I think the failure isn't IPv6 itself, but the dreadful prospect of having to step from the boat to the dock: simultaneous production implementations of ipv4 an

I think the failure isn't IPv6 itself, but the dreadful prospect of having to step from the boat to the dock: simultaneous production implementations of ipv4 and IPv6 in the same organization. Mostly this is a DNS headache, especially if you're already segmented into a bunch of different split horizons due to organizational boundaries or vpns or what have you, and unlike software networking is almost necessarily done direct to production as staging is extraordinarily complex and expensive, and again unlike software development is not part of the daily iteration cycle but instead of once or twice in a decade type of thing. anyone managing or responsible for managing that infrastructure is going to need an extremely compelling reason to pursue a switch thanks to the two systems obstacle.