> : Wouldn't it be better to define a minimal time between two redirects
> : sent to the same machine? If we don't it, we can simply wast a substantial
> : part of network bandwidth by sending redirects for an each packet received
> : from a misconfigured host. What do you think of it?
>
> It is not only good, we must make it. It is one of RFC requirements.
OK. Is anyone going to implement it or shall I do it instead?
Martin