@Blaise, yes, if there is evidence of textual copying, the bar for affirmative defenses for infringement should be very high.

AFAICT, this case didn't consider what the bar should be if you copied nothing at all, but simply wrote something compatible that happened to look similar. That's the bubble sort example I always use: we all know bubble sort so well that it's basically impossible for us to write bubble sort and have it not look the same as someone else's bubble sort. But, does that mean I've committed copyright infringement? I don't think the court considered that question at all, because of their false assumptions.

@pam, do you really think a jury is going to reject a fair use defense when considered in isolation?