@lapxo/obligations 0.4.2 → 0.4.3

package/TARGET.bound

@@ -1,6 +1,6 @@
 bound-lock/1 about="# Contributing\n\nClone, `npm ci && npm run build`. Bring a counterexample with its inputs and printed output.\n\nOpen research demands are named in `TARGET.bound`; a result needs its model, conditions and falsifier. No count is inferred from missing observations.\n\nLines land through the instrument. Submit proposed lines for review rather than editing an admitted lock or its rendered pages.\n" at=witness:041/retain-declared-documentation by=target form=alphabet measure=text role=writes scope=prose/en/guide shape=CONTRIBUTING.md value=lock
 bound-lock/1 about="# In phase\n\nA claim carries a size and a direction, how much it narrows and the turn it landed on, and a withdrawal is the same size the other way: two hundred claims in phase make the square of their number and the same two hundred strewn make its order, while a write and its withdrawal at one epoch leave nothing and at different epochs do not cancel.\n\n## Run it\n\n```ts\nimport { coherence, interfere } from '@lapxo/obligations/views/field';\n\nconst span = { lo: 0, hi: 0 };\nlet held = 11;\nconst next = (): number => { held = (held * 1103515245 + 12345) % 2147483648; return held / 2147483648; };\nconst together = Array.from({ length: 32 }, (_, i) => ({ origin: `o${i}`, epoch: 8, span }));\nconst strewn = Array.from({ length: 32 }, (_, i) => ({ origin: `o${i}`, epoch: Math.floor(next() * 8), span }));\n\nconsole.log('in phase  ', interfere(together, 8), 'coherence', coherence(together, 8).toFixed(2));\nconsole.log('strewn    ', interfere(strewn, 8).toFixed(0), 'coherence', coherence(strewn, 8).toFixed(2));\nconsole.log('their number is 32, and its square is', 32 ** 2);\nconsole.log('why       claims that land at one turn of the period add as the square of their number, and the same claims strewn across it add as its order');\n```\n\n```\nin phase   1024 coherence 1.00\nstrewn     50 coherence 0.22\ntheir number is 32, and its square is 1024\nwhy       claims that land at one turn of the period add as the square of their number, and the same claims strewn across it add as its order\n```\n\nIts origins meet.\n\n## What it rests on\n\n```mermaid\nflowchart LR\n  c0[\"coherence\"]\n  c1[\"interfere\"]\n  l0([\"amplitude has sign and phase\"])\n  c0 --> l0\n  c1 --> l0\n  classDef shown fill:#2da44e,color:#ffffff,stroke:#2da44e\n  class l0 shown\n```\n\n## Source\n\n[Its source](../../examples/release/in-phase.ts) is one of the examples of obligations, and it runs as it is written, against the built package, with nothing compiled for it.\n" at=witness:041/retain-declared-documentation by=target form=alphabet measure=text role=writes scope=prose/en/learn-in-phase shape=docs/learn/in-phase.md value=lock
-bound-lock/1 about="# Measures beside the states\n\nThe pure API @lapxo/obligations/measures does not change FREE, REQUIRED, FORBIDDEN or CONFLICT. Observed association never establishes independence.\n\n## Descriptive contract\nerrorMatrix takes explicit binary event/origin/error rows and minCells >= 2. Identical redelivery is idempotent; conflicting observations refuse; incomplete shared events are unmeasured. covariance is population covariance of the handed rows, divided by their event count. participationRatio describes effective rank; zero variation is unmeasured.\n\nspectrum requires absolute, relative and maxRotations. Asymmetric or materially non-PSD matrices refuse. commonMass is the largest observed variance fraction, not a probability of false agreement. loadings require an isolated top eigenvalue; repeated eigenspaces have no unique top loading. commonMassOf requires an explicit cutoff and whole eigenspaces. The unanimity no-sample prints observed commonMass >= 0.5 beside FREE, without claiming strength.\n\ndescentCount counts distinct proper descendants in a declared restsOn graph. Identical edges add nothing. Missing declarations remain unmeasured; duplicate node declarations and cycles refuse. Declared descent is not observed activity or independence.\n\n## Controlled procedures\nunanimityFloor and unanimityThreshold implement the declared T1 special case. With P_k=(1-c)p^k(n-1)^(1-k), F_k=(c+P_k)/(c+P_k+(1-c)(1-p)^k). For nondegenerate parameters, F_(k+1) >= F_k iff p*c >= (1-p-p/(n-1))*P_k. The c=.01,p=.1,n=2 no-sample decreases before k=3; its threshold is 3. dissentExperiment and interestExperiment report conditional counts under explicitly handed generators: xorshift32, nonzero seed and trial count. Their denominators are events and falseEvents, and zero events are unmeasured. Interest is an explicitly handed bias toward one wrong value, not inferred from a missing observation. E3: a school answers truth with probability q, otherwise its own wrong value. These are controlled procedures, never laws or state changes.\n\n## Native publication evidence\nobservations, spectrum, unanimity and descent each have yes/no cases in samples/measures/<group>.json and installed-instrument captures in vectors/measures/<group>.cli.json. M/3 has samples/measures/m3.json and vectors/measures/m3.cli.json. Each vector ran on two independent places, repeated the positive reading byte-for-byte and refused its negative by name. State-preservation tests compare all four actual algebraic states before and after measurements.\n\n## Model boundary\nM/3 below is the only reference model. The older descriptive D/commonMass and independence claims are superseded. modelRequired returns unmeasured for an absent model; no measurement silently supplies a causal model. Factor attribution, dissent and probe results are conditional on M/3's handed parameters, not empirical independence certificates.\n\n# obligations 0.4.0 — model M/3 (authoritative contract)\n\nM/3 is the reference conditional model shipped beside the descriptive measures. Every M/3\nresult carries `model=M/3`, its handed parameters and its conditions. The descriptive\nmeasures are independent of M/3. Nothing changes a state. This document supersedes all\nearlier M/1, M/2 and amendment texts.\n\n## 1. Definition\nWitnesses W (unique identifiers); cells with truth t and integer cardinality n ≥ 2;\nfactors F, each with unique identifier, sharing set a_f ⊆ W, firing probability c_f and a\nunique declared index (total order); private error probability p_w for every witness\ninvolved. Per cell, independently:\n1. each factor fires independently with probability c_f;\n2. a witness is governed by the fired factor containing it with the lowest index, if\n   any; otherwise its private error fires with probability p_w;\n3. labels: each fired governing factor is a factor label; each privately erring witness\n   is a private label (namespaces are distinct: a factor label and a private label can\n   never merge even with identical identifier strings); every label draws one wrong\n   value uniformly from the n−1 wrong values; draws of distinct labels are mutually\n   independent and independent of all firing and private indicators; everyone else says t.\nSingleton factors (|a_f| = 1) are retained in declared order; the generator is never\nnormalized.\n\n## 2. Validation (refuse by name)\nS ⊆ W and S nonempty; every a_f ⊆ W; unique witness and factor identifiers; unique factor\nindices; c_f, p_w ∈ [0,1]; n integer ≥ 2; an explicitly handed p_w for every witness in\nS and in every sharing set touching S; a sharing set with duplicate members is refused.\n\n## 3. Exact evaluator\nEnumerate fired subsets of the factors touching S (2^|touching|) and private patterns of\nungoverned members of S (2^|free|). In each configuration, erring witnesses are grouped by\nlabel; L = number of labels; weights = witnesses per label. Collisions of wrong values\nfollow set partitions π of the labels with\nP(π) = (n−1)^{↓|π|} / (n−1)^L (falling factorial), and P(π) = 0 when |π| > n−1.\nEach partition yields block totals; unanimity ⇔ no correct witness and one block;\nmajority = strictly more than |S|/2 witnesses on one value; ties and pluralities ≤ half\nare \"no majority\".\nExactness: all c_f, p_w are handed as rationals (or decimal strings converted exactly);\naccumulation is in rational arithmetic; displayed decimals are rounded half-even to a\ndeclared number of digits. Budget: before any work, compute\nwork = Σ_configurations Bell(L) ≤ `evaluator/budget` (declared); if exceeded, return\n`unmeasured` with cause `evaluator/budget` and no partial accumulation.\nOutputs per (S, x): P_true, P_false (false unanimity), P_fired (false unanimity with some\ntouching factor fired), P_dissent = 1 − P_true − P_false (identity, asserted),\nmaj_true∧dissent, maj_false∧dissent, maj_none∧dissent (exclusive, sum = P_dissent).\nClosed form admitted only for P_true = Π_{f touches S}(1−c_f) · Π_{w∈S}(1−p_w).\n\n## 4. Quantities and denominators\n```\nΦ   = P_false / (P_true + P_false)        false-unanimity share\nD   = P_fired / (P_true + P_false)        factor-attribution share\nreliability = 1 − Φ, labelled \"P(correct | unanimous) under M/3 with handed parameters\"\nP(false maj | dissent)            = maj_false∧dissent / P_dissent\nP(false maj | any maj ∧ dissent)  = maj_false∧dissent / (maj_true∧dissent + maj_false∧dissent)\n```\nA ratio with zero denominator is `unmeasured` with cause; a zero numerator over a positive\ndenominator is a measured 0. Φ − D is private coincidence; 1 − D is never reliability.\n\n## 5. Dissent\nTheorem: if a factor covering S fires and no fired factor of lower index touches S\npartially, S is unanimous; disagreement inside S certifies this configuration did not\noccur — nothing more. False majorities under dissent remain possible through partial\nfactors and are reported through the two conditionals above. T1 and T5 stay controlled\nexperiments with declared PRNG, seed, trials and denominators (T5 uses the second\nconditional).\n\n## 6. Probe update\nFor a witness u private-only and independent of every factor touching S (or, with its\ntotal marginal error probability, labelled so):\n```\na = Φ · p_u/(n−1);  b = (1−Φ)·(1−p_u);  Φ′ = a/(a+b);  unmeasured iff a+b = 0\n```\nStrict improvement Φ′ < Φ holds for 0 < Φ < 1 iff p_u < (n−1)/n; boundary cases are\nvalid (Φ=0, p_u=0.2 → 0; Φ=1, p_u=0.2 → 1; p_u=1 → 1).\n\n## 7. Identification and fitting\nLet, for each witness w,\n```\nr_w = (1−p_w) · Π_{singleton f ∋ w} (1−c_f)        unresolved singleton/private survival\ns_w = r_w · Π_{shared f ∋ w} (1−c_f)                total correctness probability\ns_uv = P(e_u = 0 ∧ e_v = 0)\n```\nPopulation identification (declared sharing sets, s_u > 0 for every involved witness):\nlog(s_uv/(s_u·s_v)) = Σ_{shared f ∋ u,v} −log(1−c_f) is linear in −log(1−c_f); if the\npair × shared-factor system has full column rank, every shared c_f is identified, and then\nr_w from s_w; r_w's decomposition into p_w and singleton c_f is unresolved by error data.\nStructure is never discovered by a measure. Finite-data fitting is a separate declared\nprocedure: constrained least squares on the log system, declared budget, residual\ndiagnostics against a declared tolerance, bootstrap uncertainty with declared B/seed/PRNG,\npooling only under declared parameter compatibility; no pseudocounts, no clamping, no\npromotion of fitted structure; zero joint-correct counts or incompatible moments →\n`model/M misfit`, no c_f reported. Φ and D are reported only with handed parameters and\nlabelled `handed`, never `fitted`.\n\n## 8. Time, stake, commonMass\nFreshness and interest enter only as hypothetical factors with declared firing\nprobabilities; otherwise annotations and `unmeasured`. Descriptive commonMass is read as\nexcess over 1/|S| only as deviation from the equal-variance independent reference; it\nnever identifies dependence.