for wholesale copy.

          1. 11.3.1.2.1

            The function headers are typically just a small fraction of a software library. The de minimus argument I was thinking of was along the lines of: we only copied 6000 “words” from a software library that includes millions of “words.”

            1. 11.3.1.2.1.1

              Genghis, we are talking about the headers are we not? The question is whether these headers are copyrightable. If they are copyrightable, there is no excuse for wholesale copying of the headers. Period. Fair use in no way justifies wholesale copying.

              1. 11.3.1.2.1.1.1

                I copy the table of contents (and the chapter titles) wholesale but the rest of my 10,000 page tome is different. Regardless of whether or not the table of contents is itself copyrightable, isn’t there a de-minimus argument here?

                1. Genghis,

                  May I recommend that you recognize (as does the courts), that even a single line (from a book) may be viewed to be copyright infringement.

                  The test is not one of mere size comparison, but includes as well “impact.”

                  (A bit dry, but I also recommend reading straight through Nimmer to obtain a working knowledge)

                2. Genghis, there is NO question that Google did not copy the functional code. The question is whether the headers are copyrightable.

                  IF they are, then Google is a copyright infringer. The can be no serious question of fair use.

                3. Alright, no de minimus argument here (or anywhere for that matter – it seems the “de minimus doctrine” is dead for all practical purposes).

              2. 11.3.1.2.1.1.2

                First off, Java doesn’t use headers. The Java API is based on a hierarchical organization of methods (functions) into packages and namespaces.

                I haven’t fully thought through the issues at hand yet and I’m unaware of all the rights that Sun/Oracle may have, but my initial reaction is that a software API is a tool to facilitate interaction with a system, not a method of expression. For the Java API in question, Sun said “here is a bunch of software functionality (the built-in Java libraries), and here is how programmers should interact with it (the API for those libraries).” What Google did was say “we have a bunch of Java programmers who already know how to interact with that API. We’ve got our own platform and functionality but we want those programmers to be able to interact with our functionality in the same way they’re used to interacting with Java.”

                An off-the-cuff analogy: if I dial 911 from my cell phone in New York, my call is geolocated and I get the local emergency dispatcher. If I dial 911 on my cell-phone in Los Angeles, my call is geolocated and I get a different emergency dispatcher, not the people in New York.

                It’s the same interface (“dial 911”), and the same intended functionality (“make an emergency call”), but it’s a different platform (emergency cell network in NY vs. LA), and a different implementation (NY dispatchers and responders vs. LA dispatchers and responders).

                If this is a valid analogy, then the question for the Court would be equivalent to “is ‘dial 911 to make an emergency call’ protectable under copyright?” If New York had come up with “dial 911” first, and LA copied it because they liked the idea and consistency in such things is a good idea, would there be a copyright claim?

                FYI, here are the APIs in question (sticking with the example of java.lang.Math):
                Android: link to developer.android.com
                Oracle (formerly Sun):
                link to docs.oracle.com

      2. 11.3.2

        Thus an alphabet is functional with respect to a language.

        What does that mean? How does this fit with the allowed claims to Morse?

        1. 11.3.2.1

          And what does it have to do with API, which is neither an alphabet nor a language?

        2. 11.3.2.2

          Anon, your fictional belief that Morse patented his code is the problem here. We have gone over this multiple times. What Morse patented was his system of marks and spaces, not his code.

          The Supreme Court found infringement even though the infringer did not use the code of Morse. What else is there to say?

          1. 11.3.2.2.1

            system of marks and spaces, not his code.

            You keep on saying this as if it has any meaning.

            It does not.

            (ps, you have not provided an answer as to how much Morse’s “space” weighs)

            /off sardonic bemusement

            1. 11.3.2.2.1.1

              J U S T because you do not understand what Morse patented, have not read the Morse case, and do not understand what is going on, is no excuse for you to ignore that the Supreme Court held that the infringer was infringing despite using a different code because they were using marks and spaces.

              1. 11.3.2.2.1.1.1

                I am not the one broadcasting their lack of understanding Ned (that would be you).

                For some reason now (twice), you want to kick up the dust of infringement when that it is clearly not the point of the reference.

                1. Your insistence, anon, that the code was patented is at odds with the holding of the court. It is significant here that your manner of sustaining your position is to simply state that the other party is wrong without providing any substantive argument whatsoever. This has long been your notorious way of behaving on these fora.

                2. Ned – you are playing that game of merely saying something with no substantive arguement. Trying to kick up dust with tangents that just are not on point f001s no one.

                  By the way, have you weighed that “space” yet?

