In message <199511062018.PAA08597@home.merit.edu>, "Steven J. Richardson" write
s:
Uh... Michael, when we were running the NSFNET, as Hans-Werner and
many readers of this list are well aware, we did _not_ accept 10% packet
loss on any link or across the network. These problems stayed with
the NSFNET NOC until resolution by the provider, MCI. We only considered
-0%- loss to be acceptable.
Steve,
Enough of your wild stories of -0%- loss. 
The correct figure was
10^-5 for acceptance with 10^-4 being the maximum threshold we would
accept on a running circuit before contacting MCI to take the circuit
in a maintenance window for diagnostics. That doesn't mean we
wouldn't bug MCI to get the circuits back perfectly clean. 
We still have the same criteria. I think MCInet is also as vigilant.
Curtis