\n\n## 9. Captured vectors (exact; rational reference; decimals rounded half-even, 8 digits)\n- two witnesses, covering c=1/20, p=1/5, n=64: P_true 0.608, P_false 0.05060317,\n  P_fired 0.05, P_dissent 0.34139683, Φ 0.07683409, D 0.07591825.\n- impossible conditioning (no factors, p_A=0, p_B=1): P_true = P_false = 0, P_dissent 1,\n  Φ, D unmeasured (cause: no unanimity mass).\n- three witnesses, covering {A,B,C} c=1/5 idx 0, partial {A,B} c=3/10 idx 1, p=0:\n  P_true 0.56, P_false 0.2, P_dissent 0.24, maj_false∧dissent 0.24, Φ = D = 0.26315789,\n  P(false maj | dissent) = 1, P(false maj | any maj ∧ dissent) = 1.\n- earlier singleton {A} c=3/10 idx 0, covering {A,B} c=1/5 idx 1, p=1/5: Φ 0.28387293,\n  D 0.28316249, P_dissent 0.49953016; merged-singleton no-sample Φ 0.35944805.\n- four witnesses, factors {0,1} and {2,3} each c=1/2, p=0: P_dissent 47/63 = 0.74603175,\n  maj_none∧dissent 47/63, P(false maj | dissent) = 0 (measured), P(false maj | any maj ∧\n  dissent) unmeasured (no majority mass), Φ = D = 1/64.\n- probe (Φ 0.07683409, n=64): p_u=0.2 → 0.00033; 0.98 → 0.06080; 0.99 → 0.11566; 1 → 1;\n  Φ=0,p_u=0.2 → 0; Φ=1,p_u=0.2 → 1.\n- identification: c=1/5, p_A=1/2, p_B=3/8 → c recovered exactly 1/5; (c=1/5,p_A=1,\n  p_B=3/8) vs (c=2/5,p_A=1,p_B=1/6) have identical moments → s_A = 0 → unidentified.\n- identity 1 − P_true − P_false = P_dissent holds in every case.\n\n## 10. Open, by name\nPartial value sharing (a factor firing without forcing one value) — a future model;\nprobe usefulness beyond the private-only case; discovery of sharing structure (never by a\nmeasure); adequacy policies (min cells, interval widths, budgets) as declared lines.\n\n## 11. Acceptance\nTypeScript implementation reproduces every captured vector exactly in rational\narithmetic; validation refuses by name; budget is computed before work; every result\ncarries model, handed parameters and conditions; states identical before and after;\nthe descriptive 0.4.0 measures and all algebra hashes unchanged.\n\n## 12. Capacity under a single dominant factor\n\n`capacityBound`, `capacityP0`, `mutualInformation`, and `unanimityNoFactor` use the declared single-dominant-factor model. T is uniform on n >= 2 values. F is independent of T with probability c. On F=1 all witnesses report one W, uniform among values different from T. On F=0 they use independent homogeneous private channels: correct with probability 1-p, otherwise uniform among n-1 wrong values. These hypotheses are handed, never inferred.\n\nFor every positive k, I(T;V) <= log2(n)-c*log2(n-1). The entropy chain rule gives\nH(T|V) = c*log2(n-1) + (1-c)*H(T|V,F=0) + I(T;F|V).\nThe first term is exclusion uncertainty, the second private uncertainty, the third factor ambiguity. A decoder cannot increase information about T.\n\nFor 0<p<1 and p!=(n-1)/n, the bound is attained as k tends to infinity: the empirical private vote distribution distinguishes T, and private unanimity has probability (1-p)^k+p^k/(n-1)^(k-1), tending to zero. For p=1,n>2 the missing value identifies T asymptotically. For p=1,n=2 every vote is the complement of T, giving one bit for every k. At p=0, I=log2(n)-h(c)-c*log2(n-1) for every k. At c=1, I=log2(n/(n-1)) for every k. At c=0 the uninformative private channel p=(n-1)/n gives zero for every k; c=0 does not by itself imply log2(n). At that uninformative p with a factor, the limit is c*log2(n/(n-1)), not the general upper bound.\n\nProbabilities and the conditional-entropy expression are exact rationals. Numerical entropy uses the explicitly handed `binary64-log2` policy, in bits, and is an approximation; it is not an exact rational logarithm. `mutualInformation` declares its joint-cell enumeration budget before work. `unanimityNoFactor` declares a rational-bit upper-bound budget before exponentiation. Budget exhaustion is unmeasured, never zero.\n\nCaptured finite information, n=4,c=1/5,p=3/10: k=1,2,4,7 gives approximately 0.313028978, 0.469712906, 0.914458010, 1.332321771 bits. The bound is approximately 1.683007500 bits. At p=0 the information is approximately 0.961079405 bits for every k. These decimals are illustrative approximations under the named policy.\n\n## 13. Decision results are separate\n\n`decisionInformation` accepts `plurality-uniform-ties@1` or `posterior-map-uniform-ties@1`. The second compares exact declared joint probabilities for finite k; it does not apply an intuition-based unanimity threshold. Ties are uniformly distributed among maximizers. It returns posterior information I(T;V), decision information I(T;answer), and exact rational decision accuracy as distinct quantities. The exact T/answer joint table accompanies them. A randomized answer cannot gain information over V.\n\nUnder p<(n-1)/n, asymptotically plurality emits the forced wrong value on the factor event. Posterior MAP identifies that event and samples uniformly among its n-1 alternatives. For n=4,c=1/5, plurality decision information is approximately 0.961079405 bits; MAP decision information is approximately 1.222162 bits; posterior information is approximately 1.683007500 bits. The two information gains differ. None is a finite-k or model-free guarantee.\n\n## 14. No-truth fitting\n\n`fitUnanimityModel` uses `exact-c-slices-on-declared-p-grid@1`. It takes one named lineage, cardinality, k/count/denominator rows, a declared p grid, exact absolute rate tolerance, and a work budget. For each handed p it intersects the exact inequalities for c, returning the entire compatible c interval and midpoint residual diagnostics. No claim is made about p values between slices or continuous identification. A measured fit means the declared procedure was evaluated; it does not mean parameters or causes were identified.\n\nThe observed rate is compared with c+(1-c)*[(1-p)^k+p^k/(n-1)^(k-1)]. Binary p versus 1-p ambiguity remains visible. Perfect unanimity leaves p=0 and c=1 indistinguishable. Too few informative sizes or no compatible slice are explicitly unmeasured. Mixed populations require a separately declared mixture model and are not pooled here.\n\nOptional bootstrap samples complete paired groups across sizes with a handed xorshift32 seed, B and random-word budget. It retains the full compatible regions in each replicate, not independent uncertainty intervals for each size. Groups are the declared resampling units; the procedure assumes exchangeable independent groups and does not establish that assumption. Counts alone leave uncertainty unmeasured. Replicate regions are empirical bootstrap output, not guaranteed confidence sets. Exhausted random words never produce successful uncertainty.\n\n## 15. False coincidence relative to a declared reference\n\n`coincidenceExcess` takes event/origin/value rows, truth rows (`kind=truth`), an explicit selection S and the `independent-uniform-wrong@1` reference. Order h is |S|; identity and redelivery never add origins or events. Values belong to the declared alphabet. Conflicting values or truth refuse. Incomplete shared events and insufficient declared event counts remain unmeasured. Imperfect probes need a separate contract and are not silently accepted as truth.\n\nThe reference uses handed error marginals, or explicitly estimated marginals stratified by observed truth. For handed homogeneous errors p_i, reference false coincidence is product(p_i)/(n-1)^(h-1). Estimated strata use their event weights, never a pooled error probability that confounds truth populations. Output contains the false-coincidence count, denominator, exact observed and reference rates, signed delta=observed-reference, and ratio=observed/reference. Delta is never floored at zero. At reference zero the ratio is unmeasured, while counts and delta remain available. No sampling design is inferred; uncertainty is separately unmeasured.\n\nA ratio greater than one means more coincidence than this reference, not a discovered common cause. Nonuniform wrong labels can exceed the uniform reference with independent witnesses. Independent errors with one shared wrong label give ratio (n-1)^(h-1): at n=64 this is 63 for pairs and 3969 for triples. The fair-bit XOR vector explicitly assigns one common wrong label: pairs have ratio 63, triples ratio zero and a negative delta. Indicators alone do not define the reported-value vector.\n\nUnder the separately handed dominant-factor model c=1/10,p=1/5,n=64, observed error marginals are 7/25. Relative to those marginals, delta at orders 2 and 3 is approximately 0.099326984 and 0.099996283. Its convergence to c is a conclusion under this model, never model-free identification.\n\n## 16. Research demands and boundaries\n\nThe spectrum summarizes covariance of binary error indicators. It is not a sufficient statistic of reported values, closure events or independence. False-unanimity excess summarizes one event and does not replace the full joint distribution. Higher-order closure statistics need a declared closure rule and the appropriate joint law of reported values given truth; neither is inferred from a form name.\n\nHelly is a property of a declared geometric family, not a probabilistic independence order or a truth guarantee. Geometry, factor precedence, partial label sharing, reproduction, percolation, time-varying fitting, imperfect-probe identification and model-free capacity remain explicitly open until their models, procedures and falsifiers close. Controlled experiments live in a world. No new origin, permission, algebraic amount or state is created by a measure.\n\n## 17. Publication and stability\n\nThe historical nine publication groups are linked to their contracts, yes/no data and declared Bound 0.1.7 captures in two places. The 0.4.2 operations below require new captures under their selected reader; historical implementation-bound receipts do not cover changed bytes. Validation readers are maintained TypeScript worlds with ordinary modular compilation; no .mjs bundle or private executor ships in this library. Their implementation inventory is retained once by digest. The installed archive must expose the same public exports, preserve the four states, match all published bytes and close native receipts without a private store. Declaring these requirements does not pay them.\n\n## 18. Declared failure bounds\n\n`unionBoundConfidence(failures)` requires distinct nonempty IDs, explicit rational probabilities in [0,1], and nonempty justification references. The caller must establish every failure bound in one probability space. References are retained, not verified. The returned lower bound for all events succeeding is max(0,1−Σp_i), with exact rational arithmetic. The empty family gives 1; Σp_i≥1 gives the trivial bound 0. No independence, truth probability, or algebraic state is inferred.\n\n`tolerantCountConfidence(failures,k)` requires a safe integer 0≤k≤m. For k<m it bounds failure by min(1,min_{0≤j≤k}(Σ of the m−j smallest p_i)/(k+1−j)); its exact success lower bound is the complement. For k=m success is certain. For m=3,k=1 and all p_i=1/10 the result is 17/20, not an inferred success rate. The distinct subset-intersection guarantee gives the weaker bound 7/10. Three disjoint paired failures among six events, each pair having probability 1/10, refute the former 4/5 claim: success is 7/10.\n\nThese operations compute conditional probability bounds. They select neither a geometric tolerance nor a consumer policy. Count-tolerant intersection T_k and widened-interval intersection T_eps remain named forms; tolerant-choice remains open.\n\n## 19. Declared covering cause\n\n`commonCauseCapacityBound(n,c,q,{logarithm:'binary64-log2'})` requires integer n≥2 and explicit rational c,q in [0,1]. Truth is uniform; the covering-cause indicator is independent of truth; when active, every report copies one symmetric channel with accuracy q. The inactive private channel is unrestricted. Revealing the indicator gives the upper bound log2(n)−c[h(q)+(1−q)log2(n−1)] in approximate bits. Copy count is no input. This is neither finite-sample mutual information nor an achievability claim.\n\nAt n=4,c=1,q=1/4 the bound is exactly zero; q=0 gives log2(4/3), and q=1 gives two bits. Missing q refuses. The operational declaration or detection of activation remains open in identification.model. Unconditioned coincidence does not identify q.\n\nM/3 and M/4 retain their existing definitions. Partial-factor governance proposals do not supply a conditional joint observation and selection law. The information carried by dissent, including a bottom symbol, is open for those unspecified proposals.\n" at=witness:042/native-quartet-closeout by=target form=alphabet measure=text role=writes scope=prose/en/measures shape=docs/measures.md value=lock
