I don't blame people for experimenting with new ways of visualizing the "scrolling" mechanic. (Though, it is so widely done these days that I can hardly give NP

I don't blame people for experimenting with new ways of visualizing the "scrolling" mechanic. (Though, it is so widely done these days that I can hardly give NPR One credit for "experimenting" in this particular case.) I think it's fun to see people testing new ways to interact. But if you do that, it had better be well implemented -- and that is clearly where this falls short. User input translates into on-screen behavior inconsistently, arrow keys are basically broken. When you're trying to replace something so basic as scrolling, this is an interesting case study in just how frustrated users will get with even small frustrations. I've found that even the (now) more typical load-on-demand pages drive me nuts when there's no more than a momentary hitch in the scrolling to load more content. Though maybe I'm just easily bothered by such things.