Comment 59 for bug 878820

I would love to see evidence from an actual user test showing that the original behavior was "confusing", because I'm skeptical that any such evidence exists.

Can somebody find even *one* user who was "confused"?