3 ms·

There isn't a misunderstanding, Oracle initially made a lot of other copyright and patent claims which were thrown out by the court. Now instead of just yieldin

by avar 14y ago

There isn't a misunderstanding, Oracle initially made a lot of other copyright and patent claims which were thrown out by the court. Now instead of just yielding the case they've fallen back on arguing that they should be awarded for the copying of a trivial function that any programmer could have written in 2 minutes.

TazeTSchnitzel 14y ago
And even 2 minutes is a stretch.
pbhjpbhj 14y ago
>they should be awarded for the copying of a trivial function that any programmer could have written in 2 minutes // The crucial question with copyright is "did they copy it". If it's a trivial portion of code that anyone could have written then why go out of your way (assuming the rest of it is not copied) to copy that particular part. If indeed it is not copied then Oracle will not be able to show it was, not even on the balance of probabilities [that's used in civil tort cases in the UK I gather, not sure of the measure applicable here]. It appears to be a bug in the statute if the test is whether the copied part is de minimis. The UK statute IIRC specifies that it must not be a 'significant part', which seems to lean more to qualitative considerations than quantitative ones.
- usea 14y ago
  It's my understanding that copyright isn't primarily concerned with whether you were looking at the original when you produced the same work. It may be a factor in some cases, but the primary concern is similarity. If the works are the same, the latter work infringes regardless of access to the original. For example, if you write a song which is substantially similar to another song released some time before your own, you've infringed even if you'd never heard the prior work. edit: apparently I am mistaken[1] on this point. My apologies. The point with this particular code is it's so obvious that many (most?) programmers would come up with the exact same code when given the task. It's part of the bigger picture that code isn't prose, it's instructions. How do you accomplish this task in this given system? First A, then B, then C. The first person to give some simple set of instructions should not be the only one allowed to. [1]: http://en.wikipedia.org/wiki/Substantial_similarity http://en.wikipedia.org/wiki/Substantial_similarity
  - pbhjpbhj 14y ago
    >but the primary concern is similarity. // Absolutely not. [Provably] Not copying is an absolute defence.
    - specialist 14y ago
      Yup. aka Clean room design. Alas, not a defense for patent IP.
- rbanffy 14y ago
  It's also a pretty obvious implementation. In fact, the function is so simple, it's hard to implement in any other way.
- ZeroGravitas 14y ago
  He admitted to copying it. The reason he did so was because he intended to (and later did) contribute the code to Oracle's OpenJDK. He wanted to write it as if it was part of that package, and as such it would re-use the original version of that code. So he went "out of his way" to help Oracle and this is how they've repaid him. I understand that the circumstances of an ex-employee contributing to an open source codebase and wanting to do things right by not unnecessarily duplicating code might be too much for your average jury, but it should be made clear to those who can understand that this is utterly appalling behaviour by Oracle, rather than a case of a sloppy Google employee getting away with a small amount of IP theft.

by avar 14y ago

TazeTSchnitzel 14y ago

And even 2 minutes is a stretch.

pbhjpbhj 14y ago

>they should be awarded for the copying of a trivial function that any programmer could have written in 2 minutes // The crucial question with copyright is "did they copy it". If it's a trivial portion of code that anyone could have written then why go out of your way (assuming the rest of it is not copied) to copy that particular part. If indeed it is not copied then Oracle will not be able to show it was, not even on the balance of probabilities [that's used in civil tort cases in the UK I gather, not sure of the measure applicable here]. It appears to be a bug in the statute if the test is whether the copied part is de minimis. The UK statute IIRC specifies that it must not be a 'significant part', which seems to lean more to qualitative considerations than quantitative ones.

usea 14y ago
It's my understanding that copyright isn't primarily concerned with whether you were looking at the original when you produced the same work. It may be a factor in some cases, but the primary concern is similarity. If the works are the same, the latter work infringes regardless of access to the original. For example, if you write a song which is substantially similar to another song released some time before your own, you've infringed even if you'd never heard the prior work. edit: apparently I am mistaken[1] on this point. My apologies. The point with this particular code is it's so obvious that many (most?) programmers would come up with the exact same code when given the task. It's part of the bigger picture that code isn't prose, it's instructions. How do you accomplish this task in this given system? First A, then B, then C. The first person to give some simple set of instructions should not be the only one allowed to. [1]: http://en.wikipedia.org/wiki/Substantial_similarity http://en.wikipedia.org/wiki/Substantial_similarity
- pbhjpbhj 14y ago
  >but the primary concern is similarity. // Absolutely not. [Provably] Not copying is an absolute defence.
  - specialist 14y ago
    Yup. aka Clean room design. Alas, not a defense for patent IP.
rbanffy 14y ago
It's also a pretty obvious implementation. In fact, the function is so simple, it's hard to implement in any other way.
ZeroGravitas 14y ago
He admitted to copying it. The reason he did so was because he intended to (and later did) contribute the code to Oracle's OpenJDK. He wanted to write it as if it was part of that package, and as such it would re-use the original version of that code. So he went "out of his way" to help Oracle and this is how they've repaid him. I understand that the circumstances of an ex-employee contributing to an open source codebase and wanting to do things right by not unnecessarily duplicating code might be too much for your average jury, but it should be made clear to those who can understand that this is utterly appalling behaviour by Oracle, rather than a case of a sloppy Google employee getting away with a small amount of IP theft.