MAC address duplication within N2N can lead to the same problems that occur in ethernet networks so care must be taken to avoid conﬂicts. Edge nodes can have dynamic IP addresses by means of a DHCP server attached to an edge node in a community. As N2N TAP interfaces behave like real (e.g. ethernet) interfaces, it is also possible to run other services such as those deﬁned by the IETF Zeroconf Working Group including, but not limited to, multicast DNS (mDNS) and DNS Service discovery . Unlike most P2P systems, a node naming scheme  for locating peers is redundant in N2N.4. N2N Evaluation and TestingAs N2N is a technology that is designed to interconnect heterogeneous computers, the authors ported it to three common platforms: Linux, MacOS X including BSD variants, and Windows. From the software point of view the code base is the same. Platform-speciﬁc code was needed for supporting the various TAP APIs. The authors acknowledge that the OpenVPN project has done a signiﬁcant amount of work, so that today it is relatively simple to use TAP devices in a multivendor environment.In the test setup, the supernode was installed on a Linux PC with a public IP address although this is not really necessary as long as the supernode UDP port is publicallyreachable. N2N nodes were placed behind several types of NAT/ﬁrewall devices, including symmetric NAT and multi-NAT (i.e. cascaded NAT devices) that are often used by ISPs. Edge nodes have been used on all above listed platforms in order to evaluate interoperability. As soon as the edge application starts, the node is part of the N2N community and can communicate with remote N2N peers. Several protocols were tested successfully, ranging from SSH to dynamic FTP.