A conversation with a Fediverse core developer revealed the structural problem: donation models fail, and even specialized contributors can't make a living. Infrastructure costs are real, but open protocols can't paywall features—you can't compete with free providers.
Only three realistic options exist:
Full decentralization + donations → developer poverty & slow collapse
Centralized infra + open protocol → de facto centralization
Public funding / municipal operation → political dependence risk
Bluesky/AT Protocol faces the same dilemma. The ideal of PDS federation vs. the reality of Relay consolidation. "Fair competition" is structurally unsustainable—network effects ensure the largest provider becomes the de facto standard.
For a Japan-specific network, combining options 2 & 3 is realistic: depend on Bluesky Social infrastructure while building Japan-focused service layers (corporate consortia/municipal partnerships), differentiating through cultural specificity (moderation standards, legal compliance).
This isn't ideological defeat—it's reconciliation with reality. The question isn't "how to avoid dominance" but "whose dominance, of what quality, do we accept?" Preserving Public Benefit Corporation structure, community governance participation, and technical guarantees of forkability become the last line of defense.
The tragedy: we haven't found a third way. Open networks may be fundamentally incompatible with sustainable economics—at least not yet.
postscript:
AT Protocol Monetization Structure Analysis
Conversation with Fediverse core developer revealed: open networks structurally contradict economic sustainability.
AT Protocol layer monetization viability:
Unit A (Integrated: Bluesky Social) PDS+Client+Relay/AppView → ✅Easy to monetize Subscriptions, custom domains, ads enable lock-in
Unit B (Infrastructure: Independent Relay/AppView) Relay alone → ❌Hardest to monetize Resource-intensive but no clear customer. Caught between upstream (PDS) and downstream (AppView) with no revenue source. Middle-layer dilemma: valuable but unmonetizable.
Unit C (Data: Independent PDS) PDS alone → △Limited viability Niche market (enterprise/community) only. 10-account limit prevents scale.
Conclusion: Decentralized design, yet economic incentives concentrate centrally. Independent Relay operation becomes charity. Mixi model (cross-subsidization) doesn't work. For Japan-specific networks: enter as Unit A or target Unit C niche. Avoid Unit B.
Reconciliation between open network ideals and economic reality remains unsolved.
postscript 2:
AT Protocol economic structure: Unit A/B/C distinction is unrealistic. Only flagship (A) and satellite PDS (C) exist in practice.
Independent Relay (B) theoretically possible but economically unsustainable. Unlike utilities: no monopoly, no customers, easily replaceable. Open protocols intentionally destroy monopolies, making infrastructure layer monetization structurally impossible.
Result: Despite "decentralization" rhetoric, converges to flagship+satellite two-tier model. Bluesky Social dominance inevitable. Only question: quality of that dominance.