Six More Amendments: Give Power to the People Who Do the Work

My previous proposal ("Six Amendments That Enforce Themselves") was well received, and the thread pushed it somewhere better — one commenter proposed pre-vote impact statements on the same registrar rails, and asked whether ordinary registrants, who ship nothing but pay the renewals that fund the share of operating costs the Endowment doesn't cover (kpk's 2025 figures, per the thread), deserve a seat. They do. This proposal starts from that concession and goes further: if the last one was about making power checkable, this one is about making power earned. Six more amendments, each borrowed from a real experiment somewhere — including the failures, because the failures teach more. The question underneath the custody fight Strip away the org charts and this debate is one question: who should hold power over ENS — and every current answer says "whoever holds tokens." Labs' proposal moves token-granted power to a board; most amendments (mine included) keep token-granted power with the DAO. But tokens measure purchase, not contribution. The people with the strongest claim — engineers who ship the protocol, service providers who deliver, registrants who pay renewals year after year — hold power only by coincidence of token ownership. Both camps in this fight are defending different arrangements of the same flawed measure. These amendments experiment with better measures. All six are time-boxed pilots with published success metrics and auto-sunset — in the spirit of my last proposal, each experiment carries the means of its own evaluation. Amendment 7 — A second chamber measured in work, not tokens. Create an Earned Chamber alongside token voting: membership from verifiable contribution — merged protocol PRs, authored ENSIPs, completed SPP deliverables, working-group service — and, per this thread, sustained registrant patronage: years of renewals as a qualification in its own right, with at least one seat per cycle drawn by lot from verified long-standing renewers. Proof of payment counts, not just proof of code; registrants are the constituency that funds the gap the Endowment doesn't cover, and today they hold exactly zero structural voice. Two design rules from others' scar tissue: membership is non-transferable (you can't buy your way in), and it decays — Colony's insight that reputation must fade so power reflects recent work, not ancient glory. Structural changes then require concurrence of both chambers: token holders and proven contributors, checking each other. This is Optimism's two-house principle, but with the second house defined by work rather than citizenship badges. Amendment 8 — Pre-vote impact statements for anything touching what registrants pay. Adopted directly from the thread: any resolution changing registrant-facing pricing, renewal mechanics, or resolution service must publish, before the vote and on the same registrar rails from my last proposal, a machine-readable impact statement — cost change per name-year by tier, active names affected, alternatives rejected — timestamped ahead of the resolution, not reported after it. The payer's exposure becomes resolvable before anyone votes to impose it. Credit where due: this is the commenter's mechanism, and it costs nothing extra to build. Amendment 9 — Conviction voting for the small-grants stream. For grants below a threshold, replace snapshot votes with conviction voting, the mechanism 1Hive has run in production for years: supporters stake preference on proposals continuously, conviction accumulates the longer they hold it, and funding releases automatically when accumulated conviction crosses a bar scaled to the ask. No campaigning windows, no vote fatigue, no committee gatekeeping the small stuff — sustained genuine interest, not momentary attention, moves money. Pilot it on a capped envelope from the ecosystem stream; measure spam rate and time-to-funding against the current process. Amendment 10 — One futarchy pilot a year, with a metric chosen like it matters. Where a decision has a genuinely measurable outcome, let a prediction market advise it: Optimism ran exactly this in March 2025, distributing 500k OP through conditional markets. Their honest lesson: the experiment stumbled because the chosen metric (USD-denominated TVL) moved with ETH's price, not project performance. So the amendment is narrow: one decision per year, advisory-only at first, with the metric ratified in advance by both chambers — denominated in protocol-native terms (renewal retention, name growth) that can't be swamped by market beta. If the market's advice beats the counterfactual over two cycles, expand it; if not, it sunsets. Markets as instruments, not rulers. Amendment 11 — A room with chairs: in-person deliberation before constitutional-scale money moves. James of Fire Eyes called for a facilitated roundtable; this council lists one under alternatives; and there's a reason Taiwan's vTaiwan process and citizens' assemblies keep producing consensus where forums produce factions — deliberation changes minds, voting only counts them. Amendment: any resolution moving more than a set share of the Endowment requires, before the vote, a facilitated in-person assembly — sortition-selected mix of contributors, registrants, Labs, and delegates, convened where ENS already gathers (Devconnect, EthCC), with remote participation for those who can't travel, producing a published findings document the eventual onchain vote must formally acknowledge. Not binding — sovereignty stays with the vote — but no nine-figure decision should be made by people who have never sat in a room together. The forum thread this whole gathering responds to is itself the evidence: text amplifies the loudest, deliberation surfaces the middle. Amendment 12 — Exit rights, learning from the $27M mistake. The deepest check on any government is that people can leave. Nouns DAO built this literally — fork with your pro-rata treasury share — and its first fork drained roughly $27M in days, mostly to arbitrageurs, not dissenters building anew. The lesson isn't "never build exits"; it's that treasury-coupled exit invites looting. ENS's version should be exit of the protocol, not the purse: a standing, DAO-ratified Forkability Guarantee — all critical resolution infrastructure open-source and reproducible, registry data continuously exportable, no dependency that couldn't be re-stood-up by a successor community — audited annually and attested on the same registrar rails. The community's ultimate leverage over both DAO and Foundation isn't taking the money; it's the credible ability to take the names and walk. A Foundation that knows the community can credibly fork is a Foundation with permanent incentive to stay worth not forking. Why these six belong together Amendments 7 and 8 redistribute voice toward the people who do the work and pay the bills. Amendments 9 and 10 move decision-making off pure token weight onto mechanisms — time, conviction, markets — that reward being right over being rich. Amendment 11 adds the one medium where humans actually change their minds. Amendment 12 is the backstop that disciplines everything above it. None is permanent: every pilot ships with its success metric published in advance and an auto-sunset if it fails to clear it — the same self-verifying standard as my last proposal, applied to the experiments themselves. Trade-offs and what would change my mind Earned chambers can gatekeep: whoever defines "contribution" holds real power, and decay parameters are politics wearing math's clothes — the definitions must be ratified by token vote at the start, which partially reimports the problem, and I don't have a cleaner answer. Conviction voting has mostly run at smaller treasuries than ENS's; scale may surface failure modes 1Hive never hit. Futarchy's record is genuinely mixed — I've cited its stumble on purpose, and if the pilot's advisory record is bad, the sunset does its job. In-person assemblies cost real money and can be captured by whoever writes the agenda; independent facilitation is load-bearing and I've priced that honestly as a risk, not a detail. And if someone demonstrates that the Forkability Guarantee is technically hollow — that some dependency can't be made reproducible — then Amendment 12 is theater, and I'd rather know now. The single thing that would most change my mind: evidence that plural chambers produce deadlock rather than balance in DAOs that tried them. If that evidence exists, bring it to the thread; deadlock at ENS's scale would be worse than the disease.