3 ms·

What really sucks is this was a perfect time for a lawyer to use Justice Thomas' analogy to better explain the difference. I would have tried something like: "

by fdye 6y ago

What really sucks is this was a perfect time for a lawyer to use Justice Thomas' analogy to better explain the difference. I would have tried something like: "Close your honor, but let me expand on this further. Imagine instead that your opponent publishes their calls and they are regularly available to all individuals of the general public. However, while they may describe a call for going long, they do not give any specifics if the receiver starts from the left, center, right side of the field, how fast he should go, who should block for him, etc. They essentially are just assigning a call to a general concept of 'go long'. Now as a competitor team, I look through their publicly available calls and say "Hey, thats a good idea, we should have a call for 'go long' as well". However, in practice my team does start from the left, and then cut in the center of the field to receive the ball. This may, or may not be totally different then what my opponents choose to do when they call "go long" The additional benefit is if I transfer a new coach onto my team, he can use the call "go long" and even though the team accomplishes the goal, possibly by a different route, he can make the call to accomplish essentially the same thing." Note: Not a huge sports guys, so perhaps my football analogy broke down a bit there, but hopefully you get the idea.

ballenf 6y ago
I think the playbook is just a bad analogy because it misses that point of what an interface is. It's the means of connecting two dissimilar mechanisms. A translation layer. I'd have argued it's more like copyrighting doorknob placement (opposite hinge, at hand height) and operation (twisting) on doors vs. patenting a novel mechanism inside the doorknob. A football playbook is more like the blueprints of the inside of the doorknob. Of course it's patentable if playbooks are patentable (I presume they are if considered works of choreography as in a dance).
- Shakahs 6y ago
  I think the QWERTY analogy quoted in the article is perfect. QWERTY is on typewriters, computers, and smartphones. Same interface, completely different implementations. And this is technology that everyday people are familiar with.
  - Chris_Newton 6y ago
    It has always seemed to me that a good analogy for the API situation would be the design of typefaces (as distinct from a particular implementation of that design through, for example, instructions in a font file), which receives special treatment in US law and is not subject to copyright in that jurisdiction. As far as I can see (but please note that I am not even in the US, never mind a legal scholar) the relevant US law is 37 CFR § 202.1 about “Material not subject to copyright” (https://www.law.cornell.edu/cfr/text/37/202.1 https://www.law.cornell.edu/cfr/text/37/202.1). This enumerates several cases that are explicitly excluded from copyright protection. One is “typeface as typeface”, which was apparently interpreted as described above in a litigated case (Eltra Corp. vs. Ringer). An argument has certainly been made, though I can’t immediately find the original source, that this exclusion is justified because of the utilitarian nature of a typeface design: allowing one party to control all use of it would be against the public interest, because ultimately that principle could allow for all printed communication to become controlled. It seems to me that APIs fulfil a similar purpose. They are not useful in themselves, lacking a corresponding implementation. They are, however, an essential basis for standardisation, interoperability and communication between practitioners. Under 202.1(b) of the law cited above, another class of works excluded from copyright protection is “Ideas, plans, methods, systems, or devices, as distinguished from the particular manner in which they are expressed or described in a writing”. I wonder whether a similar argument about the utilitarian nature of APIs as a specification distinct from its implementation could be made based on that.
- jacobr1 6y ago
  But we are talking about copyright, not patents. So to stretch your analogy. Imagine I'm a door manufacturer, or maybe a manufacturer of tools for construction so that professionals can make their own doors. I create some kind of stencil that has the doorknob placement cut-out. And maybe a few variants with places for windows and such. The general idea of such a stencil might be patentable, even if the general standard dimensions are not. The tool is a unique idea (maybe). But we are talking about copywritten material. The specific stencil design, that you could photocopy is the thing in question. And that is one of several arguments Oracle made, that api interface is a stencil. I think I would agree the law supports that argument from what I know about it, copyright is pretty broad. But at the same time, I think that is perverse and we should change the law.
- mamon 6y ago
  Or compare APIs to the car interface: steering wheel + pedals. If one car manufacturer was able to copyright them, then what are other manufacturers supposed to do? The second one can use joystick for steering, and maybe a lever for braking and acceleration. The third one has to invent something even weirder. Everybody loses: manufacturers, because they can't easily create competing products, and users, because they have to re-learn driving every time they change car brand.
defen 6y ago
If I were Oracle's lawyer, I'd respond by saying that you can have the concept of "go long", but you can't literally name it "go long" because there was an act of creativity involved in mapping that name onto the concept of a receiver running far down the field. So you have to call it "go deep" or whatever.
- Spivak 6y ago
  It's not about the names though. It's that when you look at playbooks for both teams you see that the names of all the plays are identical, and under the same headings! And you come to the conclusion that the second team must have copied the names from the first teams playbook. Such a thing might be considered fair use if the second team's argument was they indeed copied the playbook because they had a recruited some players from the first team and keeping the names the same made it easier for them during practice. But for now the judge is suspicious that they've come up with this explanation after the fact.
  - defen 6y ago
    Yeah, I don't like it, but given my (probably poor) understanding of copyright, I don't see how Oracle is wrong here. It would be one thing if we were just talking about something like Math.max where there's really only 1 possible way to declare it; but it's the entirety of the API.

