If it results in companies actually spending the time to design better APIs I would absolutely support them being copyrighted.
Millions/billions of dollars is wasted each year because of terribly designed APIs. And I am a massive fan of a future where every system (internal and external) has a clean, interoperable API to build next-generation mashups. But right now that is a pipe dream.
Would this mean that it would be dangerous to read API documentation for potential competitors? Would it mean that you'd be in danger of violating copyright if you implemented a compatible implementation of someone else's API? Would it mean that you could get a copyright on an API that you designed but never implemented?
The argument as I read it seemed to be attempting to assert that implementing an API was a derivative work of that API's documentation.
It's hard to imagine the second and third order effects of such a change in the understanding of law around software. I'm trying not to be just change averse and actually imagining what that world would look like, but damn it's hard.
Oracle goofed and premised a lot of their copyright arguments on something that was not supported by their own copyright registrations. Unless they can find some documents that prove that they have registered what they need to, they're in big trouble, because it's too late to fix things.
I think the judge said they would be allowed to introduce any new evidence, assuming they can find anything that says they own what they need to. But unless they can do that....
