On Mar 11, 2022, at 12:20 PM, John Levine <johnl@iecc.com> wrote:
It appears that Joe Maimon <jmaimon@jmaimon.com> said:
higher penetration of native v6, I would restate that a bit more conservatively as
Google's statistics are likely a fair barometer for USA usage in the large content provider arena which have a strong mobile representation.
AT&T, Comcast, and Charter/Spectrum, the three largest cable companies, have IPv6 support.
As do (so I hear) mobile providers, which is increasingly how people around the world get access to the Internet. However, this discussion has drifted a bit — it wasn’t (supposed to be) a discussion about IPv6 deployment per se, but rather network operations reality as they impact IPv6 deployment. There was an assertion (that I am not questioning) that there are various kit vendors who claim IPv6 support, but when network operators attempt to deploy that kit, the IPv6 support is found to be show-stoppingly buggy, lacking in required features, or otherwise causing said network operators frustration/irritation/etc and/or to give up on deploying IPv6 “until it is more mature” (or “more/any customers demand it”). For whatever reason, there appears to be a reluctance to name names in such cases. My question was whether it might be helpful in encouraging IPv6 deployment (or at least reducing the amount of disappointment) for network operators to be more public when reality does not match vendor claims, just as “timed full disclosure” has helped in addressing (some) security-related issues. Regards, -drc