Re: unreliability

(D. J. Bernstein wrote:

> Reliability (like security) isn't a game of ``Gee,
> I don't _know_ any problems, so it's okay.''

I don't think so. A secure system is one that's unlikely to be broken by
someone who's trying to break it: i.e., a fairly unpredictable force. A
reliable system is just one that's unlikely to break. Sure, it's hard to
test, but a really good pragmatic test, if you can manage it, is enough.

Jason Grossman