5 ms·

Quick question: Why can't an API be copyrighted? My understanding is that the purpose of a copyright is to protect a specific arrangement of ideas. A piece o

Quick question: Why can't an API be copyrighted? My understanding is that the purpose of a copyright is to protect a specific arrangement of ideas. A piece of music, a book, a painting. All of those things are an arrangement, and the arrangement itself is what's protected by copyright, not the physical object that was created. So, given that an API is an arrangement of ideas, and that copyright apparently protects arrangements of ideas, why is copyright inapplicable here? If the argument is that copyrights shouldn't apply to APIs, then I understand. It's probably more harmful than beneficial, and thus should be prevented. But if the argument is that copyright is incompatible with the domain of APIs, as if you were trying to copyright a taste like chocolate or copyright a smell, then I don't understand that at all. Would anyone mind clarifying?

ewillbefull 12y ago
The Copyright Act states: "In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work." The common belief is that an API meets this "system or method of operation" standard.
- sillysaurus3 12y ago
  Hmm... What's a "method of operation" in the context of computer programs? This isn't helpful: http://legal-dictionary.thefreedictionary.com/Method+of+operation http://legal-dictionary.thefreedictionary.com/Method+of+oper... I really don't understand how an API is a method of operation in any sense. It's an understanding between people about how to use a software library. The only sense that I could see it being a method of operation is literally "This is a description of how you operate the software library." In other words, an API is an instruction booklet for people, not computers. But in that case, why are instruction booklets copyrightable, but not software APIs?
  - ewillbefull 12y ago
    Honestly, just read the brief, it goes into detail the precedent involved in their argument: https://www.eff.org/files/2014/11/07/google_v_oracle_computer-scientists-certpetition-amicus-brief_14-410_final.pdf https://www.eff.org/files/2014/11/07/google_v_oracle_compute... for example... Ninth Circuit law—which should have been controlling in this case—was the same: the “functional requirements for compatibility” between computer programs “are not protected by copyright.” Sega Enters., Ltd., v. Accolade, Inc. Also, an API is much more than you think. An API might be Windows' system APIs which applications use to interact with the operating system -- put a license behind that and WINE is gone overnight. An API is also totally indistinguishable from a network protocol, in both a technical and legal sense.
  - dragonwriter 12y ago
    > What's a "method of operation" in the context of computer programs? The more dangerous question is "what is a computer program except a method of operation?"
    - AnthonyMouse 12y ago
      That isn't really a problem. You can have two programs that do the same thing, e.g. BSD cat and GNU cat, but you're copyrighting the specific implementation rather than the method of operation. Each method of operation can have arbitrarily many different implementations.
      - throwawaykf05 12y ago
        By that logic, copyrighting APIs is not really a problem either: Each API can have arbitrarily different concepts, paradigms, organization, and naming and calling conventions. Hence even APIs for doing the exact same things can (and do) look completely different from each other. Compare, for examples, the Java APIs with those for C++, C#, Python, Ruby, Go, etc. etc.
        AnthonyMouse 12y ago
- vegedor 12y ago
  Please name one thing that doesn't meat any of these standards, maybe then I can understand the difference.
magicalist 12y ago
One thing that many in the tech world get hung up on is the idea that this is arguing that designing APIs is not a creative endeavor. That's not what's being argued here. There are many creative acts which are not covered by copyright, and the argument is that APIs are one of them, since they are a functional description (the "method of operation"/"functional requirements for compatibility" covered by ewillbefull). It's not saying that API design is an incidental and uninteresting byproduct of the design of software, it's saying that as a method of operating that code, they aren't copyrightable. Incidentally, that may mean that these APIs could be covered by patents, but Oracle's patent argument in the original suit was extremely weak and so they dropped it, so it wouldn't be relevant to this case. It would also generally be very difficult to get patents on APIs without the Federal Circuit working a whole lot more of their patent-expansion magic.
- pron 12y ago
  > it's saying that as a method of operating that code, they aren't copyrightable That's an argument I could never understand. And remember, even though the term API has been borrowed in the past few years to mean "protocol", API in the context of this case pertains only to its classical meaning. As such, a specific API is indeed a method of operating code, but also a very elaborate text -- including specific class names and method names (and a specific ordering of parameters). None of the names is essential for replicating the functionality of the code. For example, String.length() could have just as easily been named string.len() or Str.size(). Because Googles "JDK" isn't actually the JDK -- and therefore does not interoperate with any actual JDK classes -- but simply a copy of them, the names, organization, etc. are absolutely not necessary in order to operate Google's code. Google's code was never meant to interoperate with Java, but to replicate it (once they've disregarded Java's GPL license this may not matter, but they've also broken the JDK's license in other ways, namely, they've made a partial, incompatible implementation). How, then, is the API being "a method of operating code" relevant to the matter at hand?