+bound-lock/1 about="# Measures beside the states\n\nThe pure API @lapxo/obligations/measures does not change FREE, REQUIRED, FORBIDDEN or CONFLICT. Observed association never establishes independence.\n\n## Descriptive contract\nerrorMatrix takes explicit binary event/origin/error rows and minCells >= 2. Identical redelivery is idempotent; conflicting observations refuse; incomplete shared events are unmeasured. covariance is population covariance of the handed rows, divided by their event count. participationRatio describes effective rank; zero variation is unmeasured.\n\nspectrum requires absolute, relative and maxRotations. Asymmetric or materially non-PSD matrices refuse. commonMass is the largest observed variance fraction, not a probability of false agreement. loadings require an isolated top eigenvalue; repeated eigenspaces have no unique top loading. commonMassOf requires an explicit cutoff and whole eigenspaces. The unanimity no-sample prints observed commonMass >= 0.5 beside FREE, without claiming strength.\n\ndescentCount counts distinct proper descendants in a declared restsOn graph. Identical edges add nothing. Missing declarations remain unmeasured; duplicate node declarations and cycles refuse. Declared descent is not observed activity or independence.\n\n## Controlled procedures\nunanimityFloor and unanimityThreshold implement the declared T1 special case. With P_k=(1-c)p^k(n-1)^(1-k), F_k=(c+P_k)/(c+P_k+(1-c)(1-p)^k). For nondegenerate parameters, F_(k+1) >= F_k iff p*c >= (1-p-p/(n-1))*P_k. The c=.01,p=.1,n=2 no-sample decreases before k=3; its threshold is 3. dissentExperiment and interestExperiment report conditional counts under explicitly handed generators: xorshift32, nonzero seed and trial count. Their denominators are events and falseEvents, and zero events are unmeasured. Interest is an explicitly handed bias toward one wrong value, not inferred from a missing observation. E3: a school answers truth with probability q, otherwise its own wrong value. These are controlled procedures, never laws or state changes.\n\n## Native publication evidence\nobservations, spectrum, unanimity and descent each have yes/no cases in samples/measures/<group>.json and installed-instrument captures in vectors/measures/<group>.cli.json. M/3 has samples/measures/m3.json and vectors/measures/m3.cli.json. Each vector ran on two independent places, repeated the positive reading byte-for-byte and refused its negative by name. State-preservation tests compare all four actual algebraic states before and after measurements.\n\n## Model boundary\nM/3 below is the only reference model. The older descriptive D/commonMass and independence claims are superseded. modelRequired returns unmeasured for an absent model; no measurement silently supplies a causal model. Factor attribution, dissent and probe results are conditional on M/3's handed parameters, not empirical independence certificates.\n\n# obligations 0.4.0 — model M/3 (authoritative contract)\n\nM/3 is the reference conditional model shipped beside the descriptive measures. Every M/3\nresult carries `model=M/3`, its handed parameters and its conditions. The descriptive\nmeasures are independent of M/3. Nothing changes a state. This document supersedes all\nearlier M/1, M/2 and amendment texts.\n\n## 1. Definition\nWitnesses W (unique identifiers); cells with truth t and integer cardinality n ≥ 2;\nfactors F, each with unique identifier, sharing set a_f ⊆ W, firing probability c_f and a\nunique declared index (total order); private error probability p_w for every witness\ninvolved. Per cell, independently:\n1. each factor fires independently with probability c_f;\n2. a witness is governed by the fired factor containing it with the lowest index, if\n   any; otherwise its private error fires with probability p_w;\n3. labels: each fired governing factor is a factor label; each privately erring witness\n   is a private label (namespaces are distinct: a factor label and a private label can\n   never merge even with identical identifier strings); every label draws one wrong\n   value uniformly from the n−1 wrong values; draws of distinct labels are mutually\n   independent and independent of all firing and private indicators; everyone else says t.\nSingleton factors (|a_f| = 1) are retained in declared order; the generator is never\nnormalized.\n\n## 2. Validation (refuse by name)\nS ⊆ W and S nonempty; every a_f ⊆ W; unique witness and factor identifiers; unique factor\nindices; c_f, p_w ∈ [0,1]; n integer ≥ 2; an explicitly handed p_w for every witness in\nS and in every sharing set touching S; a sharing set with duplicate members is refused.\n\n## 3. Exact evaluator\nEnumerate fired subsets of the factors touching S (2^|touching|) and private patterns of\nungoverned members of S (2^|free|). In each configuration, erring witnesses are grouped by\nlabel; L = number of labels; weights = witnesses per label. Collisions of wrong values\nfollow set partitions π of the labels with\nP(π) = (n−1)^{↓|π|} / (n−1)^L (falling factorial), and P(π) = 0 when |π| > n−1.\nEach partition yields block totals; unanimity ⇔ no correct witness and one block;\nmajority = strictly more than |S|/2 witnesses on one value; ties and pluralities ≤ half\nare \"no majority\".\nExactness: all c_f, p_w are handed as rationals (or decimal strings converted exactly);\naccumulation is in rational arithmetic; displayed decimals are rounded half-even to a\ndeclared number of digits. Budget: before any work, compute\nwork = Σ_configurations Bell(L) ≤ `evaluator/budget` (declared); if exceeded, return\n`unmeasured` with cause `evaluator/budget` and no partial accumulation.\nOutputs per (S, x): P_true, P_false (false unanimity), P_fired (false unanimity with some\ntouching factor fired), P_dissent = 1 − P_true − P_false (identity, asserted),\nmaj_true∧dissent, maj_false∧dissent, maj_none∧dissent (exclusive, sum = P_dissent).\nClosed form admitted only for P_true = Π_{f touches S}(1−c_f) · Π_{w∈S}(1−p_w).\n\n## 4. Quantities and denominators\n```\nΦ   = P_false / (P_true + P_false)        false-unanimity share\nD   = P_fired / (P_true + P_false)        factor-attribution share\nreliability = 1 − Φ, labelled \"P(correct | unanimous) under M/3 with handed parameters\"\nP(false maj | dissent)            = maj_false∧dissent / P_dissent\nP(false maj | any maj ∧ dissent)  = maj_false∧dissent / (maj_true∧dissent + maj_false∧dissent)\n```\nA ratio with zero denominator is `unmeasured` with cause; a zero numerator over a positive\ndenominator is a measured 0. Φ − D is private coincidence; 1 − D is never reliability.\n\n## 5. Dissent\nTheorem: if a factor covering S fires and no fired factor of lower index touches S\npartially, S is unanimous; disagreement inside S certifies this configuration did not\noccur — nothing more. False majorities under dissent remain possible through partial\nfactors and are reported through the two conditionals above. T1 and T5 stay controlled\nexperiments with declared PRNG, seed, trials and denominators (T5 uses the second\nconditional).\n\n## 6. Probe update\nFor a witness u private-only and independent of every factor touching S (or, with its\ntotal marginal error probability, labelled so):\n```\na = Φ · p_u/(n−1);  b = (1−Φ)·(1−p_u);  Φ′ = a/(a+b);  unmeasured iff a+b = 0\n```\nStrict improvement Φ′ < Φ holds for 0 < Φ < 1 iff p_u < (n−1)/n; boundary cases are\nvalid (Φ=0, p_u=0.2 → 0; Φ=1, p_u=0.2 → 1; p_u=1 → 1).\n\n## 7. Identification and fitting\nLet, for each witness w,\n```\nr_w = (1−p_w) · Π_{singleton f ∋ w} (1−c_f)        unresolved singleton/private survival\ns_w = r_w · Π_{shared f ∋ w} (1−c_f)                total correctness probability\ns_uv = P(e_u = 0 ∧ e_v = 0)\n```\nPopulation identification (declared sharing sets, s_u > 0 for every involved witness):\nlog(s_uv/(s_u·s_v)) = Σ_{shared f ∋ u,v} −log(1−c_f) is linear in −log(1−c_f); if the\npair × shared-factor system has full column rank, every shared c_f is identified, and then\nr_w from s_w; r_w's decomposition into p_w and singleton c_f is unresolved by error data.\nStructure is never discovered by a measure. Finite-data fitting is a separate declared\nprocedure: constrained least squares on the log system, declared budget, residual\ndiagnostics against a declared tolerance, bootstrap uncertainty with declared B/seed/PRNG,\npooling only under declared parameter compatibility; no pseudocounts, no clamping, no\npromotion of fitted structure; zero joint-correct counts or incompatible moments →\n`model/M misfit`, no c_f reported. Φ and D are reported only with handed parameters and\nlabelled `handed`, never `fitted`.\n\n## 8. Time, stake, commonMass\nFreshness and interest enter only as hypothetical factors with declared firing\nprobabilities; otherwise annotations and `unmeasured`. Descriptive commonMass is read as\nexcess over 1/|S| only as deviation from the equal-variance independent reference; it\nnever identifies dependence.\n\n## 9. Captured vectors (exact; rational reference; decimals rounded half-even, 8 digits)\n- two witnesses, covering c=1/20, p=1/5, n=64: P_true 0.608, P_false 0.05060317,\n  P_fired 0.05, P_dissent 0.34139683, Φ 0.07683409, D 0.07591825.\n- impossible conditioning (no factors, p_A=0, p_B=1): P_true = P_false = 0, P_dissent 1,\n  Φ, D unmeasured (cause: no unanimity mass).\n- three witnesses, covering {A,B,C} c=1/5 idx 0, partial {A,B} c=3/10 idx 1, p=0:\n  P_true 0.56, P_false 0.2, P_dissent 0.24, maj_false∧dissent 0.24, Φ = D = 0.26315789,\n  P(false maj | dissent) = 1, P(false maj | any maj ∧ dissent) = 1.\n- earlier singleton {A} c=3/10 idx 0, covering {A,B} c=1/5 idx 1, p=1/5: Φ 0.28387293,\n  D 0.28316249, P_dissent 0.49953016; merged-singleton no-sample Φ 0.35944805.\n- four witnesses, factors {0,1} and {2,3} each c=1/2, p=0: P_dissent 47/63 = 0.74603175,\n  maj_none∧dissent 47/63, P(false maj | dissent) = 0 (measured), P(false maj | any maj ∧\n  dissent) unmeasured (no majority mass), Φ = D = 1/64.\n- probe (Φ 0.07683409, n=64): p_u=0.2 → 0.00033; 0.98 → 0.06080; 0.99 → 0.11566; 1 → 1;\n  Φ=0,p_u=0.2 → 0; Φ=1,p_u=0.2 → 1.\n- identification: c=1/5, p_A=1/2, p_B=3/8 → c recovered exactly 1/5; (c=1/5,p_A=1,\n  p_B=3/8) vs (c=2/5,p_A=1,p_B=1/6) have identical moments → s_A = 0 → unidentified.\n- identity 1 − P_true − P_false = P_dissent holds in every case.\n\n## 10. Open, by name\nPartial value sharing (a factor firing without forcing one value) — a future model;\nprobe usefulness beyond the private-only case; discovery of sharing structure (never by a\nmeasure); adequacy policies (min cells, interval widths, budgets) as declared lines.