IPv6 demonstrates a real flaw in the IETF. It has little ability to throw away a standard that suffers from flaws that aren't of a technical nature and start over. It's been clear for a long time that IPv6 was suffering from huge adoption barriers. There have been countless IETF transition standards proposed, adopted and deployed meant as stopgaps for ip6 like CGNAT while 6 continues to languish. Carriers, an obvious critical stakeholder in protocol adoption, didn't have an opportunity to participate in the standards process at the time. Yet we continue cry that the problem is the people that won't adopt it, not the protocol that people won't adopt. DNSSEC has suffered from a similar but less dramatic market failure.
Why should we expect carriers to adopt IPv6+1 any quicker than IPv6? Are there specific elements of IPv6 that are so carrier-unfriendly that correcting them would be worth throwing away all the progress that's been made?
Because we'd be talking about something closer to ipv4+1, something that allowed core equipment to not have to process packets in a different way and maintain completely separate routing tables for granularity of traffic that only matters at the edge.
