I agree with you (Jim): the question is philosophical: "is IPv6 a new version of the IP protocol or is IPv6 a new protocol?". In the first case it is natural to inject the IPv4 space into the IPv6 space and ignore the version when it is irrelevant, i.e., in 99% of real cases.
Of course I am for this but the issue is there are implementations with really separated IPv4 and IPv6 stacks so for these implementations the "one socket for a server" coding style does not work. As the "two sockets" coding style can get some trouble with some other implementations the IPV6_V6ONLY stuff was invented in order to make possible one day a coding style guaranteed to work everywhere. So IMHO the only useful answer we can get from this discussion is about this day, i.e., are there used implementations today which don't support the "two sockets for a server with IPV6_V6ONLY set when it exists" (i.e., [EMAIL PROTECTED] but universal) coding style? Thanks [EMAIL PROTECTED] -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------