\n\n## 11. Acceptance\nTypeScript implementation reproduces every captured vector exactly in rational\narithmetic; validation refuses by name; budget is computed before work; every result\ncarries model, handed parameters and conditions; states identical before and after;\nthe descriptive 0.4.0 measures and all algebra hashes unchanged.\n\n## 12. Capacity under a single dominant factor\n\n`capacityBound`, `capacityP0`, `mutualInformation`, and `unanimityNoFactor` use the declared single-dominant-factor model. T is uniform on n >= 2 values. F is independent of T with probability c. On F=1 all witnesses report one W, uniform among values different from T. On F=0 they use independent homogeneous private channels: correct with probability 1-p, otherwise uniform among n-1 wrong values. These hypotheses are handed, never inferred.\n\nFor every positive k, I(T;V) <= log2(n)-c*log2(n-1). The entropy chain rule gives\nH(T|V) = c*log2(n-1) + (1-c)*H(T|V,F=0) + I(T;F|V).\nThe first term is exclusion uncertainty, the second private uncertainty, the third factor ambiguity. A decoder cannot increase information about T.\n\nFor 0<p<1 and p!=(n-1)/n, the bound is attained as k tends to infinity: the empirical private vote distribution distinguishes T, and private unanimity has probability (1-p)^k+p^k/(n-1)^(k-1), tending to zero. For p=1,n>2 the missing value identifies T asymptotically. For p=1,n=2 every vote is the complement of T, giving one bit for every k. At p=0, I=log2(n)-h(c)-c*log2(n-1) for every k. At c=1, I=log2(n/(n-1)) for every k. At c=0 the uninformative private channel p=(n-1)/n gives zero for every k; c=0 does not by itself imply log2(n). At that uninformative p with a factor, the limit is c*log2(n/(n-1)), not the general upper bound.\n\nProbabilities and the conditional-entropy expression are exact rationals. Numerical entropy uses the explicitly handed `binary64-log2` policy, in bits, and is an approximation; it is not an exact rational logarithm. `mutualInformation` declares its joint-cell enumeration budget before work. `unanimityNoFactor` declares a rational-bit upper-bound budget before exponentiation. Budget exhaustion is unmeasured, never zero.\n\nCaptured finite information, n=4,c=1/5,p=3/10: k=1,2,4,7 gives approximately 0.313028978, 0.469712906, 0.914458010, 1.332321771 bits. The bound is approximately 1.683007500 bits. At p=0 the information is approximately 0.961079405 bits for every k. These decimals are illustrative approximations under the named policy.\n\n## 13. Decision results are separate\n\n`decisionInformation` accepts `plurality-uniform-ties@1` or `posterior-map-uniform-ties@1`. The second compares exact declared joint probabilities for finite k; it does not apply an intuition-based unanimity threshold. Ties are uniformly distributed among maximizers. It returns posterior information I(T;V), decision information I(T;answer), and exact rational decision accuracy as distinct quantities. The exact T/answer joint table accompanies them. A randomized answer cannot gain information over V.\n\nUnder p<(n-1)/n, asymptotically plurality emits the forced wrong value on the factor event. Posterior MAP identifies that event and samples uniformly among its n-1 alternatives. For n=4,c=1/5, plurality decision information is approximately 0.961079405 bits; MAP decision information is approximately 1.222162 bits; posterior information is approximately 1.683007500 bits. The two information gains differ. None is a finite-k or model-free guarantee.\n\n## 14. No-truth fitting\n\n`fitUnanimityModel` uses `exact-c-slices-on-declared-p-grid@1`. It takes one named lineage, cardinality, k/count/denominator rows, a declared p grid, exact absolute rate tolerance, and a work budget. For each handed p it intersects the exact inequalities for c, returning the entire compatible c interval and midpoint residual diagnostics. No claim is made about p values between slices or continuous identification. A measured fit means the declared procedure was evaluated; it does not mean parameters or causes were identified.\n\nThe observed rate is compared with c+(1-c)*[(1-p)^k+p^k/(n-1)^(k-1)]. Binary p versus 1-p ambiguity remains visible. Perfect unanimity leaves p=0 and c=1 indistinguishable. Too few informative sizes or no compatible slice are explicitly unmeasured. Mixed populations require a separately declared mixture model and are not pooled here.\n\nOptional bootstrap samples complete paired groups across sizes with a handed xorshift32 seed, B and random-word budget. It retains the full compatible regions in each replicate, not independent uncertainty intervals for each size. Groups are the declared resampling units; the procedure assumes exchangeable independent groups and does not establish that assumption. Counts alone leave uncertainty unmeasured. Replicate regions are empirical bootstrap output, not guaranteed confidence sets. Exhausted random words never produce successful uncertainty.\n\n## 15. False coincidence relative to a declared reference\n\n`coincidenceExcess` takes event/origin/value rows, truth rows (`kind=truth`), an explicit selection S and the `independent-uniform-wrong@1` reference. Order h is |S|; identity and redelivery never add origins or events. Values belong to the declared alphabet. Conflicting values or truth refuse. Incomplete shared events and insufficient declared event counts remain unmeasured. Imperfect probes need a separate contract and are not silently accepted as truth.\n\nThe reference uses handed error marginals, or explicitly estimated marginals stratified by observed truth. For handed homogeneous errors p_i, reference false coincidence is product(p_i)/(n-1)^(h-1). Estimated strata use their event weights, never a pooled error probability that confounds truth populations. Output contains the false-coincidence count, denominator, exact observed and reference rates, signed delta=observed-reference, and ratio=observed/reference. Delta is never floored at zero. At reference zero the ratio is unmeasured, while counts and delta remain available. No sampling design is inferred; uncertainty is separately unmeasured.\n\nA ratio greater than one means more coincidence than this reference, not a discovered common cause. Nonuniform wrong labels can exceed the uniform reference with independent witnesses. Independent errors with one shared wrong label give ratio (n-1)^(h-1): at n=64 this is 63 for pairs and 3969 for triples. The fair-bit XOR vector explicitly assigns one common wrong label: pairs have ratio 63, triples ratio zero and a negative delta. Indicators alone do not define the reported-value vector.\n\nUnder the separately handed dominant-factor model c=1/10,p=1/5,n=64, observed error marginals are 7/25. Relative to those marginals, delta at orders 2 and 3 is approximately 0.099326984 and 0.099996283. Its convergence to c is a conclusion under this model, never model-free identification.\n\n## 16. Research demands and boundaries\n\nThe spectrum summarizes covariance of binary error indicators. It is not a sufficient statistic of reported values, closure events or independence. False-unanimity excess summarizes one event and does not replace the full joint distribution. Higher-order closure statistics need a declared closure rule and the appropriate joint law of reported values given truth; neither is inferred from a form name.\n\nHelly is a property of a declared geometric family, not a probabilistic independence order or a truth guarantee. Geometry, factor precedence, partial label sharing, reproduction, percolation, time-varying fitting, imperfect-probe identification and model-free capacity remain explicitly open until their models, procedures and falsifiers close. Controlled experiments live in a world. No new origin, permission, algebraic amount or state is created by a measure.\n\n## 17. Publication and stability\n\nThe historical nine publication groups are linked to their contracts, yes/no data and declared Bound 0.1.7 captures in two places. The 0.4.2 operations below require new captures under their selected reader; historical implementation-bound receipts do not cover changed bytes. Validation readers are maintained TypeScript worlds with ordinary modular compilation; no .mjs bundle or private executor ships in this library. Their implementation inventory is retained once by digest. The installed archive must expose the same public exports, preserve the four states, match all published bytes and close native receipts without a private store. Declaring these requirements does not pay them.\n\n## 18. Declared failure bounds\n\n`unionBoundConfidence(failures)` requires distinct nonempty IDs, explicit rational probabilities in [0,1], and nonempty justification references. The caller must establish every failure bound in one probability space. References are retained, not verified. The returned lower bound for all events succeeding is max(0,1−Σp_i), with exact rational arithmetic. The empty family gives 1; Σp_i≥1 gives the trivial bound 0. No independence, truth probability, or algebraic state is inferred.\n\n`tolerantCountConfidence(failures,k)` requires a safe integer 0≤k≤m. For k<m it bounds failure by min(1,min_{0≤j≤k}(Σ of the m−j smallest p_i)/(k+1−j)); its exact success lower bound is the complement. For k=m success is certain. For m=3,k=1 and all p_i=1/10 the result is 17/20, not an inferred success rate. The distinct subset-intersection guarantee gives the weaker bound 7/10. Three disjoint paired failures among six events, each pair having probability 1/10, refute the former 4/5 claim: success is 7/10.\n\nThese operations compute conditional probability bounds. They select neither a geometric tolerance nor a consumer policy. Count-tolerant intersection T_k and widened-interval intersection T_eps remain named forms; tolerant-choice remains open.\n\n## 19. Declared covering cause\n\n`commonCauseCapacityBound(n,c,q,{logarithm:'binary64-log2'})` requires integer n≥2 and explicit rational c,q in [0,1]. Truth is uniform; the covering-cause indicator is independent of truth; when active, every report copies one symmetric channel with accuracy q. The inactive private channel is unrestricted. Revealing the indicator gives the upper bound log2(n)−c[h(q)+(1−q)log2(n−1)] in approximate bits. Copy count is no input. This is neither finite-sample mutual information nor an achievability claim.\n\nAt n=4,c=1,q=1/4 the bound is exactly zero; q=0 gives log2(4/3), and q=1 gives two bits. Missing q refuses. The operational declaration or detection of activation remains open in identification.model. Unconditioned coincidence does not identify q.\n\nM/3 and M/4 retain their existing definitions. Partial-factor governance proposals do not supply a conditional joint observation and selection law. The information carried by dissent, including a bottom symbol, is open for those unspecified proposals.\n\n## Obligations 0.4.3 — proposed conditional measures\n\nL7: conditional bound on one declared common probability space: P(N>k)<=min(1,sum over |S|=k+1 of min p_i), with rational declared marginal failure upper bounds and 0<=k<=m. Trivial iff computed upper bound is one. Counterexamples refute the former top-k+1 success guarantee. Joint-independent Bernoulli convolution is a separate declared hypothesis; rho=0 and pairwise independence do not supply it. Existing stronger 0.4.2 count bound is unchanged and needs no joint independence.\n\nL8: law by enumeration only for 3<=n<=64, positive integer divisors p|n and integer 0<=s<p. D=log2(n/p)-h(s/p), explicit binary64-log2 approximation. Odd-period minimum uses h(floor(p/2)/p), even-period minimum uses one. Complete maximum at zero; nonzero domain has endpoints 1,p-1, empty when p=1. No information or worst-case interpretation follows from nonnegativity. T87 retains its separate wrapped-phase amplitude identity.\n\nL9: law by counterexample of existence of noninjective weighted mix, and measure with a declared prior on a finite explicitly supplied set of named-factor configurations. Equal-weight ab and ba give the same bottom tie and different named forks. Loss is H(configuration | mix)=H(configuration)-H(mix); a uniform prior must be explicitly selected, absent prior is unmeasured. The four-state uniform example has mean loss 0.5 bits and one bit conditioned on tie; the injective named fork has zero loss. Defines no new algebraic act, governance selection, M/3 or M/4 law, or destroyed field.\n" at=witness:043/three-regions by=target form=alphabet measure=text role=writes scope=prose/en/measures shape=docs/measures.md value=lock