              2. 11.3.2.2.1.1.2

                …and you STILL haven’t told me how much the “space” of Morse’s allowed claims weighs…

                /off sardonic bemusement

                1. Doesn’t answer the question Ned (then again, you already knew that, right?)

                  Why have you not answered the question Ned? Just answer the question Ned.

                  /off sardonic bemusement

      3. 11.3.3

        was preventing competition in equivalent software.

        That was true in the Borland case. Here, arguably the only reason Google re-coded the functionality was to avoid a copyright licensing fee, which is similar to what Aereo did.

  8. 10

    If Google had copied just the function calls headers but Implemented different functionalities, there were be no case (not in the legal sense of a cause of action, in the sense of Oracle would never have sued).

    Oracle is leveraging a claim in a copyright in an interface to protect the “stuff” behind the interface. Not only should 102 apply, but it should be copyright abuse.

    1. 10.2

      If Google had copied just the function calls headers but Implemented different functionalities,

      That would defeat the purpose.

      1. 10.2.1

        Exactly.

        Part of the nuance here is that the item involved has more than one aspect. One argument that was rejected (and has been rejected in several cases such as Traffix), was that IP protections should be mutually exclusive. That is, if something has aspects of functionality, then only patent protection should apply.

        Such is not – and never has been – the case. Most rudimentary IP courses cover this. Any particular item may very well have different aspects that garner protection from the different IP laws (both federal and state – as they may apply).

  9. 9

    I have a question, why doesn’t the government explicitly write into the copyright/patent laws what compooter programs can be copywritten/patented etc so that the people may know the law with certainty?

    1. 9.1

      The editing controls are out of control.

      Again.

      Somehow pointing out that 101 has to be written in an open-ended manner because innovation cannot be known a priori is deemed “offensive”…?

      This while Malcolm’s antics continue unabated.

      Here’s a suggestion Prof: write a set of objective rules and enforce those rules universally. That way, you can avoid the perception that you favor particular viewpoints in what you allow and what you do not allow.

      1. 9.1.1

        you can avoid the perception that you favor particular viewpoints

        Because favoring particular viewpoints is the worst thing ever!

        The mere idea that Dennis might have an opinion on some subject or the other, or that certain expressions of opinion aren’t worthy of publishing … The horror! The horror!

        101 has to be written in an open-ended manner

        No, it doesn’t.

        innovation cannot be known a priori

        But general classes of endeavor in which patent-worthy “innovation” occurs can certainly be known and identified. It’s really not that difficult. In any event, Congress can always rewrite the statute if the public (as opposed to a couple incredibly short-sighted jurists completely unfamiliar with the “technology” or the issues associated with it) demands that the gates of the patent system be opened wide to some previously unanticipated and fundamentally different new class of “innovation”.

        The bottom line, as many, many people — including 6 and myself — will happily admit is that the subject matter eligibility statute we have was written by incredibly short-sighted and/or incompetent people, and to the extent it’s ever redrafted “from scratch” there’s zero chance that it’ll be as nonsensical on its face as it is now.

  10. 8

    I agree with the Federal Circuit’s reasoning behind the rejection of the merger doctrine, but I can’t help but think that their reasoning conflated “API” with “source code.” This is indicative in the dicta, “Many of Google’s arguments, and those of some amici, appear premised on the believe that copyright is not the correct legal ground upon which to protect intellectual property rights to software programs; they opine that patent protection for such programs, with its insistence on non-obviousness, and shorter terms of protection, might be more applicable and sufficient.” (Emphasis added.)

    As I see it, API is a command structure — a system or method of operation, a functional component which differs greatly from the expressive artistry that can emerge from the expression of source code. This would clearly place an API under 102(b) as a “procedure, process, system, [or] method of operation,” as Lotus Development held. Whereas source code can inherit copyright through its structure, sequence, and organization a la Computer Associates v. Altari, an API only uses structure, sequence, and organization as a functional aspect to ensure interoperability (which Google argued poorly). The A-F-C test expressly filters out any expression required for purposes of interoperability or compatibility. Thus, the organizational aspects of an API, such as Java, are inherently filtered out of a copyright application analysis in ways that organizational aspects of source code would not. Oracle initially got around this by relying on Whelan, but Whelan is exceptionally overbroad and assigns copyrightability to anything that extends beyond the mere “idea” of a program. This runs contrary to 102(b), which precludes classes beyond that of an “idea.”

    In short, the Federal Circuit should have never gotten to the merger doctrine question because an API (which is purely structurally and organizationally functional) should be precluded from copyright via 102(b). Comparatively, a “system” can be expressed in various different ways (as is often done in patent law through various descriptions of embodiments), yet it is still explicitly precluded from copyright consideration under 102(b).

    1. 8.2

      You do realize that one item may have different aspects that are protected under different IP laws (and such not limited to just patents and copyrights), right?

