5 ms·

We always see the anti-Oracle/FreeAPI’s regards this story, but I’m curious who out there is on Oracles side in this? Are there any other major players or even

by bori5 7y ago

We always see the anti-Oracle/FreeAPI’s regards this story, but I’m curious who out there is on Oracles side in this? Are there any other major players or even small ones that people know of that would like to see Oracle win this case ?

unlinked_dll 7y ago
I'm on Google's side professionally because it affects my livelihood directly. I'm on Oracle's side dogmatically because I think there's nothing in the law that says they can't copyright an API. I'm on neither side personally because I think this whole case is a massive failing by the law to adapt to new use cases.
- Aloha 7y ago
  Copyright is a grant, not a natural right, which generally means you can only copyright what the legislature says you can.
  - paulddraper 7y ago
    Sure. But if I can copyright a screenplay, what's to say I can't copyright an API? It's not clear to me why one copyright would be legitimate and the other not.
- saurik 7y ago
  Yeah... a big part of my feelings on this case come down to an idea that "the law is broken and needs to be fixed, and applications of the law to this that try to be consistent yield nonsensical results that no one really likes"; so I guess part of me kind of hopes Oracle wins because then it might feel like a set of broken laws is being more consistently applied and thereby maybe someone will finally bother to actually fix it? Copyright as an entire concept is just so arbitrary and broken when you really get into the weeds, and I find it almost impossible to draw any bright lines from first principles :(. I mean, Google clearly wanted to build something similar in many ways to Java and, honestly, did _not_ do so for trivial "interoperability" purposes (which is the most common defense of what Google did, as "interoperability" has often been used as a successful defense of fair use and even has many explicit carve outs and exemptions in the DMCA) as existing Java software did not work on their implementation and they didn't _really_ care (and in fact benefited); the best part of the lawsuit to me was when Google's lawyers failed to find even a single Java program that could run on Android when challenged by Oracle. > District Court noted that Oracle in essence was arguing that Google’s interoperability argument would have been stronger if Google copied more--the structure of all 166 Java API packages--rather than just 37. https://www.law.berkeley.edu/wp-content/uploads/2015/09/Jonathan-Band-oracle-vs-google-interoperability.pdf https://www.law.berkeley.edu/wp-content/uploads/2015/09/Jona... I mean, this is just "good TV", and doesn't lead to a situation to me where anyone is "obviously correct" to me given either the law or the current precedent. At the end of that PDF, the case of Lotus v. Borland is brought up, where Borland had a menu variant compatible with Lotus 1-2-3 for their Quattro Pro. Is the API of a programming environment more like this kind of user interface (which was left without a national precedent as it _split evenly_ at the Supreme Court level: four to four with one recusal) or more like the API of the IBM BIOS that had to be replicated by Compaq (which _was_ considered to be copyrightable)? I really feel like most people are largely on Google's side for, as you state, "professional" reasons, as it causes all of us issues if it turns out that Google wasn't allowed to do this, as we all _like_ being able to build something "kind of similar to what this other person built but not compatible enough so that our use is neutral". Imagine if this got applied to protocols... what kind of effect would that have on the web? Oracle has also been so classically evil--and there's still this residual delusion that Google isn't--that I think we all have blinders on with respect to whether they might be right.
  - cromwellian 7y ago
    In 2008, no smartphone could really handle the full Java API. Java failed on mobile precisely because Desktop Java was a pig, and the mobile variants were either pigs too, or so non-featured, that they represented no market parity. Does anyone really thing Swing, or PersonalJava, or MIDP on phones, or even JavaFX, was going to compete with iOS? The whole write-once-run-anywhere promise of Java is a sham. Any non-trivial professional Java app, used some non-portable APIs. Take the popular ones like Limewire, or IntelliJ, and they end up on each platform, hooking into MacOS, Windows, et al, specific APIs that require JNI code. " Imagine if this got applied to protocols" That's exactly what happened since the 80s. Do you know how many versions of IMAP there are if you count the extensions? Or Jabber? The internet was built on "rough consensus and running code". People shared code and ideas freely. That's what built the net. Copyrighting APIs with 100+ year time spans is INSANE.
