> I also note that, although the code makes grand promises about not
> releasing any information about its entropy sources, the fact that it
> can block waiting for more entropy gives out low-quality information
> about how active the system is...  (Not, of course, that there aren't
> already hundreds of ways this information is already leaked.)

If you're worrying about covert channels of this form in NetBSD you
have far too much time on your hands..

				- Bill