As I wrote: > As I already pointed out in the past here, having "interface > identifier" is a feature inherited from XNS having running code. > > But, that XNS and IPv6 have running code dose not mean they are > good. On 2021/11/03 1:45, Vasilenko Eduard wrote:
+1. It was a big mistake to break OSI model and include L2 address (MAC) inside L3 address (IPv6). Half of the address bits were wasted. Many people are coming and coming asking that they would like too to get some bits of IPv6 address for some new protocol. Eduard
Indeed. Originally, as proposed by IPng directorates, interface ID was 48bit long. However, as I pointed out that IEEE1394 has 64bit MAC, it was extended to be 64bit long. But, I haven't noticed that it is layer violation. Thank you. Masataka Ohta