 bound-lock/1 about="# Price\n\n<p align=\"center\"><img src=\"../img/price.svg\" alt=\"Two origins on one scale: seller claims 2 to 9 and buyer claims 5 to 12; what they hold together runs from a floor of 5 to a ceiling of 9.\" width=\"640\"></p>\n\nA claim from one origin is potential; a cell is information only where two independent origins touch it and their intervals meet.\n\n## Run it\n\n```ts\nimport { chain, disagree, freedom, meetAt, origins, point, unit } from '@lapxo/obligations';\n\nconst said = (cell: Parameters<typeof meetAt>[0]): string =>\n  (origins(cell) < 2 ? 'grey' : disagree(cell) ? 'split' : 'agreed');\n\nconst price = point('price', 0, [\n  { origin: 'seller', span: { lo: 2, hi: 9 } },\n  { origin: 'buyer', span: { lo: 5, hi: 12 } },\n]);\nconst scale = chain(Array.from({ length: 16 }, (_, i) => `l${i}`));\n\nconsole.log('verdict   ', said(price));\nconsole.log('they hold ', JSON.stringify(meetAt(price)));\nconsole.log('still free', freedom(scale, unit('price', { floor: 'l5', ceiling: 'l9' })).toFixed(2), 'bits');\nconsole.log('why       a seller and a buyer are two origins; where their marks meet is a price, and what is left between the marks is what nobody has decided yet');\n```\n\n```\nverdict    agreed\nthey hold  {\"lo\":5,\"hi\":9}\nstill free 2.32 bits\nwhy       a seller and a buyer are two origins; where their marks meet is a price, and what is left between the marks is what nobody has decided yet\n```\n\nIts origins meet.\n\n## What it rests on\n\n```mermaid\nflowchart LR\n  c0[\"point\"]\n  c1[\"unit\"]\n  l0([\"encounter\"])\n  l1([\"debit in bits\"])\n  c0 --> l0\n  c1 --> l1\n  classDef shown fill:#2da44e,color:#ffffff,stroke:#2da44e\n  class l0 shown\n```\n\n## Source\n\n[Its source](../../examples/release/price.ts) is one of the examples of obligations, and it runs as it is written, against the built package, with nothing compiled for it.\n" at=witness:041/retain-declared-documentation by=target form=alphabet measure=text role=writes scope=prose/en/learn-price shape=docs/learn/price.md value=lock
 bound-lock/1 about="# Reference\n\nEvery operation this package offers, every law it states and every form it declares: 242 operations, 126 laws, 4 forms.\n\n## Notation\n\n```text\nc = ⟨⊥, ⊤, O, ⊑⟩        over L at r\n```\n\n| formula | law | sentence |\n|---|---|---|\n| `meet(c) = ⊤ ∧ ⋀_{o∈O} span_o` | meet | two bounds on one cell fold by the meet of their form; the order they arrive in never decides |\n| `info(c) ⟺ \\|ext(O)\\| ≥ 2 ∧ meet(c) ≠ ∅` | encounter | a claim from one origin is potential; a cell is information only where two independent origins touch it and their intervals meet |\n| `state(c) ∈ {FREE, REQUIRED, FORBIDDEN, CONFLICT}` | four states | one origin is required; two or more that meet within the ceiling are free; that meet outside it are forbidden; that do not meet are in conflict; fewer states collapse two of them |\n| `x ⊨ c ⟺ ⊥ ⊑ x ⊑ ⊤` | carrier | every quantity has two poles, a floor that is required and a ceiling that is permitted, and a cell is the pair |\n| `F(c) = log₂ \\|{x ∈ L : ⊥ ≤ x ≤ ⊤}\\|` | debit in bits | the debit of a cell is log2 of the states between its floor and ceiling; a closed cell has zero |\n| `claim = (o, span, e), e ≺ e′ ⟺ e → e′` | epoch | every claim carries when it holds; order is causal, never a clock |\n| `ext(o) = {(x, span_o(x))}, ext(o) = ext(o′) ⟹ o = o′` | origin is extension | an origin is the set of (coordinate, interval) it has produced; two origins with one extension are one; independence is distance between extensions |\n| `o ⊢_{r,e} c for o, c ∈ C, and ⊢ ∈ C` | everything is a bound | one relation: an origin bounds a cell at a resolution and an epoch; cells, origins, locks and laws are cells under it; the relation bounds itself, and there is no level above it |\n| `sign: ⊤(c′) ← ⊤(c′) ∧ s ∀ c′ ⊒ c` | signatures are meets | a signature narrows a ceiling and travels to every cell that rests on it as a meet, and meets do not remember the order they were taken in, so no order of any number of signatures ever shows in the field they leave; widening is not a meet and does not travel, which is why a narrowing costs one signature and an error costs one at every cell, and folding the field and signing it commute even with a withdrawal present |\n| `F(C) = Σ_{c∈C} F(c), F(a(c)) ≤ F(c) for every act a that narrows` | freedom conserved | the freedom of a system is the sum of its cells' debits in bits; an encounter never raises it; only a new coordinate does |\n| `⊥(c) = ⋁ required ∨ ⋁_{o∈O} lo_o` | the obligatory | the object is c = ⟨⊥, ⊤, O, ⊑⟩ over L at r: four parts in two pairs, the value (⊥, ⊤) and the relations (O, ⊑), with the lattice L it is given and the resolution r it is read at belonging to the reading and the sign to the act; an origin is (o, φ), and the poles are read, never stored: ⊥ is what is required joined with what its live claims hold, ⊤ what is signed met with what they allow; eight acts in four pairs and seven names answer for it: meet and join on the ceiling, local, a join only with a witness; observe and withdraw on the floor, local, a withdrawal being an observation at −1 that restores exactly; sign and require on both poles, travelling by rest, commuting and both closing freedom; encounter and refine with no pole; and no widening travels |\n| `⊤(c) = ⋀ signed ∧ ⋀_{o∈O} hi_o` | the obligatory | the object is c = ⟨⊥, ⊤, O, ⊑⟩ over L at r: four parts in two pairs, the value (⊥, ⊤) and the relations (O, ⊑), with the lattice L it is given and the resolution r it is read at belonging to the reading and the sign to the act; an origin is (o, φ), and the poles are read, never stored: ⊥ is what is required joined with what its live claims hold, ⊤ what is signed met with what they allow; eight acts in four pairs and seven names answer for it: meet and join on the ceiling, local, a join only with a witness; observe and withdraw on the floor, local, a withdrawal being an observation at −1 that restores exactly; sign and require on both poles, travelling by rest, commuting and both closing freedom; encounter and refine with no pole; and no widening travels |\n| `A(c) = \\|Σ_{o∈O} s_o w_o e^{iφ_o}\\|²` | amplitude has sign and phase | a claim carries a size and a direction, how much it narrows and the turn it landed on, and a withdrawal is the same size the other way: two hundred claims in phase make the square of their number and the same two hundred strewn make its order, while a write and its withdrawal at one epoch leave nothing and at different epochs do not cancel |\n| `⊤(c) = ⋀_{b ⊑ c} signed(b) ∧ ⋀_{o∈O} hi_o, one line per act` | signatures live where they are signed | a signature or requirement is a line at the bound, never a value copied into resting cells; a cell's poles are read up its chain of rest as the meet of the live signatures above it; an absorbed signature stays live so withdrawing a tighter one reveals it. Signing and withdrawing cost one line; reading costs the depth of rest; nothing propagates, and A2 with T3 makes the reading equal to any replay. |\n| `\\|e^{ia} + e^{ib}\\|² < 2 ⟺ \\|a − b\\| > π/2` | the quarter turn | two claims reinforce within a quarter turn of each other and cancel beyond it, since \\|e^{ia} + e^{ib}\\|² < 2 exactly when \\|a − b\\| > π/2: a fork is destructive interference, and a field is laminar while fewer than half its pairs lie beyond a quarter turn, forks over closes below one |\n| `state(c) ∈ {FREE, REQUIRED, FORBIDDEN, CONFLICT}` | state | a cell is in one of four states: FREE, REQUIRED, FORBIDDEN, CONFLICT — green, grey, red, fork are their renders |\n| `frag(c) = min { \\|W\\| : W ⊆ O(c), ⊥(c∖W) ≠ ⊥(c) or ⊤(c∖W) ≠ ⊤(c) }` | fragility is the fewest withdrawals | a cell is as fragile as the fewest origins whose withdrawal moves one of its poles: a pole one origin sets breaks with that origin, and a pole several origins set together holds until all of them withdraw |\n\n```text\ncell  : at ×  × [Claim] × [restsOn] → Obligatory\nparts : Lattice × Obligatory × World → Cell\nstate : Lattice × Obligatory × World → state\nmark  : Mark → Mark\nlive  : [L] → [L]\nstate(c) ∈ {FREE, REQUIRED, FORBIDDEN, CONFLICT}\n```\n\n## States\n\n```mermaid\nstateDiagram-v2\n  state \"FREE · green\" as FREE\n  state \"REQUIRED · grey\" as REQUIRED\n  state \"FORBIDDEN · red\" as FORBIDDEN\n  state \"CONFLICT · fork\" as CONFLICT\n  FORBIDDEN --> CONFLICT: dangerous\n  FORBIDDEN --> FREE: a witnessed join reopened it\n  FREE --> CONFLICT: dangerous\n  FREE --> FORBIDDEN: a signature closed below what they hold\n  FREE --> REQUIRED: a withdrawal took one back\n  REQUIRED --> CONFLICT: dangerous\n  REQUIRED --> FREE: a second origin met the first\n  classDef green fill:#2da44e,color:#ffffff,stroke:#2da44e\n  classDef grey fill:#8c959f,color:#ffffff,stroke:#8c959f\n  classDef red fill:#cf222e,color:#ffffff,stroke:#cf222e\n  classDef fork fill:#8250df,color:#ffffff,stroke:#8250df\n  class FREE green\n  class REQUIRED grey\n  class FORBIDDEN red\n  class CONFLICT fork\n```\n\n## What the object rests on\n\n```mermaid\nflowchart BT\n  o([\"the obligatory\"])\n  r0([\"carrier\"])\n  r1([\"encounter\"])\n  r2([\"meet\"])\n  r3([\"epoch\"])\n  r4([\"origin is extension\"])\n  r5([\"freedom conserved\"])\n  r6([\"everything is a bound\"])\n  o --> r0\n  o --> r1\n  o --> r2\n  o --> r3\n  o --> r4\n  o --> r5\n  o --> r6\n  classDef object fill:#2da44e,color:#ffffff,stroke:#2da44e\n  class o object\n```\n\n## Operations\n\n| operation | view | signature | refuses | law |\n|---|---|---|---|---|\n| anyOf | chain | `anyOf(c: Scale, units: readonly Unit[]): Unit[]` | 1 | no law |\n| coarsen | chain | `coarsen(place: Point, steps: number): Point` | 1 | no law |\n| composeState | chain | `composeState(c: Scale, units: readonly Unit[]): State` | 1 | no law |\n| conflictDebits | chain | `conflictDebits(c: Scale, u: Unit): number` | 1 | no law |\n| downset | chain | `downset(c: Scale, names: readonly string[]): readonly string[]` | 2 | D2-ideal |\n| entails | chain | `entails(c: Scale, a: Unit, b: Unit): boolean` | 1 | no law |\n| foldRegion | chain | `foldRegion<T>(places: readonly Point[], where: Region, f: (met: readonly Point[]) => T): readonly Fold<T>[]` | 1 | no law |\n| freedom | chain | `freedom(c: Scale, u: Unit): number` | 3 | no law |\n| missing | chain | `missing(c: Scale, have: Unit, want: Unit): { raiseFloorTo: string[]; raiseCeilingTo: string[] }` | 2 | no law |\n| size | chain | `size(at: readonly string[], steps?: number): number` | 1 | A3-resolution |\n| sufficientResolution | chain | `sufficientResolution<T>(finest: number, verdict: (steps: number) => T, same: (a: T, b: T) => boolean = Object.is): number` | 1 | T7-sufficient-resolution |\n| unit | chain | `unit(subject: string, bounds: { floor: string; ceiling: string }): Unit` | 1 | D4-debit-in-bits |\n| unitToIdeal | chain | `unitToIdeal(c: Scale, u: Unit): IdealDebit` | 2 | no law |\n| upset | chain | `upset(c: Scale, names: readonly string[]): readonly