      Prof. – great teaching moment for you here…

    2. 8.3

      API is a command structure

      Kinda like a command module except it’s weightless and you can build it with a pencil.

      1. 8.3.1

        I want one of those weightless computers that was built with a pencil. Wow!!

          1. 8.3.1.1.1

            Not really – but I think that he was just toying with you.

            The better point of course – reminding everyone (yes, it is in the archives) of your admission against interests concerning knowing and understanding the controlling law in regards to certain manufactures of the hand of man being in writing and having patentable weight (the exceptions to the judicial doctrine of printed matter), is that you are merely dissembling (again).

            What a chump.

    3. 8.4

      Warhawke said “As I see it, API is a command structure — a system or method of operation, a functional component which differs greatly from the expressive artistry that can emerge from the expression of source code.

      I wouldn’t call an API a “system or method of operation.” An API is just a set of function names. The “method of operation” is in the implementation of an API, not in the API. But I don’t think APIs should be copyrightable for policy reasons.

  11. 7

    Some people suggest that one could (and indeed should) protect software by using copyright law. This case is proof that only those who have never written a single line of code or those who are biased against software (there could be a multitude of reasons for being biased) would say something like that.

    This case shows that trying to protect software using copyright law doesn’t work. If the Supreme Court reverses, Oracle may not be able to even protect the software API, let alone the actual code using copyright law.

    I feel that, first, the Fed. Cir. and the SCOTUS reach a pragmatic conclusion based on the facts of a case, and then, they try to figure out a way to “fit” that conclusion into the “law.” I know such outcome based decision making is looked down upon by legal scholars and purists, but it makes a lot of practical sense to me.

      1. 7.1.1

        Well, it depends…

        In the context of morality, I agree with that statement wholeheartedly. But, in engineering, the “ends” almost always justify the “means” – in fact, that’s pretty much what engineering is about.

        1. 7.1.1.1

          In the context of morality, I agree with that statement wholeheartedly.

          Whatever that means.

          This case shows that trying to protect software using copyright law doesn’t work.

          No, it doesn’t show that at all.

          This case is proof that only those who have never written a single line of code or those who are biased against software (there could be a multitude of reasons for being biased) would say something like [copyright can be used to protect software

          No, it isn’t proof of that at all.

          1. 7.1.1.1.1

            Genghis said “In the context of morality, I agree with that statement wholeheartedly..” MM said “Whatever that means.

            It is well known concept. It means that when justifying whether or not a particular act is moral, the end for which the act is being performed should not be used as the justification.

            1. 7.1.1.1.1.1

              It is well known concept.

              It’s a cliche.

              when justifying whether or not a particular act is moral, the end for which the act is being performed should not be used as the justification.

              Because you say so? Why do you believe this “wholeheartedly”?

              I’m curious

              I bet you are.

              1. 7.1.1.1.1.1.1

                Because you say so?

                Do you have another meaning?

                Why do you believe this “wholeheartedly”?

                Why not?

                I bet you are.

                I was just a little, but not any more.

                1. Why do you believe this “wholeheartedly”?

                  Why not?

                  For starters, it seems like an unnecessary and silly handicap to impose on oneself.

                  I trespassed and stole a traffic cone to save a busload of schoolchildren, including your own. Is that a moral failure? Most people wouldn’t think so, even as most people (myself included) would acknowledge without hesitation that theft is “wrong”.

                  Cliches and idioms are fun for philosophers but this one — in the phrasing you endorsed — strikes me as a blunt and not-terrifically-useful rule for guiding one’s decision-making. But to each his own.

                2. Translation: “it doesn’t match Malcolm’s Belieb system.”

                  Thus sayeth the Red Queen.

                3. Is that a moral failure?

                  No, in the example you provided the end did justify the means. But of course that wasn’t the example I was thinking of when I posted my comment. If it were, I wouldn’t have used the word “wholeheartedly.”

                  Is there an upper bound to the number of comment levels on this blog?

                4. of course that wasn’t the example I was thinking of when I posted my comment. If it were, I wouldn’t have used the word “wholeheartedly.”

                  Tell us the example you were thinking of when you used the word “wholeheartedly.”

                  Most of us, I think, would try to come up with several examples on the spectrum before writing what you wrote.

                  Have you noticed a pattern, Genghis? Live and learn …

          2. 7.1.1.1.2

            MM, in another thread you mentioned that you have a Ph.D. Can you tell me the subject? Someone also mentioned that you used to go by the name Malcolm Mooney, and I found this article on this blog about you – link to patentlyo.com.

            I’m curious – what makes you tick? Do you really enjoy posting the type of comments you post here?

