Mark Andrews wrote: > Well draft-omar-ipv10-XX and draft-omar-ipmix-XX provide NOTHING THAT IS NEW. > You don’t even summarise the existing technology accurately. Traffic will > switch over to IPv6. All that is required is time. Large amounts of traffic > already goes from IPv6 hosts at one end to IPv4 hosts at the other end. Some > of that is initiated from IPv4, some from IPv6. Most equipment that you buy > today is IPv6 capable. More to the point, draft-omar-ipv10 requires substantial protocol updates on not just all operating systems but also all TCP/IP APIs and applications, and the application level support is no more backwards compatible than native ipv6 support. This is not viable because the amount of time and effort required to make "omar-ipv10" work at all would be better spent on making ipv6 work better. In practice, it means that even if draft-omar-ipv10 worked at a technical level (and there are no working prototypes), it would be undeployable. > As for draft-omar-krp-XX, it is completely unrealistic. I suggest that you > talk to network operators to understand why. KRP seems to require hierarchical routing, and hierarchical routing is not how the internet works. In other words, both draft-omar-ipv10 and draft-omar-krp are unworkable. This is why no-one is interested in discussing them in any of the working groups. Nick