The unrolling of IPv6 (draft)

This is part of my series on [ (Trying to) foresee IPv6 deployment]. IANA has exhausted its IPv4 pool. As I'm writing, this is only a matter of days before APNIC, the first RIR to be depleted, exhausts its own IPv4 supply. The question in the back of the mind of everybody working on the IPv4 to IPv6 transition (including myself) is: when and what will be the first IPv6 only machine connected to the Internet. By that, I don't mean some lab experiment where someone intentionally left an IPv6 with no IPv4 counter part. By the first machine, I mean a host connected to the Internet through IPv6, that its network operator would have connected using IPv4, had it had the choice. After that event, IPv6 deployment should develop in logical phases. == Waiting for the spark == Everyone in IT who reads the news should now know that it's time to act and jump into the IPv6 train. However, the IPv6 problem is that costs and risks are easy to see, but a compelling business case has yet to be made. Thus many people are dragging their feet on the one hand, but It's coming nonetheless on the other. Now if you have to make decision about IPv6 deployment, it could help to have a vision of '''how''' it will happen. And here are == My facts and speculations == * [[ipv6-big-content-vs-important-content|Big Content vs important content]] * [[ipv6-and-chinese-student-doing-skype-grandpa|Regional core and the spreading of IPv6: the Chinese student]] * [[is-google-boosting-ipv6-adoption|Is Google boosting IPv6 adoption ?]] * [ IPv6 momentum: a story of pool and poules]: can we "segment" the IPv6 market and try to predict the path to adoption ? == More to come == * IPv6 on 3G is the next phase ? I very much doubt so. LTE is years away, and 3GPP dual-stack technology is still wet. * IRC, (certificate validation ?) * Small businesses may be hurt first * Will IPv6 kill the VPN business ? * DNS, mail: the role of the infrastructure.