On Tue, 26 May 2020 at 09:52, Daniel Ankers <md1...@md1clv.com> wrote: > The thing about Y2K and 2038 is that they are absolutely fixed dates. No > amount of arguing or pleading would move them. On the other hand, if a flag > day for IPv4 shutoff was chosen it would be arbitrary and could, if needed, > be moved. While the vast majority of the internet is IPv4 first there will > be pressure to move the date (and I believe that pressure will be too strong > to resist), and if people think the date might be moved then they won't > migrate to IPv6.
On Tue, 26 May 2020 at 12:17, Nick Hilliard <n...@foobar.org> wrote: > ipv4 will fade when it becomes more expensive and troublesome than ipv6. +1 look at the efforts to deprecate python2 and the number of times the deadline has been pushed back. and how much microsoft were able to charge for keeping windows 7 on life support. At the moment, having IPv4 is an externality for most customers of ISPs, but if it becomes an explicit part of the bill, much like many ISPs charge for static IPv4 addresses, then perhaps most customers would accept having to use dns64/nat64 via their provider's nat64 gateway and that would accelerate the change? I wonder how content filtering/logging for anti-piracy/illegal porn will be affected by that, it will surely be harder to disentangle all the customer IPv4 traffic?
