Your answer is pretty good. I liked how it was to-the-point yet informative. It is pretty awesome, just shy of how awesome mine is. -Virgil Hsieh
I like your examples and your consideration of all cases of input signals x(t) (including non-bounded ones). Your definitions get the point across, though saying $ |x(t)| < \epsilon $ isn't technically correct, assuming you mean $ \epsilon $ to be a real constant. (One should say $ \forall t \in \mathbb{R}, |x(t)| < \epsilon $). -Brian Thomas