            1. 7.1.1.1.2.1

              Nine years and running Malcolm has been a blight here – that’s well before I started posting and is a sentiment oft repeated (except for his little crowd of anti-patent [be it anti-software, anti-business method, or just plain anti] echoes).

            2. 7.1.1.1.2.2

              the type of comments

              Tell you what, Genghie: show everyone here that you’re an equal opporunity critic of a certain “type of comment” and I’ll start paying attention to your concerns about those “types of comment”. Until then, consider a mild descent from the “you must have never programmed a computer before” high horse. You’re only about the fiftieth commenter over the years here who has tried the “people who disagree with me don’t understand programming” gambit in one context or another. If you’ve got a great “technical” argument to make, by all means you should offer it up.

              1. 7.1.1.1.2.2.1

                I agree the “if you don’t agree with me then you don’t understand programming” comment wasn’t very polite. But the amount of impoliteness I’ve received from you is significantly more.

                As far as “equal opportunity critic” – I don’t care what you say to others or what others say to you. However, you are the only one who has been consistently impolite to me and that is why I made that comment. If someone else was consistently impolite to me, I would say the same thing to them.

                1. Genghis,

                  He is like that to any (and every) one who does not Belieb in his special Belieb system.

                  I did try to warn you.

                2. you are the only one who has been consistently impolite to me

                  I don’t see where I was impolite to you in this thread. I asked you questions and I pointed out where you weren’t being clear. I’m going to keep doing that.

                  If you put your opinions in the comment section of a blog where opinions are debated, people will debate them. You should expect questions to be asked about your comments. Try to anticipate those questions before you hit the “post’ button, particularly if you are sensitive about being questioned or corrected.

                3. MM said “I don’t see where I was impolite to you in this thread.

                  I didn’t say in this thread. I just told you how I felt.

                  MM said “Try to anticipate those questions before you hit the “post’ button, particularly if you are sensitive about being questioned or corrected.

                  I will try to anticipate. I’m not sensitive about being questioned or corrected. I’m sensitive about the tone in which it is done.

                4. MM said “Buy some kleenex.

                  I’m going to try my best to be polite to everyone. I guess you have a different approach. But, to each his own.

              2. 7.1.1.1.2.2.2

                …as opposed to Malcolm and his total lack of technical (and legal) arguments.

                Go figure.

      2. 7.1.2

        So,

        If I am hungry and I make myself a PB&J sandwich and eat that sandwich to satisfy my hunger, making the sandwich was not justified?

    1. 7.2

      I wrote quite a bit of code in my time and wholeheartedly believe code should be protected by copyright. Remember, this simply means someone can’t copy your code; they can make something with the same function as long as they do not copy your code (even if the code looks similar).

      Here, Google did copy much of the “code”.

      1. 7.2.1

        Let’s wait and see what the SCOTUS thinks.

        I don’t think APIs should be protected – I think APIs should be standardized and freely available. However, I think the actual algorithm that implements the functionality of an API call should be protected by a patent.

        Let’s say we have a standardized API function called “speechToText.” I don’t want that to be protected. I want applications to use that API function freely. However, I want inventors to improve the current speech recognition technology, and one way to encourage that is to protect non-obvious software improvements through patents that have claims that Goldilocks would approve (not too narrow, not too broad, just right).

        1. 7.2.1.1

          In that case, you should be posting that the USSC should reverse in full.

          Google did not copy the code inside the APIs. They copied only the name and structure.

          In other words, Java’s MAX might have been…

          public int Math.Max(int a, int b)
          {
          if (b > a)
          return b;
          else
          return a;
          }

          Where as, Android Max might have been…

          public int Math.Max (a, b)
          {
          if (a > b)
          return a;
          if (b > a)
          return b;
          return a;
          }

          These two functions are operationally the same. However, they are not written the same.

          Ironically enough, both will reduce down to the exact same bit of machine code (since they are functionally equivalent).

          Oracle is not trying to use Copyright to protect how they implemented Math.Max(a, b). They are trying to copyright Math.Max(a, b).

          BTW: I completely and totally think that software should be covered by copyright (I write code for a living, and I utilize it daily, including putting in copyright notices in my code).

          For you to say that I do not understand coding because I think copyright works for software is asinine at the least, and extremely insulting. Also from your comment, you obviously havent’ even bothered to read the case. I’ve been following it for over 5 years. Go read exactly what is going on , then come back and comment.

          1. 7.2.1.1.1

            For you to say that I do not understand coding because I think copyright works for software is asinine at the least, and extremely insulting.

            OK, I take that back.

            BTW: I completely and totally think that software should be covered by copyright (I write code for a living, and I utilize it daily, including putting in copyright notices in my code).