olalonde 12y ago
> purpose of a copyright is to protect a specific arrangement of ideas The purpose of copyright is to encourage innovation and creativity. In a broader sense, the ultimate goal is to benefit society. I think the real question we you should be asking is "will society benefit more from copyrighted APIs or not?". I personally think the answer is that it would be a net negative for society but that's my own opinion.
jrochkind1 12y ago
There's lots of previous caselaw in the U.S. suggesting API's are not copyrightable, basically on the theory that functionality itself, as well as factual statements or descriptions of something, are not copyrightable. For instance, if you write a recipe, the -exact text- of that recipe is copyrightable, but the actual recipe itself is not -- someone else can reword it in their own words, with the same basic ingredients combined using the same operations, and that is not protected by the recipe-writer's copyright. This is (or was?) established in U.S. law. Similarly, in the U.S. _functionality_, behaviors or operations themselves, are not copyrightable (although it may be patent-able). You can copy a competitors functionality exactly, as long as you do it with your own original implementation -- the functionality is not protected by copyright. (This is why people do 'clean room' compatible implementations by developers who never saw the original code they are copying the functionality of). In general, the right reverse engineer and reproduce the functionality of a technical work without the permission of the copyright holder of the original work -- is (or has been?) legal and protected. (I am not sure the judges involved realized the relationship between promoting the copyrightability of API, and what this does to the previously well-established right to reverse engineer and create compatible things). So the belief, previous to this case, was that an API is more like functionality, or the steps in the recipe. An implementation of an API is copyrightable -- but the description of the API itself, the classes, methods and signatures, are more like functionality or statements of fact, more like the steps of a recipe or a description of behavior. Both of which are not copyrightable -- you can't prevent someone from "doing the same thing", only from copying your particular code or text), then they are like a creative work in fixed medium themselves (the exact text of a particular written recipe, or a particular code implementation. (Patent law lets you prevent others from 'doing the same thing', but that's a different law with different requirements for what is patentable vs what is copyrightable, different legal rules for determining if a violation occured and what the penalties are, etc.) You are (or were?) allowed to copy an API in your own implementation by reverse engineering in order to create a compatible implementation, and it doesn't make a lot of sense if you lose this right because you can get the API from docs instead of reverse engineering it. I'd imagine you can get more details from various briefs involved these cases, as far as cites to specific previous law. But to many, the fact that API's are not copyrightable is what's consistent with existing law over copyright, including the lack of copyright over the actual operations of a recipe themselves and the protected ability to reverse engineer someone elses thing to create your own compatible implementation. It seems to many that making API's copyrightable is inconsistent with much existing law.
shmerl 12y ago
Simple answer - it's too generic of a level to allow copyrighting it. It ends up in vendor lock-in and anticompetitive behavior. Same reason abstract ideas aren't allowed to be copyrighted - it gives too much monopoly to those who would claim them.
tzs 12y ago
I don't recall which case this was in, but the court in one of the early API cases noted that allowing copyright of APIs would give the copyright owner control over the use of their software that is almost the same as the control a patent gives a patentee over the use of their invention--except for a much much longer term, and without any requirement of novelty or non-obviousness. Copyright is supposed to protect just the expression of an idea, not the underlying idea itself. With API copyrights, there isn't room left for others to express the same idea.
ghshephard 12y ago
It's important (and sometimes difficult for us technically oriented types) to recognize that the law isn't necessarily (despite what Scalia would have us all believe) a slavish dedication to the precise words, and that the broader context and purpose of the law is taken into account when making judgements. Just because a particular reading of copyright law might suggest that APIs could be copyrighted, doesn't mean that in the presence of a greater good ("...To promote the progress of science and useful arts..") the court might just decide that allowing the copyright of APIs would be incorrect.

