Interesting, thanks! Probably not as cool, but you could also just get the output of the regular top, parse it and generate HTML from it (a bit like http://tracesof.net/uebersicht/)