            Let me explain. I have no problem with software being protected by copyright law. I have a problem with the following argument: “copyright protection for software is sufficient; no type of patent protection for software is required.” It is in that context I used the word “should.”

            Oracle is not trying to use Copyright to protect how they implemented Math.Max(a, b). They are trying to copyright Math.Max(a, b).

            Yes, I understand the difference. I don’t think Oracle should should be allowed to copyright a list of the function names. A list of function names is not software code which, as I said above, should be copyrightable.

            1. 7.2.1.1.1.1

              Genghis: I have a problem with the following argument: “copyright protection for software is sufficient; no type of patent protection for software is required.”

              That’s not an argument. It’s an opinion.

              Exactly why do you have a problem with that opinion? You insulted people who disagree with you but never provided any explanation except for some incomprehensible “analysis” of the CAFC’s Oracle case.

              1. 7.2.1.1.1.1.1

                That’s not an argument. It’s an opinion.

                It is a conclusion that arguments I have read on this message board seem to reach.

                Exactly why do you have a problem with that opinion?

                I want to protect a non-obvious algorithm that I created. I understand how patents can be used to protect such an algorithm. I have not seen a convincing explanation of how copyright law will protect a competitor from rewriting a non-obvious algorithm that I created.

                You insulted people who disagree with you but never provided any explanation except for some incomprehensible “analysis” of the CAFC’s Oracle case.

                I said sorry, and I will be more careful in the future with my tone.

                What does insulting people have to do with providing an analysis of a case? Are you suggesting it is OK to insult if you provide an analysis? In my opinion it is not OK to insult regardless of whether or not you provide an analysis.

                1. That’s just it Genghis – you are trying to be too nice to someone who WILL insult anyone with an opinion or an argument that does not align with his Red Queen edicts.

                  Your apologies here – especially to Malcolm – are like throwing pearls to the swine.

                2. I want to protect a non-obvious algorithm that I created. I understand how patents can be used to protect such an algorithm.

                  Algorithms are ineligible subject matter.

                  Whatever you think you “understand” I’m going to go waaay out on a limb and guess that you are either (1) wrong or (2) you haven’t thought about it very deeply.

                  Mayb you can tell us how “patents can be used to protect algorithms” in a way that doesn’t turn the patent system into a joke (the ad hoc make-it-up-as-we-go gambit hasn’t been working very well so far for you and your fellow travelers).

                  For starters, define exactly what you mean by “algorithm”. Then explain how the PTO is supposed to compare “algorithms” in the prior art to your awesome “new” “algorithm”. Is there some searchable database of “algorithms” that you are aware of? What form must your “algorithm” take to be described with sufficient definiteness? Be specific and provide some examples.

                  Show us how this is supposed to work. I’ve been waiting years to see a practical description of the “algorithm patent system” that isn’t just an end-around the existing limits that were put in place by some near-illiterate, backward thinking clowns back in the 50s.

                3. Algorithms are ineligible subject matter.

                  I know that.

                  I’m going to go waaay out on a limb

                  You are right.

                4. that isn’t just an end-around the existing limits that were put in place by some near-illiterate, backward thinking clowns back in the 50s.

                  That is your opinion.

                5. Genghis: That is your opinion.

                  I don’t have an opinion on your proposed “algorithm patenting system”, Genghis, because you haven’t provided one single detail about how it is supposed to work. Wasn’t it you who just finished raving about how “the ends never justify the means”? It’s all about “the means”, isn’t it, Genghis? Share your proposed “means” with everyone.

                  I won’t hold my breath waiting for you, of course. I’ve been watching the pro-software patent crowd for too long to expect anything resembling a well-considered answer. They seem perfectly content to simply stamp their feet and whine as their house of cards crumbles. Any idea why that is, Genghis?

          2. 7.2.1.1.2

            MDT – BTW, do you agree/disagree with the following statement: “copyright protection for software is sufficient; no type of patent protection for software is required.”

            1. 7.2.1.1.2.1

              Apparently my response was moderated out of existence. Not sure why, I didn’t curse, only stated an opinion (and wasn’t one of those ‘it is ebil!’ things).

              In an attempt to not get moderated, I’ll just say, I am not kneejerk ‘no’ to patents on software, but I don’t think most software patents granted today that I read about making it to trial are worthwhile. Although my position is that of a non-patent attorney, it’s as a person of skill in the art.

              1. 7.2.1.1.2.1.1

                Careful there MDT, as PHOSITA is a term of art and does NOT mean what you seem to imply: the political/philosophical viewpoint of any average group of practitioners. It is an easy “Slashdot/Techdirt” type of mistake to make (and one often purposefully made by others to obscure the conversation).

