Hard to imagine they estimate about a year's worth left until exhaustion. And IPv6 isn't remotely close to being fully adopted.
Is it safe to assume that IPv4 would continue to be used even after exhaustion, but we would need to start using IPv6 for any further growth? Is there any inherent problems with a system with mixed addresses?
Popular centralized services (name servers, certain web servers) will want to stick with IPv4 for a long time. They need to support customers who cannot or will not upgrade their clients.
Edge services (ISP customers, file sharing) can easily move to IPv6 as soon as their ISPs come to their senses.
Currently, TLS/SSL consumes an IP address for every secure web site. Does anybody think we will see virtual hosting of HTTPS using the UPGRADE method? Or will IPv6 win out before this becomes necessary?
Is it safe to assume that IPv4 would continue to be used even after exhaustion, but we would need to start using IPv6 for any further growth? Is there any inherent problems with a system with mixed addresses?