string[]` | 2 | D2-ideal |\n| weakest | chain | `weakest(methods: Iterable<Method>): Method` | 2 | D4-debit-in-bits |\n| and | complement | `and(a: Packed, b: Packed): Packed` | 13 | no law |\n| asBoolean | complement | `asBoolean(c: Scale, u: Unit): BoolView` | 1 | no law |\n| bitLoss | complement | `bitLoss(c: Scale): { readonly possible: number; readonly exactly: number; readonly lossy: number; readonly unsayable: number; }` | 1 | no law |\n| canonical | complement | `canonical(pk: Packer, p: Packed): Packed` | 2 | T8-exact-withdrawal |\n| opinion | complement | `opinion(demanded: number, permitted: number, limiting: number, required: number): Packed` | 2 | T8-exact-withdrawal |\n| amplitude | field | `amplitude(claim: Claim): number` | 2 | T49-amplitude-has-sign-and-phase |\n| bites | field | `bites<T>(L: Lattice<T>, world: World<T>, resting: ReadonlyMap<string, readonly string[]>, name: string, span: T): { readonly cells: readonly string[]; readonly visits: number }` | 2 | T94-a-signature-stops-where-the-meet-is-identity |\n| budget | field | `budget(cells: number, resolution: number): number` | 2 | H5-hilbert-is-the-obligatory-without-four-fields |\n| cancels | field | `cancels(a: Origin, b: Origin): boolean` | 5 | T87-the-quarter-turn, T104-concurrent-withdrawal-cancels |\n| cell | field | `cell<T>(at: string, epoch = 0, seen: readonly Claim<T>[] = [], restsOn: readonly string[] = []): Obligatory<T>` | 7 | D11-the-obligatory |\n| choose | field | `choose(lattice: Lattice<Band>, world: readonly Obligatory<Band>[], weights: Weights, held: ReadonlySet<string>): string` | 2 | T55-a-policy-is-four-weights |\n| coherence | field | `coherence(claims: readonly Claim[], period: number): number` | 3 | T49-amplitude-has-sign-and-phase |\n| contracts | field | `contracts(regions: readonly Region[], touched: readonly Point[]): boolean` | 1 | no law |\n| converge | field | `converge(width: number, maturation: number, forgetting: number, epochs: number): { readonly width: number; readonly trend: 'widens' \\| 'narrows' \\| 'holds' \\| 'zero' }` | 1 | T6-present-converges |\n| coordinate | field | `coordinate(name: string, steps = '/'): readonly string[]` | 12 | D0-coordinate |\n| covers | field | `covers(met: readonly Point[], of: { readonly where: readonly Region[]; readonly steps: Interval; readonly when: Interval }): boolean` | 1 | no law |\n| decohere | field | `decohere(claims: readonly Claim<Band>[], environment: number, period: number): number` | 2 | H5-hilbert-is-the-obligatory-without-four-fields |\n| disagree | field | `disagree(place: Point): readonly [Sighting, Sighting] \\| null` | 1 | no law |\n| distinct | field | `distinct(lattice: Lattice<Band>, world: readonly Obligatory<Band>[], width: number): number` | 2 | H5-hilbert-is-the-obligatory-without-four-fields |\n| disturbance | field | `disturbance(lattice: Lattice<Band>, one: Obligatory<Band>, claim: Claim<Band>): number` | 2 | T47-a-two-pole-object |\n| encounter | field | `encounter<T>(L: Lattice<T>, c: Obligatory<T>, world: World<T> = new Map()): { readonly origins: number; readonly held: T }` | 4 | D11-the-obligatory |\n| encounter | field | `owed` | 0 | no law |\n| encounter | field | `owed` | 0 | no law |\n| encounter | field | `owed` | 0 | no law |\n| encounter | field | `owed` | 0 | no law |\n| encounter | field | `owed` | 0 | no law |\n| encounter | field | `owed` | 0 | no law |\n| endsBeforeNow | field | `endsBeforeNow(now: number, leading: readonly (readonly [number, number])[], lagging: readonly (readonly [number, number])[]): boolean` | 1 | D8-present |\n| evolve | field | `evolve(world: readonly Obligatory<Band>[], epochs: number, budget_: number, seed: number): readonly (readonly Obligatory<Band>[])[]` | 2 | T52-directing-where-origins-land-closes-the-horizon |\n| exclusive | field | `exclusive(mine: readonly Region[], others: readonly (readonly Region[])[]): readonly Region[]` | 1 | no law |\n| foldIn | field | `foldIn<T>(L: MeetSemilattice<T>, demands: readonly T[]): T` | 2 | A4-epoch |\n| fragility | field | `fragility<T>(L: Lattice<T>, c: Obligatory<T>, world: World<T> = new Map()): number` | 2 | T101-fragility-is-the-fewest-withdrawals |\n| fringes | field | `fringes(claims: readonly Claim[], period: number, delays: readonly number[]): readonly number[]` | 2 | T48-epoch-is-phase |\n| interfere | field | `interfere(claims: readonly Claim[], period: number): number` | 2 | T49-amplitude-has-sign-and-phase |\n| join | field | `join<T>(L: Lattice<T>, c: Obligatory<T>, span: T, witness: string, at?: number): Obligatory<T>` | 5 | D11-the-obligatory, A7-freedom-conserved |\n| join | field | `owed` | 0 | no law |\n| join | field | `owed` | 0 | no law |\n| join | field | `owed` | 0 | no law |\n| join | field | `owed` | 0 | no law |\n| join | field | `owed` | 0 | no law |\n| join | field | `owed` | 0 | no law |\n| key | field | `key(regions: readonly Region[]): string` | 1 | no law |\n| laminar | field | `laminar(origins: readonly Origin[]): boolean` | 2 | T89-disagreement-is-where-to-look |\n| latest | field | `latest(places: readonly Point[]): Point \\| null` | 1 | no law |\n| live | field | `live<L extends { readonly id?: string; readonly takes?: string }>(lines: readonly L[]): readonly L[]` | 2 | D11-the-obligatory |\n| looksLeft | field | `looksLeft(L: Lattice<Interval>, world: World<Interval>, origin: string): number` | 2 | T93-every-act-is-paid-by-its-origin |\n| mark | field | `mark<T>(one: Mark<T>): Mark<T>` | 3 | D11-the-obligatory, A7-freedom-conserved |\n| meet | field | `meet<T>(L: Lattice<T>, c: Obligatory<T>, world: World<T> = new Map()): T` | 7 | D11-the-obligatory |\n| meet | field | `owed` | 0 | no law |\n| meet | field | `owed` | 0 | no law |\n| meet | field | `owed` | 0 | no law |\n| meet | field | `owed` | 0 | no law |\n| meet | field | `owed` | 0 | no law |\n| meet | field | `owed` | 0 | no law |\n| meetAt | field | `meetAt(place: Point): Interval` | 1 | no law |\n| observe | field | `observe<T>(c: Obligatory<T>, claim: Claim<T>, s: Act = claim.takes === undefined ? 1 : -1): Obligatory<T>` | 3 | D11-the-obligatory |\n| observe | field | `owed` | 0 | no law |\n| observe | field | `owed` | 0 | no law |\n| observe | field | `owed` | 0 | no law |\n| observe | field | `owed` | 0 | no law |\n| observe | field | `owed` | 0 | no law |\n| observe | field | `owed` | 0 | no law |\n| origins | field | `origins(place: Point): number` | 18 | no law |\n| pairs | field | `pairs(origins: readonly Origin[]): { readonly forks: number; readonly closes: number }` | 3 | T87-the-quarter-turn |\n| parts | field | `parts<T>(L: Lattice<T>, c: Obligatory<T>, world: World<T> = new Map()): Cell<T>` | 3 | D11-the-obligatory |\n| passes | field | `passes(place: Point \\| Region, step: string): boolean` | 1 | no law |\n| pay | field | `pay(L: Lattice<Interval>, world: World<Interval>, origin: string): World<Interval>` | 3 | T93-every-act-is-paid-by-its-origin, T100-an-act-is-paid-with-attention |\n| phase | field | `phase(claim: Claim, period: number): number` | 3 | T48-epoch-is-phase |\n| play | field | `play(lattice: Lattice<Band>, world: readonly Obligatory<Band>[], weights: Weights, epochs: number, seed: number, spread: number): { readonly closed: number; readonly forks: number; readonly reopened: number; readonly coverage: number; }` | 2 | T55-a-policy-is-four-weights |\n| point | field | `point(name: string, epoch = 0, seen: readonly Sighting[] = [], steps = '/'): Point` | 2 | A1-encounter |\n| potential | field | `potential(place: Point): boolean` | 1 | no law |\n| present | field | `present(place: Point): Interval \\| 'potential' \\| null` | 2 | no law |\n| price | field | `price<T>(bits: (held: T) => number, before: T, after: T): number` | 1 | T100-an-act-is-paid-with-attention |\n| refine | field | `refine<T>(c: Obligatory<T>, steps: number): Obligatory<T>` | 2 | D11-the-obligatory |\n| refine | field | `owed` | 0 | no law |\n| refine | field | `owed` | 0 | no law |\n| refine | field | `owed` | 0 | no law |\n| refine | field | `owed` | 0 | no law |\n| refine | field | `owed` | 0 | no law |\n| refine | field | `owed` | 0 | no law |\n| region | field | `region(name: string, steps = '/'): Region` | 3 | D1-region |\n| require | field | `require<T>(L: Lattice<T>, world: World<T>, name: string, span: T, line: Line = {}): World<T>` | 3 | D11-the-obligatory |\n| require | field | `owed` | 0 | no law |\n| require | field | `owed` | 0 | no law |\n| require | field | `owed` | 0 | no law |\n| require | field | `owed` | 0 | no law |\n| require | field | `owed` | 0 | no law |\n| require | field | `owed` | 0 | no law |\n| resolution | field | `resolution(place: Point \\| Region): number` | 8 | no law |\n| restOf | field | `restOf<T>(c: Obligatory<T>, world: World<T> = new Map()): readonly string[]` | 3 | T99-the-fold-is-total |\n| restingOn | field | `restingOn<T>(world: World<T>): ReadonlyMap<string, readonly string[]>` | 2 | T94-a-signature-stops-where-the-meet-is-identity |\n| sameObject | field | `sameObject(lattice: Lattice<Band>, a: Obligatory<Band>, b: Obligatory<Band>, width: number): boolean` | 2 | H5-hilbert-is-the-obligatory-without-four-fields |\n| seenAs | field | `seenAs(cell: Point, period: number): readonly Claim[]` | 2 | T49-amplitude-has-sign-and-phase |\n| selfFold | field | `selfFold(lattice: Lattice<Band>, one: Obligatory<Band>, epochs: number, width: number, spread: number, seed: number): readonly Obligatory<Band>[]` | 2 | T73-a-cell-that-folds-itself-converges |\n| sign | field | `sign<T>(L: Lattice<T>, world: World<T>, name: string, span: T, line: Line = {}): World<T>` | 3 | D11-the-obligatory |\n| sign | field | `owed` | 0 | no law |\n| sign | field | `owed` | 0 | no law |\n| sign | field | `owed` | 0 | no law |\n| sign | field | `owed` | 0 | no law |\n| sign | field | `owed` | 0 | no law |\n| sign | field | `owed` | 0 | no law |\n| state | field | `state<T>(L: Lattice<T>, c: Obligatory<T>, world: World<T> = new Map()): State` | 5 | D11-the-obligatory, A8-four-states |\n| turned | field | `turned(before: readonly Point[], after: readonly Point[]): readonly Point[]` | 2 | no law |\n| within | field | `within(inner: Point \\| Region, outer: Region): boolean` | 3 | D1-region |\n| witnessesIn | field | `witnessesIn<T>(L: MeetSemilattice<T>, demands: readonly T[]): readonly [T, T] \\| null` | 1 | no law |\n| alphabets | galois | `alphabets(): Poset<Alphabet>` | 1 | no law |\n| relate | galois | `relate<T>(L: Poset<T>, a: T, b: T): Relation` | 2 | no law |\n| closureViolationsIn | ideal | `closureViolationsIn(c: Scale, xs: readonly IdealDebit[]): readonly string[]` | 1 | no law |\n| complexity | ideal | `complexity(c: Scale, u: Unit): number` | 1 | no law |\n| dangerous | ideal | `dangerous(m: Matrix): readonly { from: State; to: State; n: number }[]` | 2 | no law |\n| fromIdeals | ideal | `fromIdeals(c: Scale): Lattice<IdealDebit>` | 1 | no law |\n| holds | ideal | `holds(c: Scale, units: readonly Unit[], index: number): boolean` | 3 | no law |\n| idealComplexity | ideal | `idealComplexity(c: Scale, d: IdealDebit): number` | 1 | no law |\n| idealDebit | ideal | `idealDebit(c: Scale, demand: readonly string[], permit: readonly string[]): IdealDebit` | 1 | no law |\n| idealEntails | ideal | `idealEntails(a: IdealDebit, b: IdealDebit): boolean` | 1 | no law |\n| idealJoin | ideal | `idealJoin(a: IdealDebit, b: IdealDebit): IdealDebit` | 1 | no law |\n| idealMeet | ideal | `idealMeet(a: IdealDebit, b: IdealDebit): IdealDebit` | 1 | no law |\n| isDistributive | ideal | `isDistributive(c: Scale): boolean` | 1 | no law |\n| isDownset | ideal | `isDownset(c: Scale, names: readonly string[]): boolean` | 1 | no law |\n| isIdealPair | ideal | `isIdealPair(c: Scale, d: IdealDebit): boolean` | 