                1. You will note I specifically said I was ‘of skill in the art’, not, PHOSITA. I very specifically chose those words. I am someone of skill in the art, and have the degrees and nearly 20 years on my resume to prove it.

                  There is a difference between PHOSITA and someone ‘of skill in the art’. Which is why I chose the wording I chose. One is a construct invented by people not skilled in the art, and has meanings and baggage that have nothing to do with the skill in the art. The other words indicate someone who has skill in the art.

                  Note I didn’t say ‘skill in the tech’, although that could have been used as well. Much like writing, programming is both a science and an art. The science of programming is the type of work Google and Sun did (creating a programming language). The art of programming is using the structure and words of the language to create something.

                2. One is a construct invented by people not skilled in the art, and has meanings and baggage that have nothing to do with the skill in the art.

                  This is especially true when it comes to patent claims designed to protect software. As it presently stands, precisely zero skill in the art is needed to conceive, write and obtain patent claims that turn highly skilled people into infringers.

                3. MDT:
                  PHOSITA and POSITA are legally equivalent. Any use of “person [having/with/of] skill in the art” is quite obviously trading on the term of art bearing nearly the same name. Your semantic argument isn’t fooling anyone here. I take it as a software developers attempt to sound “legalese-y” but its just poorly advised. Just call yourself a skilled programmer and be done with it.

              2. 7.2.1.1.2.1.2

                I am not kneejerk ‘no’ to patents on software,

                Neither am I. Nor am I a kneejerk “yes” to patents on software.

                I don’t think most software patents granted today that I read about making it to trial are worthwhile

                I agree.

        2. 7.2.1.2

          Genghis: I want inventors to improve the current speech recognition technology, and one way to encourage that is to protect non-obvious software improvements through patents that have claims that Goldilocks would approve (not too narrow, not too broad, just right).

          What’s the difference between copyright protection that is ‘just right’ and patent protection that is ‘just right’?

          Unless you provide the details of what “just right” means, you really aren’t saying much of anything at all.

          1. 7.2.1.2.1

            Unless you provide the details of what “just right” means, you really aren’t saying much of anything at all.

            Fair enough. I don’t think I can provide a complete definition. This is the best I can do: “just right protection” protects, for a certain time period, non-obvious algorithms (all software and/or hardware implementations), and only protects, for a certain time period, a particular software and/or hardware implementation of an obvious algorithm.

          2. 7.2.1.2.2

            Unless you provide the details of what “just right” means, you really aren’t saying much of anything at all.

            The exact same thing is said of the Supreme Court – and yet you glorify in their nonsense.

            Go figure.

            1. 7.2.1.2.2.1

              The exact same thing is said of the Supreme Court

              No doubt such things are “said.” But we all know that you and your cohorts have an unrelenting tendency to equate “no guidance” with “guidance that isn’t favorable to my favorite patents.”

              The Supreme Court has provided reams more guidance on subject matter eligibility than Genghis or you have. Genghis, at least, has an excuse (in theory).

    2. 7.4

      Genghis, copyright in software exists only to protect the copying of expression in a particular work. It cannot be used to protect the functionality of a system beyond the particular work.

      Whether as fair use or copyrightable subject matter, the API expression cannot be protected under the copyright laws so as to protect the API interface functionality.

      I assume a novel API can be protected under the patent laws. But that is not what Sun and Oracle wanted to protect – some novel functionality. They want to protect the specific interface. The question then becomes whether the specific interface can be protected under state unfair competition laws or whether the area so dominated by patent and copyright laws and policy as to prohibit such protection. See the US Supreme Court case of Sears, Compco and Bonito. Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225, 84 S. Ct. 784, 11 L. Ed. 2d 661 (1964). Compco Corp. v. Day-Brite Lighting, Inc., 376 U.S. 234, 84 S. Ct. 779, 11 L. Ed. 2d 669 (1964). Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 109 S. Ct. 971, 103 L. Ed. 2d 118 (1989).

      I think the answer to the hypothetical question posed here is very clear considering the 3 cases.

      1. 7.4.1

        Genghis, copyright in software exists only to protect the copying of expression in a particular work. It cannot be used to protect the functionality of a system beyond the particular work.

        And that’s why the protection that copyright offers me for software is of very limited use.

        Whether as fair use or copyrightable subject matter, the API expression cannot be protected under the copyright laws so as to protect the API interface functionality.

        I agree.

        However, I want software companies to invest in projects such as the one that resulted in Java, and if there was a way they could protect their investment, I think it would make for good policy. I don’t think there is a way to do that under the current law.

        1. 7.4.1.1

          that’s why the protection that copyright offers me for software is of very limited use.