    - saurik 7y ago
      > " Imagine if this got applied to protocols" > That's exactly what happened since the 80s. You managed to invert my paragraph somehow :/. Imagine if [this: the idea that APIs are copyrightable] got applied to protocols... that would be horrible, and would have had massive (negative) impacts on the web; this categorically did not happen "since the 80s". Therefore, paying no attention to the law or the precedent or even the facts of the case, we all have this visceral reaction to the idea behind this lawsuit and want it to be decided in a way that feels like it works for us... but maybe the issue is that the law is broken. > Copyrighting APIs with 100+ year time spans is INSANE. I would argue copyrighting anything with 100+ year time spans is INSANE. The idea that developing and distributing techniques that even merely help people copy things is illegal (a la Section 1201(b))... the issue is that this entire area of law is INSANE :/.
- jayd16 7y ago
  >there's nothing in the law that says they can't copyright an API. Well this is copyright law and not patent law but my understanding is in patent law they've already established things like obvious plug designs cannot be patented. If we assume the same is to be said for code you easily argue that to keep compatibility for any given method signature, the reimplementation method signature is obvious.
  - boublepop 7y ago
    Copying anything is “obvious” given the source material. Your arguing against all forms of copyright not just Software interfaces.
    - jayd16 7y ago
      You can often reimplement a function body in a way that happens to be the exact way another person implimented the method. There's only so many ways to count the size of an array. That does not mean you're committing copyright infringement. Google has shown that they did not copy the bodies verbatim so its down to the API. There happens to only be a single way to write the method signatures of a compatible method.
      - tsimionescu 7y ago
        > There happens to only be a single way to write the method signatures of a compatible method. That's not actually true. For example, all of the following method headers are compatible: void foo(int a) throws Exception void foo(int x) throws IOException void foo(int index) throws RuntimeException Sure, this is a technicality, but it is not completely irrelevant. However, the whole discussion is moot, because Google had no intention of being compatible with anything, as shown by the fact that there is no real-world Java program that runs as-is on Android. Also, copyright only applies to actual copies of the original work. Google could have done a clean-room redesign of the Java API starting from Java programs, they could have arrived at the same API that they have today, and could not have been successfully accused of copying.
tsimionescu 7y ago
Personally, although I generally despise their practices, I think that Oracle is fully in the right here. APIs are obviously the result of human creativity and they do deserve protection. Google only chose to copy the Java API for their new platform because they liked it aesthetically and didn't want to invest into creating a new API to use for their platform (a decision I'm sure they are regretting). What I think would make most sense is to allow APIs to be copyrighted, but to enshrine a specific fair-use exception for interoperability. If that were granted, WINE and other emulators would be allowed to copy Windows' API under that fair-use exception, whereas Google would still not be allowed to do what they did.
- extropy 7y ago
  The concept is tempting. However in most cases the implementation is only partial. As in Wine and Samba, etc do not implement the full extent of the API. Just the part that's practical for their purposes. And that's essentially what Google did - made a partial implementation of the API. I don't see a good way to rule that Google use was protected but other not 100% complete implementations are.
dragonwriter 7y ago
> Are there any other major players or even small ones that people know of that would like to see Oracle win this case ? The Trump Administration. (They've filed, AFAIK, the only amicus brief supporting Oracle in the case. I would think they count as a major player.)
jariel 7y ago
I'm on Oracle's side because I think APIs are essentially the same as code and should be protected similarly. I don't believe that this will be massively disruptive to the industry. I think it's unfair that Google or anyone should have to pay anyone for doing something that was no clearly unlawful and the gov. should also step in and clear up the legality of these things. Clarity on these matters will help everyone, everywhere, except lawyers. I do not believe that the adversarial justice system is remotely the right place to be making these decisions.