ewillbefull 12y ago

The Copyright Act states: "In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work." The common belief is that an API meets this "system or method of operation" standard.

sillysaurus3 12y ago
Hmm... What's a "method of operation" in the context of computer programs? This isn't helpful: http://legal-dictionary.thefreedictionary.com/Method+of+operation http://legal-dictionary.thefreedictionary.com/Method+of+oper... I really don't understand how an API is a method of operation in any sense. It's an understanding between people about how to use a software library. The only sense that I could see it being a method of operation is literally "This is a description of how you operate the software library." In other words, an API is an instruction booklet for people, not computers. But in that case, why are instruction booklets copyrightable, but not software APIs?
- ewillbefull 12y ago
  Honestly, just read the brief, it goes into detail the precedent involved in their argument: https://www.eff.org/files/2014/11/07/google_v_oracle_computer-scientists-certpetition-amicus-brief_14-410_final.pdf https://www.eff.org/files/2014/11/07/google_v_oracle_compute... for example... Ninth Circuit law—which should have been controlling in this case—was the same: the “functional requirements for compatibility” between computer programs “are not protected by copyright.” Sega Enters., Ltd., v. Accolade, Inc. Also, an API is much more than you think. An API might be Windows' system APIs which applications use to interact with the operating system -- put a license behind that and WINE is gone overnight. An API is also totally indistinguishable from a network protocol, in both a technical and legal sense.
- dragonwriter 12y ago
  > What's a "method of operation" in the context of computer programs? The more dangerous question is "what is a computer program except a method of operation?"
  - AnthonyMouse 12y ago
    That isn't really a problem. You can have two programs that do the same thing, e.g. BSD cat and GNU cat, but you're copyrighting the specific implementation rather than the method of operation. Each method of operation can have arbitrarily many different implementations.
    - throwawaykf05 12y ago
      By that logic, copyrighting APIs is not really a problem either: Each API can have arbitrarily different concepts, paradigms, organization, and naming and calling conventions. Hence even APIs for doing the exact same things can (and do) look completely different from each other. Compare, for examples, the Java APIs with those for C++, C#, Python, Ruby, Go, etc. etc.
      - AnthonyMouse 12y ago
vegedor 12y ago
Please name one thing that doesn't meat any of these standards, maybe then I can understand the difference.

magicalist 12y ago

One thing that many in the tech world get hung up on is the idea that this is arguing that designing APIs is not a creative endeavor. That's not what's being argued here. There are many creative acts which are not covered by copyright, and the argument is that APIs are one of them, since they are a functional description (the "method of operation"/"functional requirements for compatibility" covered by ewillbefull). It's not saying that API design is an incidental and uninteresting byproduct of the design of software, it's saying that as a method of operating that code, they aren't copyrightable. Incidentally, that may mean that these APIs could be covered by patents, but Oracle's patent argument in the original suit was extremely weak and so they dropped it, so it wouldn't be relevant to this case. It would also generally be very difficult to get patents on APIs without the Federal Circuit working a whole lot more of their patent-expansion magic.

pron 12y ago
> it's saying that as a method of operating that code, they aren't copyrightable That's an argument I could never understand. And remember, even though the term API has been borrowed in the past few years to mean "protocol", API in the context of this case pertains only to its classical meaning. As such, a specific API is indeed a method of operating code, but also a very elaborate text -- including specific class names and method names (and a specific ordering of parameters). None of the names is essential for replicating the functionality of the code. For example, String.length() could have just as easily been named string.len() or Str.size(). Because Googles "JDK" isn't actually the JDK -- and therefore does not interoperate with any actual JDK classes -- but simply a copy of them, the names, organization, etc. are absolutely not necessary in order to operate Google's code. Google's code was never meant to interoperate with Java, but to replicate it (once they've disregarded Java's GPL license this may not matter, but they've also broken the JDK's license in other ways, namely, they've made a partial, incompatible implementation). How, then, is the API being "a method of operating code" relevant to the matter at hand?