          Because there’s not enough people out there for you to sue? Because the software you write really isn’t that good and nobody will bother copying it, much less buy it from you for the price that you are asking? You deserve to be rich because you know how to write software?

          Please tell us more.

          1. 7.4.1.1.1

            Because there’s not enough people out there for you to sue? Because the software you write really isn’t that good and nobody will bother copying it, much less buy it from you for the price that you are asking? You deserve to be rich because you know how to write software?

            No.

            Please tell us more.

            No.

        2. 7.4.1.2

          I want software companies to invest in projects such as the one that resulted in Java

          Define “projects such as the one that resulted in Java.”

  12. 6

    Fair use vs. copyrightable subject matter?

    It seems that when any use is fair, the subject matter should not be copyrighted in the 1st place.

    The whole point of extending copyright to software is to protect the expressive elements from copying but not to protect any functionality. But where the interface is involved, one is protecting not simply the expressive elements of the interface, but the interface itself, which is functional.

    I would think that fair use is not the right answer here, and that Supreme Court needs to clarify the law.

    1. 6.1

      Gee Ned, what other area of inte11ectual property law is expressly built to protect the functional aspects of man’s manufactures?

      (and yes, you should know this already)

      1. 6.1.1

        Trade secret laws?

        Perhaps Sun should have scrambled its API so that actual users needed to get a code from Sun to continue to operate. That way, Sun could pull the plug for violation of it Terms and Conditions.

        1. 6.1.1.1

          With all due respect Ned, your ability to be obtuse rivals that of Warden Norton.

          It just pains you beyond belief to not give the clear and ready answer, doesn’t it?

          1. 6.1.1.1.1

            It just pains you beyond belief to not give the clear and ready answer

            Coming from Mr. Innuendo himself that’s pretty rich.

        2. 6.1.1.2

          Perhaps Sun should have scrambled its API so that actual users needed to get a code from Sun to continue to operate. That way, Sun could pull the plug for violation of it Terms and Conditions.

          Are you saying that Oracle/Sun could sue Google under trade secret law even if Google recreates the unscrambled API from scratch? If so, then I guess theoretically scrambling would work but the whole point of Java was to distribute it widely so that others could develop “write once run everywhere” applications on it. So even if your approach works theoretically, it is at odds with the business purpose.

          1. 6.1.1.2.1

            Genghis: the whole point of Java was to distribute it widely

            LOL

            This is where the jokes start to write themselves.

  13. 5

    I find this case more akin to a completely new dictionary with made up (very complex) words, and Google copied those words.

    Right. So now you know why I asked the question. Here it is again:

    Can you describe a language in a book, e.g., a dictionary, and then successfully sue people who use the words you defined? Has that ever happened?”

    Imagine I’m compiling my own dictionary of, say, “recently coined words” and I want to include in that dictionary some of the words that you coined. I’m going to modify the definitions (not copy them).

    Google used “java.lang.Math.max(a,b)” when they could have used something completely different, such as android.mathematics.determineMaximum(x,y)

    Just for the record: those two phrases are not “completely different.” Not even close. But the problem of defining what “completely different” means in the context of a description of a logic process is a problem worth highlighting (again).

    1. 5.1

      MM

      Does your line of questioning broaden generally to:

      Under what circumstances and on what principle are “lists” (be it names, addresses, functions, words, etc.) protectable under copyright?

      1. 5.1.1

        Anon2,

        Even then the copyright would only extend to the list itself, not the contents of the list, right?

        1. 5.1.1.1

          What would be protected by copyright extending to “the list itself” but not to “the contents of the list”?

          I take the list to be made of content which is in the form of a list… without content the list does not exist, without a list, by definition there is no content.

          1. 5.1.1.1.1

            It means you could protect the particular way you organized the contents of the list, for instance.

            Take a telephone book wherein the telephone book organizes people’s numbers by their height. You couldn’t reproduce such a book.

            However, you’d be completely free to take the numbers from the book and organize them in some other way, say, in alphabetical order.

            1. 5.1.1.1.1.1

              Go,

              A straight up question that I am not sure that you have ever answered:

              Can you copyright math?

              (and of course, to dissuade the dissemblers, this is not a question of copyrighting a math book – let’s not move the goalposts)

              1. 5.1.1.1.1.1.1

                Anon, im a little unsure what you mean by ‘math’ in this context. Can you provide an example of what youre trying to protect?

                1. I’m not trying to protect anything.

                  It’s a simple question – you are trying too hard to make this too difficult.

                  Feel free to understand math for what math is. You do not need any special ‘context’ for that.

    2. 5.2

      People invent new languages all the time, for example Klingon, Wookie or Tsolyani. The creator of a new language is clearly the creator of an original work and holds the copyright.

