Note 30

On Why IPv6 Will Never Happen

What would a migration have to pay back before real networks adopt it?

Lu Heng · January 12, 2026 · 2 min read

Conceptual model of two equipment sets serving one facility, each with its own tools and blue cable route. — The paired equipment is a metaphor for maintaining two protocols, not a hardware requirement. Lu Heng questions whether the additional operating work earns a corresponding return.

IPv6 will never happen in the way the RIR world promised.

Not because IPv6 is “bad tech.” IPv6 works. The reason is simpler: it is against the economic interest of ISPs, cloud providers, and real networks.

Networks are not charities. They survive on thin margins, constant capex, and endless operational risk. Any migration that increases complexity must pay back in revenue or cost reduction. IPv6 does neither at the scale required.

IPv6 adds a second universe. Dual-stack is not “transition,” it is permanent overhead: more routing, more security rules, more tooling, more training, more edge cases, more outages. You pay twice and earn the same.

And the biggest issue is scarcity.

IPv6 removes scarcity by design. No scarcity means no price power. No price discovery. No asset value. It pushes networks back into infinite supply where the only competition is price-to-zero.

IPv4 scarcity is the opposite. It creates leverage. It turns IP into capital. It lets networks own scarce digital land, compound it, and escape the poverty line business model of pure bandwidth labor.

That is why, after twenty years of preaching, IPv6 still did not replace IPv4. Because operators do not vote with slogans. They vote with incentives.

Even the “IPv6 supporters” inside governance circles often treat it as social signaling. It is the correct sentence to say in the room. It keeps you inside the club. But it does not change what operators actually do when the incentives are wrong.

So what will happen instead is what is already happening.

IPv6 will exist where it is forced, where it is easy, or where it is free. It will be used as an add-on. It will grow slowly at the edges. But it will not eliminate IPv4, because eliminating IPv4 eliminates the most valuable market mechanism networks have ever been offered.

Scarcity is not a bug. Scarcity is the system creating capital.

So the real future is not “IPv6 everywhere.” The real future is IPv4 as a mature asset market, with leasing, pricing, and ownership reality shaping behavior.

Operators should stop pretending otherwise.

Stop apologizing for scarcity.

Stop donating your leverage to people who don’t pay your bills.

Treat IPv4 as capital, and build as owners.

Where this Note sits

What does the design make possible?

This Note's focus

That is why, after twenty years of preaching, IPv6 still did not replace IPv4. Because operators do not vote with slogans. They vote with incentives.

This question in the exhibition

“The Internet works because no one owns it.”Hall 3 · The World We BuildA City in Use Note 2

Reading map

Before you start

A protocol is a set of working rules, not a constitution that silently grants authority. Separate what must be coordinated from what can be decided elsewhere.

The problem
A system’s formal language can hide who may act, who can be overruled, and which assumptions have already become political power.
The proposed direction
Describe roles, limits, exit routes, and accountability as part of the architecture itself, rather than leaving them to custom or rhetoric.
Why it matters now
Technical choices harden into public rules as soon as people depend on them, so the time to make their consequences visible is before dependence becomes permanent.