by fdye 6y ago

ballenf 6y ago

I think the playbook is just a bad analogy because it misses that point of what an interface is. It's the means of connecting two dissimilar mechanisms. A translation layer. I'd have argued it's more like copyrighting doorknob placement (opposite hinge, at hand height) and operation (twisting) on doors vs. patenting a novel mechanism inside the doorknob. A football playbook is more like the blueprints of the inside of the doorknob. Of course it's patentable if playbooks are patentable (I presume they are if considered works of choreography as in a dance).

Shakahs 6y ago
I think the QWERTY analogy quoted in the article is perfect. QWERTY is on typewriters, computers, and smartphones. Same interface, completely different implementations. And this is technology that everyday people are familiar with.
- Chris_Newton 6y ago
  It has always seemed to me that a good analogy for the API situation would be the design of typefaces (as distinct from a particular implementation of that design through, for example, instructions in a font file), which receives special treatment in US law and is not subject to copyright in that jurisdiction. As far as I can see (but please note that I am not even in the US, never mind a legal scholar) the relevant US law is 37 CFR § 202.1 about “Material not subject to copyright” (https://www.law.cornell.edu/cfr/text/37/202.1 https://www.law.cornell.edu/cfr/text/37/202.1). This enumerates several cases that are explicitly excluded from copyright protection. One is “typeface as typeface”, which was apparently interpreted as described above in a litigated case (Eltra Corp. vs. Ringer). An argument has certainly been made, though I can’t immediately find the original source, that this exclusion is justified because of the utilitarian nature of a typeface design: allowing one party to control all use of it would be against the public interest, because ultimately that principle could allow for all printed communication to become controlled. It seems to me that APIs fulfil a similar purpose. They are not useful in themselves, lacking a corresponding implementation. They are, however, an essential basis for standardisation, interoperability and communication between practitioners. Under 202.1(b) of the law cited above, another class of works excluded from copyright protection is “Ideas, plans, methods, systems, or devices, as distinguished from the particular manner in which they are expressed or described in a writing”. I wonder whether a similar argument about the utilitarian nature of APIs as a specification distinct from its implementation could be made based on that.
jacobr1 6y ago
But we are talking about copyright, not patents. So to stretch your analogy. Imagine I'm a door manufacturer, or maybe a manufacturer of tools for construction so that professionals can make their own doors. I create some kind of stencil that has the doorknob placement cut-out. And maybe a few variants with places for windows and such. The general idea of such a stencil might be patentable, even if the general standard dimensions are not. The tool is a unique idea (maybe). But we are talking about copywritten material. The specific stencil design, that you could photocopy is the thing in question. And that is one of several arguments Oracle made, that api interface is a stencil. I think I would agree the law supports that argument from what I know about it, copyright is pretty broad. But at the same time, I think that is perverse and we should change the law.
mamon 6y ago
Or compare APIs to the car interface: steering wheel + pedals. If one car manufacturer was able to copyright them, then what are other manufacturers supposed to do? The second one can use joystick for steering, and maybe a lever for braking and acceleration. The third one has to invent something even weirder. Everybody loses: manufacturers, because they can't easily create competing products, and users, because they have to re-learn driving every time they change car brand.

defen 6y ago

If I were Oracle's lawyer, I'd respond by saying that you can have the concept of "go long", but you can't literally name it "go long" because there was an act of creativity involved in mapping that name onto the concept of a receiver running far down the field. So you have to call it "go deep" or whatever.

Spivak 6y ago
It's not about the names though. It's that when you look at playbooks for both teams you see that the names of all the plays are identical, and under the same headings! And you come to the conclusion that the second team must have copied the names from the first teams playbook. Such a thing might be considered fair use if the second team's argument was they indeed copied the playbook because they had a recruited some players from the first team and keeping the names the same made it easier for them during practice. But for now the judge is suspicious that they've come up with this explanation after the fact.
- defen 6y ago
  Yeah, I don't like it, but given my (probably poor) understanding of copyright, I don't see how Oracle is wrong here. It would be one thing if we were just talking about something like Math.max where there's really only 1 possible way to declare it; but it's the entirety of the API.