by bori5 7y ago

unlinked_dll 7y ago

I'm on Google's side professionally because it affects my livelihood directly. I'm on Oracle's side dogmatically because I think there's nothing in the law that says they can't copyright an API. I'm on neither side personally because I think this whole case is a massive failing by the law to adapt to new use cases.

Aloha 7y ago
Copyright is a grant, not a natural right, which generally means you can only copyright what the legislature says you can.
- paulddraper 7y ago
  Sure. But if I can copyright a screenplay, what's to say I can't copyright an API? It's not clear to me why one copyright would be legitimate and the other not.
saurik 7y ago
Yeah... a big part of my feelings on this case come down to an idea that "the law is broken and needs to be fixed, and applications of the law to this that try to be consistent yield nonsensical results that no one really likes"; so I guess part of me kind of hopes Oracle wins because then it might feel like a set of broken laws is being more consistently applied and thereby maybe someone will finally bother to actually fix it? Copyright as an entire concept is just so arbitrary and broken when you really get into the weeds, and I find it almost impossible to draw any bright lines from first principles :(. I mean, Google clearly wanted to build something similar in many ways to Java and, honestly, did _not_ do so for trivial "interoperability" purposes (which is the most common defense of what Google did, as "interoperability" has often been used as a successful defense of fair use and even has many explicit carve outs and exemptions in the DMCA) as existing Java software did not work on their implementation and they didn't _really_ care (and in fact benefited); the best part of the lawsuit to me was when Google's lawyers failed to find even a single Java program that could run on Android when challenged by Oracle. > District Court noted that Oracle in essence was arguing that Google’s interoperability argument would have been stronger if Google copied more--the structure of all 166 Java API packages--rather than just 37. https://www.law.berkeley.edu/wp-content/uploads/2015/09/Jonathan-Band-oracle-vs-google-interoperability.pdf https://www.law.berkeley.edu/wp-content/uploads/2015/09/Jona... I mean, this is just "good TV", and doesn't lead to a situation to me where anyone is "obviously correct" to me given either the law or the current precedent. At the end of that PDF, the case of Lotus v. Borland is brought up, where Borland had a menu variant compatible with Lotus 1-2-3 for their Quattro Pro. Is the API of a programming environment more like this kind of user interface (which was left without a national precedent as it _split evenly_ at the Supreme Court level: four to four with one recusal) or more like the API of the IBM BIOS that had to be replicated by Compaq (which _was_ considered to be copyrightable)? I really feel like most people are largely on Google's side for, as you state, "professional" reasons, as it causes all of us issues if it turns out that Google wasn't allowed to do this, as we all _like_ being able to build something "kind of similar to what this other person built but not compatible enough so that our use is neutral". Imagine if this got applied to protocols... what kind of effect would that have on the web? Oracle has also been so classically evil--and there's still this residual delusion that Google isn't--that I think we all have blinders on with respect to whether they might be right.
- cromwellian 7y ago
  In 2008, no smartphone could really handle the full Java API. Java failed on mobile precisely because Desktop Java was a pig, and the mobile variants were either pigs too, or so non-featured, that they represented no market parity. Does anyone really thing Swing, or PersonalJava, or MIDP on phones, or even JavaFX, was going to compete with iOS? The whole write-once-run-anywhere promise of Java is a sham. Any non-trivial professional Java app, used some non-portable APIs. Take the popular ones like Limewire, or IntelliJ, and they end up on each platform, hooking into MacOS, Windows, et al, specific APIs that require JNI code. " Imagine if this got applied to protocols" That's exactly what happened since the 80s. Do you know how many versions of IMAP there are if you count the extensions? Or Jabber? The internet was built on "rough consensus and running code". People shared code and ideas freely. That's what built the net. Copyrighting APIs with 100+ year time spans is INSANE.
  - saurik 7y ago
    > " Imagine if this got applied to protocols" > That's exactly what happened since the 80s. You managed to invert my paragraph somehow :/. Imagine if [this: the idea that APIs are copyrightable] got applied to protocols... that would be horrible, and would have had massive (negative) impacts on the web; this categorically did not happen "since the 80s". Therefore, paying no attention to the law or the precedent or even the facts of the case, we all have this visceral reaction to the idea behind this lawsuit and want it to be decided in a way that feels like it works for us... but maybe the issue is that the law is broken. > Copyrighting APIs with 100+ year time spans is INSANE. I would argue copyrighting anything with 100+ year time spans is INSANE. The idea that developing and distributing techniques that even merely help people copy things is illegal (a la Section 1201(b))... the issue is that this entire area of law is INSANE :/.
jayd16 7y ago
>there's nothing in the law that says they can't copyright an API. Well this is copyright law and not patent law but my understanding is in patent law they've already established things like obvious plug designs cannot be patented. If we assume the same is to be said for code you easily argue that to keep compatibility for any given method signature, the reimplementation method signature is obvious.
- boublepop 7y ago
  Copying anything is “obvious” given the source material. Your arguing against all forms of copyright not just Software interfaces.
  - jayd16 7y ago
    You can often reimplement a function body in a way that happens to be the exact way another person implimented the method. There's only so many ways to count the size of an array. That does not mean you're committing copyright infringement. Google has shown that they did not copy the bodies verbatim so its down to the API. There happens to only be a single way to write the method signatures of a compatible method.
    - tsimionescu 7y ago
      > There happens to only be a single way to write the method signatures of a compatible method. That's not actually true. For example, all of the following method headers are compatible: void foo(int a) throws Exception void foo(int x) throws IOException void foo(int index) throws RuntimeException Sure, this is a technicality, but it is not completely irrelevant. However, the whole discussion is moot, because Google had no intention of being compatible with anything, as shown by the fact that there is no real-world Java program that runs as-is on Android. Also, copyright only applies to actual copies of the original work. Google could have done a clean-room redesign of the Java API starting from Java programs, they could have arrived at the same API that they have today, and could not have been successfully accused of copying.

tsimionescu 7y ago

Personally, although I generally despise their practices, I think that Oracle is fully in the right here. APIs are obviously the result of human creativity and they do deserve protection. Google only chose to copy the Java API for their new platform because they liked it aesthetically and didn't want to invest into creating a new API to use for their platform (a decision I'm sure they are regretting). What I think would make most sense is to allow APIs to be copyrighted, but to enshrine a specific fair-use exception for interoperability. If that were granted, WINE and other emulators would be allowed to copy Windows' API under that fair-use exception, whereas Google would still not be allowed to do what they did.

extropy 7y ago
The concept is tempting. However in most cases the implementation is only partial. As in Wine and Samba, etc do not implement the full extent of the API. Just the part that's practical for their purposes. And that's essentially what Google did - made a partial implementation of the API. I don't see a good way to rule that Google use was protected but other not 100% complete implementations are.

dragonwriter 7y ago

> Are there any other major players or even small ones that people know of that would like to see Oracle win this case ? The Trump Administration. (They've filed, AFAIK, the only amicus brief supporting Oracle in the case. I would think they count as a major player.)

jariel 7y ago

I'm on Oracle's side because I think APIs are essentially the same as code and should be protected similarly. I don't believe that this will be massively disruptive to the industry. I think it's unfair that Google or anyone should have to pay anyone for doing something that was no clearly unlawful and the gov. should also step in and clear up the legality of these things. Clarity on these matters will help everyone, everywhere, except lawyers. I do not believe that the adversarial justice system is remotely the right place to be making these decisions.