团队文章权力与治理

The Growing Geopolitical and Legal Fragility of IPv4 Address Markets

Why a cross-border IPv4 transaction depends on more than a price: separate the agreement, registry record and working network, then examine the points of dependence.

目录

Working network cabinets span separate platforms with uneven joints between them.
A network can cross several administrative boundaries. Its continuity depends on how those relationships hold together when one changes.

A company can agree a price for an IPv4 block and still face a harder question: will the addresses remain usable through a change of provider, a disputed record or the failure of an administrator? The transaction happens between particular parties. The service built on it depends on a wider set of relationships.

That is the fragility this article examines. The important distinction is between acquiring an asset, recording responsibility for it and keeping a network operating. Treating those as one event hides where another institution can influence continuity.

Start with a service crossing several boundaries

Imagine a hosting business incorporated in one country, using a provider in another, with customers in several more. Its addresses are associated with a registry record, its routes are announced by a network, and its customers may have put those addresses into firewall rules and application settings.

A change can affect one of these relationships without changing all the others. A provider can change while the recorded holder stays the same. A transfer can update the recorded holder without moving the equipment. A customer's dependency can persist after both changes.

This is why the physical location of traffic, a registry's service region and the legal relationship between two companies must be understood separately. Drawing one boundary around them creates a certainty the situation does not provide.

Three questions that should stay separate

The agreement asks what the parties have agreed to provide and receive. The registry record asks who is recorded as responsible for the number resource. The running network asks which systems actually carry traffic and which other networks accept its announcements.

Each matters. None automatically settles every question raised by the other two. A purchase document does not configure a route. A route that works does not explain the entire history of a resource. An administrator's database does not, simply by existing, settle every claim about the activity recorded in it.

Proof of control is useful when the claim is specific. Evidence should identify what it establishes: access to an account, authority for a particular change, a routing relationship or a documented transaction. An all-purpose claim of control can hide the very gaps a business needs to understand.

The institutional risk is dependence without an exit

In Note 62, Lu Heng argues that a limited coordinating role has been presented as a much larger political mandate. A service footprint becomes a claim to speak for a region; an internal process becomes a claim to decide for people outside the room.

The commercial consequence is practical. Operators build services and bear the cost of interruption, while an upstream institution can influence the records on which those services depend. If the operator cannot leave without losing recognition or rebuilding customer relationships, the administrative service has become difficult to replace.

This explains why geography matters in his critique. Organising a registry's service area is one thing. Treating that area as a source of ownership or political authority over resources is another. The first does not establish the second.

Scarcity is not a reason to expand administrative power

Discussions of IPv4 often move too quickly from a finite address space to claims about inevitable price increases or unavoidable operational shortage. Those are different claims and need different evidence.

In Note 45, Lu Heng challenges the idea that exhaustion of a registry's unallocated pool proves that addresses are the binding economic constraint on a running service. He asks readers to examine actual cost, utility and access instead of treating an administrative label as the conclusion.

The relevance here is institutional: even when a resource has substantial value, that value does not explain why its recordkeeper should gain broader discretion over commercial life. Note 72 argues for a narrower duty: preserve uniqueness, accurate records and continuity.

Choosing IPv4, IPv6 or a combination involves the requirements of the service being operated. A change of protocol does not, by itself, answer who may expand authority at the coordination layer. The governance question remains a separate one.

Where a cross-border relationship can become fragile

Consider a provider that becomes unavailable. The immediate problem may be access to contacts, records or operational assistance. Consider a disagreement over a transfer. The issue may instead be whether the recognised record reflects the documented change. Consider conflicting claims about responsibility. The necessary evidence may sit with several organisations.

These are different failure paths. Calling all of them “geopolitical risk” makes the problem sound large while making the next question harder to identify.

A useful account names the dependency: which party performs which function, which record others rely on, which change is disputed and what must remain true for customers to keep connecting. Questions about a particular agreement or legal dispute require their own facts; they cannot be answered by a universal statement that a registry owns everything or that a purchase removes every dependency.

What a more resilient coordination layer would preserve

Lu Heng's proposed direction is a thin common layer that protects the facts networks need in order to interoperate. Unique resources must remain distinguishable. Control claims and changes must be verifiable. History must remain readable. Replacing an institution must not require destroying the useful relationships built around a resource.

In Running-Code Primacy, he takes the idea further: validity should be determined through locally verifiable rules and distributed state, rather than the continuing discretion of one indispensable recordkeeper. Later change depends on implementation and adoption.

That is a direction for system design, not a claim that present-day networks will accept any exported record. The transition must preserve security, avoid conflicting claims and provide evidence counterparties can actually use.

A concrete place to begin

Follow one address block through the business. Identify its recorded holder, its operational user, the network originating its route, the relevant authorisations and the people who can explain its history. Then consider one change: the provider leaves, the holder changes or the current recordkeeper becomes unavailable.

Which parts continue? Which require someone else's cooperation? Which cannot presently be replaced? Those answers locate the fragility more precisely than a forecast about the whole IPv4 market.

The long-term question is how to preserve useful coordination without turning dependence into permanent authority. Continue with the different states in an address resource's lifecycle and why a transfer process should protect record accuracy.