| This should all be part of QoS under NetBSD.

I don't really see what QOS has to do with this discussion.
That sounds very seperate to me.

| Perhaps a more general purpose "Event Per Quantum" needs to be developed
| at this point so it can be utilized by more than just ICMP.

If you look at the implementation, you'll see that they are
all sufficiently generic (ppsratecheck() and ratecheck())
that they are not constrained to packets or ICMP errors,
and should work for any event.

--jhawk