> If IPv6 were wholeheartedly "adopted" then I and most consumers would not require "dual-stack" at home. Nah. We're not going to shut down IPv4 service for ma

> If IPv6 were wholeheartedly "adopted" then I and most consumers would not require "dual-stack" at home. Nah. We're not going to shut down IPv4 service for many, many decades. IPv6 is primarily an global-address-space-pressure-relief mechanism. In the next five, ten, fifteen years, enough end-user sites will have IPv6 GUA service that putting every end-user behind an IPv4 CGN and reducing the IPv4 GUA to end-user ratio to 50:1 or more will be a quite reasonable thing to do. Any complaints you have about what you believe to be added complexity are mooted by one or more of the following: 1) End-users typically plug the magic internet router into the ISP-provided magic internet cable or box. They then magically get internet access in their house. They typically have no understanding of how any of this works. 2) Because the firewalls provided by Windows and OSX primarily care about which application is doing the network traffic, they're inherently IP version agnostic. I know that on the Linux side of things, nftables has been available for a decade and -by default- gives zero shits about the IP version of the traffic it's inspecting. So, for the three major consumer OSs it's easy to write firewall rules exactly once and have them apply regardless of what version of IP carries the traffic. 3) The major IP implementations were written years ago. That work's already been done. Removing one of those stacks is even more work.