      API’s have always been different because API’s are used to access functionality. When you copyright an API you require each person who develops the same functionality to find a different way of expressing how the functionality is accessed. There are only a few words in English which mean ‘max’ (max, Max, maximum, Maximum). What could a developer do if those 4 words are already copyrighted in some other API?

      If instead you say “Thats ridiculous, anyone can use the word ‘max’ “, then where do you draw the line? Can you reuse ‘min’ also? What if you copy 10 names? What if you copy 6000?

      1. 5.2.1

        The creator of a new language is clearly the creator of an original work and holds the copyright.

        Exactly what does the copyright protect is the question. The actual work? Sure. The use of the “new language” itself? Highly dubious.

        1. 5.2.1.1

          I am certain that the movie and music industry has a thing to say about “use,” Malcolm.

  14. 4

    As I recall, the allowed claims of Morse

    There was zero analysis in Morse of the patent eligibility of a “new” language so I really could care less what was “allowed.”

    But go right ahead and build a shiny citadel on that foundation if it pleases you. Then come after me when I communicate something using your “new language” (“beep!” i replied when asked whether I used an “i” or an “e” to spell my name …). Together we’ll watch your claims be immolated. Bring some marshmallows!

    1. 4.1

      MM, the code itself was not patented. It was the system of physical marks and spaces that was.

      Anon is fully aware of this, or at least, should be as we have gone over it before several times.

  15. 3

    I’m curious how the SG might have chosen to support such a conclusion when nearly the whole industry is opposed and it contradicts universal practice established for fifty years.

    If the CAFC decision holds, possessing the computer you are reading this on makes you a felon many times over because nobody in the industry ever thought APIs they were using could have been under copyright. Clean room techniques used to make all modern PCs were devised by industry lawyers specifically to take advantage of that fact.

    So perhaps the SG really believes the conclusion he writes are the text of the law against all political motives. And maybe a few lobbyists and campaign contributions don’t make a presidential administration a tool of Facebook, IBM, or Google.

    1. 3.1

      So perhaps the SG really believes the conclusion he writes are the text of the law against all political motives.

      Don’t forget: the SG also argued in Prometheus v. Mayo that you could use a patent to protect a “new correlation” by tossing in an old data-gathering step. Somehow section 103 was supposed to do the heavy lifting but the SG never explained how 103 would work to tank a patent that protected non-obvious information, nor did the government care to remember that the patent had already sailed through the PTO.

      Seems to me that if “fair use” is the best answer here (which Dennis, at least, finds suggested in the government’s brief) the government should set forth its understanding of that doctrine.

  16. 2

    Lexmark v. Static Control from the 6th Circuit is educational here. The reason to duplicate the form and function names of the Java API is for interoperability with programs already written in Java. Otherwise, those programs would not be able to work with Google’s API.

    Another way of looking at it is that the interface is merely a list of function prototypes that tell you what a function is named, what it takes as inputs, and what it gives as output. Not too different from a compiled list of facts. The implementations of the API functions are probably copyrightable, at least in part (I have trouble believing that Java’s implementation of max() is any different from all the other implementations over the years), but the interface to those functions should not be.

    1. 2.1

      An alphabet should not be copyrightable as it does not protect the expression in a particular work.

      In this area, nothing is black and white. If the color balance tends to gray, the law should not allow copyright in the first place.

  17. 1

    Java is a language for instructing computing devices.

    Can you describe a language in a book, e.g., a dictionary, and then successfully sue people who use the words you defined? Has that ever happened?

    the company copied the set of more than 6,000 function calls

    I’m curious about the importance of this number. What is the total number of function calls in the “API-toolkit”? Does it matter if 6,000 function calls were taken as opposed to 1000?

    an important case is pending before the U.S. Supreme Court regarding whether computer programs are protectable under copyright.

    I think the question is a tad more nuanced than that!

    1. 1.1

      As I recall, the allowed claims of Morse included such a “defined language.”

      Oops.

    2. 1.2

      What does this have to do with the Google v. Oracle case? “Can you describe a language in a book, e.g., a dictionary, and then successfully sue people who use the words you defined? Has that ever happened?” An API is quite different than a dictionary. For instance, Google used “java.lang.Math.max(a,b)” when they could have used something completely different, such as android.mathematics.determineMaximum(x,y). That is, it’s not necessary for them to use the same words in a “dictionary”.

      Personally, I find this case more akin to a completely new dictionary with made up (very complex) words, and Google copied those words.

    3. 1.3

      Can you describe a language in a book, e.g., a dictionary,

      No. To describe a language you need more than a dictionary. Most of us, I think, would try to come up with several examples on the spectrum before writing what you wrote.

Comments are closed.