A setback for Google against Oracle
A setback for Google against Oracle
Posted May 17, 2014 3:21 UTC (Sat) by Cyberax (✭ supporter ✭, #52523)In reply to: A setback for Google against Oracle by viro
Parent article: A setback for Google against Oracle
Posted May 17, 2014 22:34 UTC (Sat)
by giraffedata (guest, #1954)
[Link] (7 responses)
The question of whether a contract is copyrightable has exactly the same terminology problem as the question of whether an API is copyrightable.
A contract, in legal discussion, is not a writing. It is a disembodied agreement. Indeed, many contracts don't involve any writing at all. By this definition, a contract is definitely not copyrightable. If John agrees to give Mary $1000 to paint John's house and Bob hears about that and agrees to give Susan $1000 to paint Bob's house, there is no copyright infringement. Bob and Susan owe John and Mary nothing.
But in informal talk, "contract" often comes to mean the memorandum of a contract - a writing that memorializes what the parties agreed to. These are quite copyrightable. In fact, because they are written by people with highly specialized skills (lawyers), they are the most valuable per-page copyrights I know. If John's attorney writes out the details of the house-painting contract between John and Mary and, rather than hire his own attorney, Bob just photocopies John's contract and modifies the parties and dates, John's attorney has a right to be paid.
Analogously, I have always been perplexed by claims that application programming interfaces are copyrightable, because I understand an API to be the method of a program communicating with another program - what bit goes where, what effect a call has on the system, etc. But it's clear from what I've read today that what many programmers call an API is a piece of code. According to this article, the court has said that piece of code, not the function it performs, is copyrightable.
That terminology confusion might be at the heart of much of the controversy over this case.
Posted May 18, 2014 1:07 UTC (Sun)
by dlang (guest, #313)
[Link] (6 responses)
what we are talking about here is the code that declares that the function min exists and that it takes two integers as it's arguments and returns an integer (with a comment that says that it returns the smallest of it's arguments)
This isn't the function that performs this check, just the definition that this function exists and how to call it.
If you are going to write a library to replace an existing library (so that programs written in the language can still work), then the name of the function, which library the function is in, and the type and number of the parameters and return value must be the same.
The only flexibility that you have is in the names of the arguments,
the claim here is that Google copied the names and organization of the functions, and that the comments explaining what the functions do sound very similar
well, the comments had better sound pretty similar, the functions do the same thing after all.
If Google were to organize the functions differently (put them in different libraries) or name them something different (minimum instead of min for example), it would break existing programs and not be the same API
I guess they could have listed the functions in each library in a random order instead of alphabetically, but that's not a very creative choice.
In the past, the courts have ruled that you can copy the minimum that you need in order to be interoperable. There was even a case where the entire program was copied, bit for bit, and the court ruled that it was legal (the printer manufacturer checked that the entire program existed, bit for bit, in the toner cartridge, in order for aftermarket cartridges to work, they needed to have the exact same bits in the exact same place, so the entire program was copied)
The big problem here is that people are looking and saying "that was a lot of effort to do, nobody else should be able to do the same thing without paying" and this completely ignores the fact that _everything_ that we do is built on top of work that was done by other people in the past.
This post http://blog.goodstuff.im/oracle_sweet_lord by David Pollak talks about how this ruling is a disaster, especially for Oracle.
He also goes on to talk about how everyone who writes and deploys any Java program is in violation of Oracle's API copyrights if this is upheld http://blog.goodstuff.im/we_all_infringe_redux
Posted May 18, 2014 19:51 UTC (Sun)
by giraffedata (guest, #1954)
[Link] (1 responses)
This is a pretty abstract leap of metaphor. And the court in this case and every other one I know of doesn't make it. They say what you said in the rest of your post: that there is a distinction between what a computer does and any piece of code and the former is not the subject of copyright law and the latter is.
The district court said all Google copied was what the computer does, so Google was in the clear. The appeals court said Google copied a bunch of pieces of code, so is not in the clear.
I found it hard to make sense of David Pollack's essay because of the terminology confusion in the word API that I previously mentioned. I note that the court used more specific terminology than "API" in its opinion.
The court seemed to make it clear that one could implement an API with original code and thereby not infringe, but found that Google didn't do that in this case.
(And by the way, the court didn't say copying the pieces of code was an infringement in this case - only that it could be. More trial required).
Posted May 19, 2014 1:49 UTC (Mon)
by dlang (guest, #313)
[Link]
So there's a lot of ground to cover in the appeals here.
Posted May 22, 2014 5:52 UTC (Thu)
by linuxrocks123 (subscriber, #34648)
[Link] (3 responses)
Posted May 22, 2014 6:43 UTC (Thu)
by dlang (guest, #313)
[Link] (2 responses)
A setback for Google against Oracle
A setback for Google against Oracle
A setback for Google against Oracle
anything that a computer does is a piece of code
A setback for Google against Oracle
A setback for Google against Oracle
A setback for Google against Oracle