I'm not sure why so many people here seem to be surprised by this, I got the exact same impression from the hearing.
The problem for Google on the copyrightability front is that "compilations" of non-copyrightable items can be copyrightable even if the underlying items themselves are not, if the "selection, coordination, or arrangement" of those items involves sufficient creativity to be considered an "original work of authorship".
For example, individual recipes are generally not copyrightable, but if you compile a set of recipes and publish them in a cookbook, the cookbook itself becomes copyrightable.
As applied to Java, individual method signatures like min and max might not be copyrightable, but someone had to go in and organize those individual methods into classes, and organize those classes into packages. That act of organization likely implies that the Java API (taken as a whole) constitutes a copyrightable work, even if the individual method signatures are not.
What's worse for Google is that code has always been considered copyrightable so arguing that APIs are non-copyrightable is essentially carving out an exception for a specific type of code.
As is common when HN discusses legal matters I see a lot of people blaming the judges for not understanding tech, but the problem here isn't really the judges so much as the law itself. We should not be surprised if Oracle ends up winning here, at least on the question of copyrightability. Google may have an out related to fair use, but that seems like a long shot as well.
The problem for Google on the copyrightability front is that "compilations" of non-copyrightable items can be copyrightable even if the underlying items themselves are not, if the "selection, coordination, or arrangement" of those items involves sufficient creativity to be considered an "original work of authorship".
For example, individual recipes are generally not copyrightable, but if you compile a set of recipes and publish them in a cookbook, the cookbook itself becomes copyrightable.
As applied to Java, individual method signatures like min and max might not be copyrightable, but someone had to go in and organize those individual methods into classes, and organize those classes into packages. That act of organization likely implies that the Java API (taken as a whole) constitutes a copyrightable work, even if the individual method signatures are not.
What's worse for Google is that code has always been considered copyrightable so arguing that APIs are non-copyrightable is essentially carving out an exception for a specific type of code.
As is common when HN discusses legal matters I see a lot of people blaming the judges for not understanding tech, but the problem here isn't really the judges so much as the law itself. We should not be surprised if Oracle ends up winning here, at least on the question of copyrightability. Google may have an out related to fair use, but that seems like a long shot as well.