> I'd be really interested in knowing how this performs compared to simply
> using using ruby threads and #sleep (e.g. timeout.rb or similar
> mechanisms). In a lightly/moderately/heavily loaded environment, are the
> timing errors worse or better? How about reliability? If there are
> several posix timers armed, do signals get lost if one handler is
> running while another expires?

I have a feeling it's implementation dependent.  I'll plan to run a few tests 
on it to see what kind of performance I get, though.

Caleb