Post by @petrillic@hachyderm.io

2026-09-25 22:35 UTC

@paul_ipv6@infosec.exchange @wendynather@infosec.exchange it was. And PIP (our proposal) was the super future looking. So they decided to go with a monstrosity composed of a bunch of different things none of which are wrong necessarily but which don’t tell a cohesive story. Combine that with the introduction of NAT and… well here we are 30 years later.

Replies (2)

@paul_ipv6@infosec.exchange 2026-09-25 22:44

@petrillic@hachyderm.io @wendynather@infosec.exchange yup. i think that certain things in IPv6 were improvements and did let us learn from some of the lessons of IPv4. that said, it also suffered from too many people who decided it that if we'd done it in v4, it must be wrong and needed to be ripped out. that, plus the attitude that "we're smarter than you and if you think you need something like DHCP, you must be an idiot and you're running your network wrong" didn't help with enterprise or large vendors. as you say, the IETF saying "NAT is evil so we won't even try to support it", yet it did solve lots of the issues for lots of folks, gave the IETF a credibilty problem and NAT a solid foot in the door. we're doing better now than we were 25 years ago but i sure hope we've learned some lessons.

Open ##5099047
@bsdphk@fosstodon.org 2026-09-26 07:19

@petrillic@hachyderm.io @paul_ipv6@infosec.exchange @wendynather@infosec.exchange In hindsight, it is obvious that there were IPng proposals and concepts NSA really didn't want to happen. For instance flow-routing would have made their "collect everything" a lot more expensive. Same, probably to a lesser degree with multi-homing.

Open ##5099048