olalonde 12y ago

> purpose of a copyright is to protect a specific arrangement of ideas The purpose of copyright is to encourage innovation and creativity. In a broader sense, the ultimate goal is to benefit society. I think the real question we you should be asking is "will society benefit more from copyrighted APIs or not?". I personally think the answer is that it would be a net negative for society but that's my own opinion.

jrochkind1 12y ago

There's lots of previous caselaw in the U.S. suggesting API's are not copyrightable, basically on the theory that functionality itself, as well as factual statements or descriptions of something, are not copyrightable. For instance, if you write a recipe, the -exact text- of that recipe is copyrightable, but the actual recipe itself is not -- someone else can reword it in their own words, with the same basic ingredients combined using the same operations, and that is not protected by the recipe-writer's copyright. This is (or was?) established in U.S. law. Similarly, in the U.S. _functionality_, behaviors or operations themselves, are not copyrightable (although it may be patent-able). You can copy a competitors functionality exactly, as long as you do it with your own original implementation -- the functionality is not protected by copyright. (This is why people do 'clean room' compatible implementations by developers who never saw the original code they are copying the functionality of). In general, the right reverse engineer and reproduce the functionality of a technical work without the permission of the copyright holder of the original work -- is (or has been?) legal and protected. (I am not sure the judges involved realized the relationship between promoting the copyrightability of API, and what this does to the previously well-established right to reverse engineer and create compatible things). So the belief, previous to this case, was that an API is more like functionality, or the steps in the recipe. An implementation of an API is copyrightable -- but the description of the API itself, the classes, methods and signatures, are more like functionality or statements of fact, more like the steps of a recipe or a description of behavior. Both of which are not copyrightable -- you can't prevent someone from "doing the same thing", only from copying your particular code or text), then they are like a creative work in fixed medium themselves (the exact text of a particular written recipe, or a particular code implementation. (Patent law lets you prevent others from 'doing the same thing', but that's a different law with different requirements for what is patentable vs what is copyrightable, different legal rules for determining if a violation occured and what the penalties are, etc.) You are (or were?) allowed to copy an API in your own implementation by reverse engineering in order to create a compatible implementation, and it doesn't make a lot of sense if you lose this right because you can get the API from docs instead of reverse engineering it. I'd imagine you can get more details from various briefs involved these cases, as far as cites to specific previous law. But to many, the fact that API's are not copyrightable is what's consistent with existing law over copyright, including the lack of copyright over the actual operations of a recipe themselves and the protected ability to reverse engineer someone elses thing to create your own compatible implementation. It seems to many that making API's copyrightable is inconsistent with much existing law.

shmerl 12y ago

Simple answer - it's too generic of a level to allow copyrighting it. It ends up in vendor lock-in and anticompetitive behavior. Same reason abstract ideas aren't allowed to be copyrighted - it gives too much monopoly to those who would claim them.

tzs 12y ago

I don't recall which case this was in, but the court in one of the early API cases noted that allowing copyright of APIs would give the copyright owner control over the use of their software that is almost the same as the control a patent gives a patentee over the use of their invention--except for a much much longer term, and without any requirement of novelty or non-obviousness. Copyright is supposed to protect just the expression of an idea, not the underlying idea itself. With API copyrights, there isn't room left for others to express the same idea.

ghshephard 12y ago

It's important (and sometimes difficult for us technically oriented types) to recognize that the law isn't necessarily (despite what Scalia would have us all believe) a slavish dedication to the precise words, and that the broader context and purpose of the law is taken into account when making judgements. Just because a particular reading of copyright law might suggest that APIs could be copyrighted, doesn't mean that in the presence of a greater good ("...To promote the progress of science and useful arts..") the court might just decide that allowing the copyright of APIs would be incorrect.