> I never heard of this "0.0.0.0" thing before today, and frankly don't
> understand why we are making any effort to support it.

The idea that 0.0.0.0 means "this host on this network" is discussed in
RFC 1122 (host requirements) section 3.2.1.3, and RFC 1812 (router
requirements) section 4.2.2.11.  Both those RFCs are quite clear about
those addresses never being sent or accepted over the network except in
special circumstances (such as boootstrapping).  Neither of them appears
to say anything about support for the 0.0.0.0 address in local connections
where both source and destination are inside the same host, and the
packets don't cross any network.

--apb (Alan Barrett)