1 | no law |\n| isUpset | ideal | `isUpset(c: Scale, names: readonly string[]): boolean` | 1 | no law |\n| moved | ideal | `moved(m: Matrix): number` | 2 | no law |\n| packer | ideal | `packer(c: Scale): Packer` | 1 | no law |\n| permitAll | ideal | `permitAll(c: Scale, subject: string): Unit` | 1 | no law |\n| quadrants | ideal | `quadrants(demanded: WideMask, permitted: WideMask, extent: WideMask): Quadrants` | 2 | D3-state |\n| rebuild | ideal | `rebuild(c: Scale, reg: Register, subject = ''): Unit` | 1 | no law |\n| register | ideal | `register(c: Scale, u: Unit): Register` | 1 | no law |\n| runQuadrants | ideal | `runQuadrants(demanded: readonly Span[], permitted: Span \\| null, extent: Span): readonly Run[]` | 2 | D3-state |\n| state | ideal | `state(c: Scale, u: Unit): State` | 5 | D3-state, D2-ideal |\n| stateAt | ideal | `stateAt(q: Quadrants, bit: number): State \\| null` | 2 | no law |\n| statesPerDebit | ideal | `statesPerDebit(c: Scale, u: Unit): [string, State][]` | 1 | no law |\n| table | ideal | `table(a: State, b: State): DebitOutcome` | 1 | no law |\n| touches | ideal | `touches(c: Scale, mine: Unit, folded: Unit): boolean` | 3 | no law |\n| transitions | ideal | `transitions(before: Quadrants, after: Quadrants, width: number): Matrix` | 2 | no law |\n| unitComplexity | ideal | `unitComplexity(c: Scale, u: Unit): number` | 1 | no law |\n| widen | ideal | `widen(c: Scale, units: readonly Unit[]): Unit` | 1 | no law |\n| alphabetForm | lattice | `alphabetForm: Form<WideMask>` | 4 | A2-meet |\n| foldSigned | lattice | `foldSigned<T>(L: Poset<T>, values: readonly T[]): { readonly classes: readonly (readonly number[])[]; readonly standing: readonly number[]; readonly fork: boolean; }` | 2 | D7-fork |\n| latticeForm | lattice | `latticeForm: Form<Unit>` | 3 | no law |\n| meetViolationsIn | lattice | `meetViolationsIn<T>(L: MeetSemilattice<T>, xs: readonly T[]): readonly string[]` | 1 | no law |\n| posetViolationsIn | lattice | `posetViolationsIn<T>(L: Poset<T>, xs: readonly T[]): readonly string[]` | 1 | no law |\n| replay | lattice | `replay<T>(landed: readonly Point[], where: Region, verdict: (met: readonly Point[]) => T): readonly Fold<T>[]` | 2 | no law |\n| subsets | lattice | `subsets(tokens: readonly string[]): Lattice<WideMask>` | 3 | no law |\n| violationsIn | lattice | `violationsIn<T>(L: Lattice<T>, xs: readonly T[]): readonly string[]` | 1 | no law |\n| agreements | product | `agreements(cells: readonly Point[], steps: number): { readonly agreed: number; readonly forks: number; readonly cells: number }` | 2 | T36-agreement-has-a-resolution |\n| apart | product | `apart(world: readonly Obligatory<Band>[], a: string, b: string, bounds: readonly string[] = []): boolean` | 3 | T51-the-shape-of-knowing-can-be-designed |\n| area | product | `area(cell: Point): number` | 2 | T38-origins-are-a-polygon |\n| buildGraph | product | `buildGraph(edges: readonly Edge[]): Graph` | 2 | T4-composition |\n| centre | product | `centre(cell: Point): Vertex` | 3 | T38-origins-are-a-polygon |\n| compose | product | `compose(c: Scale, units: readonly Unit[]): Unit` | 2 | no law |\n| continuousForm | product | `continuousForm: Form<{ lo: number; hi: number }>` | 4 | A2-meet |\n| curvature | product | `curvature(world: readonly Point[]): number` | 2 | T40-cost-by-density |\n| density | product | `density(cell: Point): number` | 3 | T38-origins-are-a-polygon |\n| design | product | `design(world: readonly Obligatory<Band>[], budget: number, bounds: readonly string[] = []): readonly (readonly [string, string])[]` | 2 | T51-the-shape-of-knowing-can-be-designed |\n| distance | product | `distance(world: readonly Point[], from: string, to: string, bounds: readonly string[] = []): number` | 3 | T41-distance-is-the-walk-between-lookings |\n| entangled | product | `entangled(world: readonly Point[], bound: string): readonly string[]` | 4 | T45-cells-that-rest-on-one-bound-are-entangled |\n| fromScale | product | `fromScale(c: Scale): Lattice<Unit>` | 1 | no law |\n| fuse | product | `fuse(c: Scale, units: readonly Unit[]): Unit` | 1 | no law |\n| grow | product | `grow(walls: readonly number[], placed: readonly number[]): readonly { readonly at: number; readonly width: number }[]` | 2 | T39-cells-grow-until-stopped |\n| horizon | product | `horizon(world: readonly Point[], from: string, bounds: readonly string[] = []): readonly string[]` | 2 | T41-distance-is-the-walk-between-lookings |\n| horizonTotal | product | `horizonTotal(world: readonly Obligatory<Band>[], bounds: readonly string[] = []): number` | 2 | T51-the-shape-of-knowing-can-be-designed |\n| hull | product | `hull(cell: Point): readonly Vertex[]` | 2 | T38-origins-are-a-polygon |\n| intervals | product | `intervals(lo: number, hi: number): Lattice<Interval>` | 1 | no law |\n| ladderForm | product | `ladderForm: Form<Unit>` | 4 | A2-meet |\n| lone | product | `lone(name: string, epoch = 0): Obligatory<Band>` | 2 | T44-an-origin-is-worth-its-reach |\n| loom | product | `loom(world: readonly Obligatory<Band>[], epochs: number, budget: number, bounds: readonly string[] = []): readonly number[]` | 2 | T52-directing-where-origins-land-closes-the-horizon |\n| neighbours | product | `neighbours(world: readonly Point[], bounds: readonly string[] = []): ReadonlyMap<string, readonly string[]>` | 2 | T41-distance-is-the-walk-between-lookings |\n| or | product | `or(a: Packed, b: Packed): Packed` | 5 | no law |\n| outliers | product | `outliers(cell: Point): readonly string[]` | 3 | T38-origins-are-a-polygon |\n| pour | product | `pour(world: readonly Point[], lock: Lock): { readonly filled: number; readonly of: number }` | 2 | T37-a-lock-is-a-mould |\n| rank | product | `rank(world: readonly Obligatory<Band>[], bound: string): number` | 2 | T45-cells-that-rest-on-one-bound-are-entangled |\n| reach | product | `reach(world: readonly Point[], from: string, bounds: readonly string[] = []): number` | 3 | T46-a-bound-moves-cells-no-walk-reaches |\n| reachIds | product | `reachIds(g: Graph, from: number, scratch?: Uint8Array): Uint8Array` | 2 | T4-composition |\n| red | product | `red(world: readonly Point[], lock: Lock): readonly string[]` | 2 | T37-a-lock-is-a-mould |\n| residual | product | `residual(cells: readonly Point[], steps: number): number` | 2 | T36-agreement-has-a-resolution |\n| resting | product | `resting(world: readonly Point[], bound: string, was: Interval, now: Interval): readonly string[]` | 4 | T45-cells-that-rest-on-one-bound-are-entangled |\n| sited | product | `sited(world: readonly Obligatory<Band>[], sites: readonly (readonly [string, string])[], epoch = 1): readonly Obligatory<Band>[]` | 3 | T51-the-shape-of-knowing-can-be-designed |\n| steady | product | `steady(cell: Point): boolean` | 3 | T38-origins-are-a-polygon |\n| stretch | product | `stretch(lock: Lock, k: number): Lock` | 2 | T37-a-lock-is-a-mould |\n| sufficient | product | `sufficient(cells: readonly Point[], upTo: number, tolerance = 1e-9): number` | 2 | T36-agreement-has-a-resolution |\n| tense | product | `tense(lock: Lock, world: readonly Point[], rate: number): Lock` | 2 | T37-a-lock-is-a-mould |\n| tighter | product | `tighter(before: Lock, after: Lock): { readonly tightened: number; readonly loosened: number }` | 2 | T37-a-lock-is-a-mould |\n| tunnel | product | `tunnel(world: readonly Point[], places: readonly string[], bounds: readonly string[] = []): readonly (readonly [string, string])[]` | 2 | T44-an-origin-is-worth-its-reach |\n| vertices | product | `vertices(cell: Point): readonly Vertex[]` | 2 | T38-origins-are-a-polygon |\n| width | product | `width(cell: Point): number` | 3 | T38-origins-are-a-polygon |\n| answers | quotient | `answers(met: readonly Point[], where: readonly Region[]): boolean` | 1 | no law |\n| outranks | quotient | `outranks(met: readonly Point[], named: readonly (readonly [string, string])[], order: readonly string[]): string \\| null` | 1 | no law |\n| permutationBlind | quotient | `permutationBlind<T, R>(values: readonly T[], claim: (xs: readonly T[]) => R, draws = 8): boolean` | 1 | A2-meet |\n| permute | quotient | `permute<T>(xs: readonly T[], salt: number): T[]` | 1 | A2-meet |\n| retracted | quotient | `retracted(met: readonly Point[], by: readonly Region[]): readonly Point[]` | 1 | no law |\n| verified | quotient | `verified(met: readonly Point[], green: (place: Point) => boolean): boolean` | 1 | no law |\n| attainable | spectrum | `attainable(c: Scale): { readonly values: readonly string[]; readonly holes: readonly { readonly value: string; readonly after: string; readonly before: string }[]; }` | 1 | no law |\n| c_n | spectrum | `c_n(n: number): number` | 1 | T13-c_n |\n| classifyValue | spectrum | `classifyValue(c: Scale, value: string): Attainability` | 1 | no law |\n| meetBlocks | spectrum | `meetBlocks(c: Scale, u: Level, w: Level): Level[][]` | 1 | no law |\n| spectrum | spectrum | `spectrum(c: Scale, worth: Valuation, places = 9, quantity: SpectrumQuantity = 'premium'): { values: number[]; pairs: number; band: { low: number; high: number }; degeneracy: number }` | 1 | no law |\n| asUnit | valuation | `asUnit(c: Scale, d: Declaration, subject = ''): Unit` | 2 | no law |\n| best | valuation | `best(c, value, u, w)` | 1 | thm-extremal |\n| chain | valuation | `chain(levels: readonly string[]): Scale` | 5 | no law |\n| collapses | valuation | `collapses(c: Scale, value: Valuation, u: Level, w: Level, tol = 1e-9): boolean` | 1 | no law |\n| conj | valuation | `conj(c: Scale, a: Declaration, b: Declaration): Declaration` | 2 | no law |\n| costOfState | valuation | `costOfState(c: Scale, value: Valuation, u: Level, w: Level, s: Level): { shortfall: number; overshoot: number }` | 1 | no law |\n| declaration | valuation | `declaration(c: Scale, u: Unit): Declaration` | 2 | D5-declaration |\n| declarationOf | valuation | `declarationOf(c: Scale, place: Point): { readonly demand: Level; readonly limit: Level }` | 1 | no law |\n| defect | valuation | `defect(c: Scale, value: Valuation, a: Level, b: Level): number` | 1 | no law |\n| degenerate | valuation | `degenerate(c: Scale, value: (level: Level) => number): boolean` | 1 | no law |\n| disj | valuation | `disj(c: Scale, a: Declaration, b: Declaration): Declaration` | 2 | no law |\n| dropTo | valuation | `dropTo(value: Valuation, u: Level, m: Level): number` | 1 | no law |\n| frontier | valuation | `frontier(c: Scale, value: (level: Level) => number, asked: readonly Asked[]): readonly Open[]` | 2 | D10-open-question |\n| gap | valuation | `gap(c: Scale, value: Valuation, u: Level, w: Level): number` | 1 | no law |\n| height | valuation | `height(c: Scale, level: Level): number` | 1 | no law |\n| isMonotone | valuation | `isMonotone(c: Scale, value: Valuation): boolean` | 1 | no law |\n| isSupermodular | valuation | `isSupermodular(c: Scale, value: Valuation): boolean` | 1 | no law |\n| movesOf | valuation | `movesOf(c: Scale, before: Point, after: Point): Matrix` | 1 | no law |\n| order | valuation | `order(levels: readonly string[], leq: readonly boolean[][]): Scale` | 6 | no law |\n| premium | valuation | `premium(c: Scale, value: Valuation, u: Level, w: Level): number` | 2 | D6-premium |\n| premiumOf | valuation | `premiumOf(c: Scale, value: (level: Level) => number, place: Point): number` | 1 | no law |\n| price | valuation | `price(c: Scale, value: Valuation, u: Level, w: Level): Price` | 1 | no law |\n| pricesAnything | valuation | `pricesAnything(c: Scale, worth: Valuation): boolean` | 1 | no law |\n| registerOf | valuation | `registerOf(c: Scale, place: Point): readonly [string, State][]` | 1 | no law |\n| slack | valuation | `slack(c: Scale, a: Level, b: Level): { up: number; down: number } \\| null` | 1 | no law |\n| stateOf | valuation | `stateOf(c: Scale, place: Point): State` | 1 | no law |\n| unitOf | valuation | `unitOf(c: Scale, place: Point): Unit` | 1 | no law |\n| unitStream | valuation | `unitStream(seed: number): () => number` | 1 | no law |\n| vacuityOf | valuation | `vacuityOf(predicate: string, structure: string, c: Scale, test: (c: Scale, value: Valuation) => boolean, opts: VacuityOptions = {}): Verdict` | 1 | no law |\n| withdrawsExactly | valuation | `withdrawsExactly(c: Scale): boolean` | 1 | no law |\n| yieldOf | valuation | `yieldOf(c: Scale, acts: readonly Act[]): readonly Yield[]` | 2 | D9-yield |\n\n## Laws\n\n| law | kind | view | rests on | sentence | formula | yes | falsifier | reproduced |\n|---|---|---|---|---|---|---|---|---|\n| A0-carrier | axiom | lattice | nothing | every quantity has two poles, a floor that is required and a ceiling that is permitted, and a cell is the pair | `x ⊨ c ⟺ ⊥ ⊑ x ⊑ ⊤` | [yes](../samples/yes/A0-carrier.json) | derived by the reader | not a derivation |\n| A1-encounter | axiom | field | nothing | a claim from one origin is potential; a cell is information only where two independent origins touch it and their intervals meet | `info(c) ⟺ \\|ext(O)\\| ≥ 2 ∧ meet(c) ≠ ∅` | [yes](../samples/yes/A1-encounter.json) | derived by the reader | not a derivation |\n| A2-meet | axiom | lattice | nothing | two bounds on one cell fold by the meet of their form; the order they arrive in never decides | `meet(c) = ⊤ ∧ ⋀_{o∈O} span_o` | [yes](../samples/yes/A2-meet.json) | derived by the reader | not a derivation |\n| A3-resolution | axiom | chain | nothing | every observation has a scale; hierarchy is resolution; a coordinate's depth is its resolution |  | [yes](../samples/yes/A3-resolution.json) | derived by the reader | not a derivation |\n| A4-epoch | axiom | field | nothing | every claim carries when it holds; order is causal, never a clock | `claim = (o, span, e), e ≺ e′ ⟺ e → e′` | [yes](../samples/yes/A4-epoch.json) | derived by the reader | not a derivation |\n| A5-origin-is-extension | axiom | galois | nothing | an origin is the set of (coordinate, interval) it has produced; two origins with one extension are one; independence is distance between extensions | `ext(o) = {(x, span_o(x))}, ext(o) = ext(o′) ⟹ o = o′` | [yes](../samples/yes/A5-origin-is-extension.json) | derived by the reader | not a derivation |\n| A6-vacuity | axiom | field | nothing | a region no origin touches for an epoch contracts; nothing keeps a cell alive but a new observation |  | [yes](../samples/yes/A6-vacuity.json) | derived by the reader | not a derivation |\n| A7-freedom-conserved | axiom | chain | nothing | the freedom of a system is the sum of its cells' debits in bits; an encounter never raises it; only a new coordinate does | `F(C) = Σ_{c∈C} F(c), F(a(c)) ≤ F(c) for every act a that narrows` | [yes](../samples/yes/A7-freedom-conserved.json) | derived by the reader | not a derivation |\n| A8-four-states | axiom | ideal | nothing | one origin is required; two or more that meet within the ceiling are free; that meet outside it are forbidden; that do not meet are in conflict; fewer states collapse two of them | `state(c) ∈ {FREE, REQUIRED, FORBIDDEN, CONFLICT}` | [yes](../samples/yes/A8-four-states.json) | derived by the reader | not a derivation |\n| A9-everything-is-a-bound | axiom | lattice | nothing | one relation: an origin bounds a cell at a resolution and an epoch; cells, origins, locks and laws are cells under it; the relation bounds itself, and there is no level above it | `o ⊢_{r,e} c for o, c ∈ C, and ⊢ ∈ C` | [yes](../samples/yes/A9-everything-is-a-bound.json) | derived by the reader | not a derivation |\n| C1-independence-is-distance | corollary | galois | A5-origin-is-extension | the independence of two origins is the distance between their extensions: one extension is one origin, disjoint extensions share nothing |  | [yes](../samples/yes/C1-independence-is-distance.json) | derived by the reader | yes |\n| C3-bootstrap-once | corollary | field | A1-encounter, A4-epoch | one claim per generation is admitted without a second origin; a second such claim is a fork |  | [yes](../samples/yes/C3-bootstrap-once.json) | derived by the reader | yes |\n| C4-dpi-subadditive | corollary | valuation | thm-collapse, thm-modular | Data processing forces submodularity — Every valuation with the data-processing property is submodular. Among such valuations, $g\\equiv0$ if and only if $f$ is modular. |  | none | derived by the reader | yes |\n| D0-coordinate | definition | field | A1-encounter | a coordinate exists where two origins agree they observe one thing; the coordinate set of a domain is the fold of recognitions, never a schema |  | [yes](../samples/yes/D0-coordinate.json) | derived by the reader | not a derivation |\n| D1-region | definition | field | A3-resolution | a region is (coordinate prefix, length); refinement is a longer prefix; regions are derived, never stored |  | none | derived by the reader | not a derivation |\n| D10-open-question | definition | valuation | A0-carrier, A1-encounter, D6-premium | a signed ceiling with no origin: a question with form, resolution, premium, dependents and a cheapest closer |  | [yes](../samples/yes/D10-open-question.json) | derived by the reader | not a derivation |\n| D11-the-obligatory | definition | lattice | A0-carrier, A1-encounter, A2-meet, A4-epoch, A5-origin-is-extension, A7-freedom-conserved, A9-everything-is-a-bound | the object is c = ⟨⊥, ⊤, O, ⊑⟩ over L at r: four parts in two pairs, the value (⊥, ⊤) and the relations (O, ⊑), with the lattice L it is given and the resolution r it is read at belonging to the reading and the sign to the act; an origin is (o, φ), and the poles are read, never stored: ⊥ is what is required joined with what its live claims hold, ⊤ what is signed met with what they allow; eight acts in four pairs and seven names answer for it: meet and join on the ceiling, local, a join only with a witness; observe and withdraw on the floor, local, a withdrawal being an observation at −1 that restores exactly; sign and require on both poles, travelling by rest, commuting and both closing freedom; encounter and refine with no pole; and no widening travels | `⊥(c) = ⋁ required ∨ ⋁_{o∈O} lo_o` | [yes](../samples/yes/obligatory.json) | The ways the object can be got wrong that no act's descriptor derives. A cell with no ceiling is not a cell; a meet that assumes the reals is a meet of one lattice pretending to be the meet; a state written as a word in the code is a state the lattice did not give. A signature that edits an origin's claim has forged a reading, and a withdrawal matched by value takes back a claim nobody made. And a floor raised by some composition of the other acts, taken in every order, would make require a second name for what was already there. What a descriptor forces is derived from the descriptors and run over every act, never written here: a widening that travels, narrows or lands with nobody standing for it, a travelling act that misses a cell resting on its bound, a narrowing that opens freedom, a pole stored, a withdrawal that does not restore. | not a derivation |\n| D2-ideal | definition | ideal | A0-carrier | a cell's bound is an ideal (upset, downset): what is required and what is forbidden, as a pair |  | none | derived by the reader | not a derivation |\n| D3-state | definition | ideal | D2-ideal | a cell is in one of four states: FREE, REQUIRED, FORBIDDEN, CONFLICT — green, grey, red, fork are their renders | `state(c) ∈ {FREE, REQUIRED, FORBIDDEN, CONFLICT}` | [yes](../samples/yes/D3-state.json) | derived by the reader | not a derivation |\n| D4-debit-in-bits | definition | chain | A0-carrier, A3-resolution | the debit of a cell is log2 of the states between its floor and ceiling; a closed cell has zero | `F(c) = log₂ \\|{x ∈ L : ⊥ ≤ x ≤ ⊤}\\|` | [yes](../samples/yes/D4-debit-in-bits.json) | derived by the reader | not a derivation |\n| D5-declaration | definition | valuation | A0-carrier | a declaration is a demand u and a limit w on a valued lattice; its cost is forced by two axioms |  | none | derived by the reader | not a derivation |\n| D6-premium | definition | valuation | D5-declaration, D4-debit-in-bits | the premium of a cell is what closing it costs: the residual information under the declaration |  | none | derived by the reader | not a derivation |\n| D7-fork | definition | lattice | A2-meet | two incomparable bounds on one cell from two signers; rendered with both, closed only by a signature |  | none | derived by the reader | not a derivation |\n| D8-present | definition | field | A4-epoch, A3-resolution | the present is the meet of overlapping observation bands; there is no instant to certify |  | [yes](../samples/yes/D8-present.json) | derived by the reader | not a derivation |\n| D9-yield | definition | valuation | A5-origin-is-extension, D4-debit-in-bits | an origin's yield is bits it closed minus bits reopened against it, per act, over an epoch |  | [yes](../samples/yes/D9-yield.json) | derived by the reader | not a derivation |\n| H1-uncertainty | conjecture | chain | A3-resolution, A1-encounter | resolution times independent encounters per cell stays above a floor c; the sufficient resolution is where the product saturates |  | [yes](../samples/yes/H1-uncertainty.json) | The falsifier: a cell read at four resolutions whose product of resolution and independent encounters comes out above the floor while the encounters fall away with the sharpening. A product that rises as the reading sharpens would say the floor is not a floor but an accident of how coarsely the cell was read. | not a derivation |\n| H10-the-period-is-a-line-of-the-region | conjecture | field | T48-epoch-is-phase, T87-the-quarter-turn | the period a phase is read against, the one T48 and T87 turn on, is a line of the region and never a fact of the instrument: a concert, a match and a sensor-s cadence each define theirs, and two heads that read one region with its period agree on its coherence with no clock between them; structural resemblance, one origin, not a claim about the world |  | [yes](../samples/yes/H10-the-period-is-a-line-of-the-region.json) | The falsifier: two heads that read one region, each with the period of its own clock, and agree on its coherence. A period each instrument brought would be the instrument's, and the heads would need a clock between them to agree. | not a derivation |\n| H2-density-attracts | conjecture | field | A1-encounter, A6-vacuity | where reading already is, the next reading lands: over eight epochs of placement in proportion to the origins a cell already carries, the ratio between the densest tenth and the sparsest rises at every epoch and the deserts stay empty, while the same budget placed blind to density holds that ratio near two and touches three times as many cells |  | [yes](../samples/yes/H2-density-attracts.json) | The falsifier: encounters that land independently of the density already there. Then the ratio does not rise with the epochs and the cells touched grow with the budget, which is what the blind placement in the yes sample shows. | not a derivation |\n| H3-a-distant-signature-moves-a-cell-that-did-not-move | conjecture | lattice | T46-a-bound-moves-cells-no-walk-reaches, T47-a-two-pole-object | stated as a hint and never as a claim about the world: if what is is a field of floors and ceilings, a distant signature changes a cell whose own record never moved, and in two hundred trials of one cell holding one value the verdict came out both ways, so nothing stored at the cell explains the difference and what explains it is the bound the cell rests on; the resemblance is structural and never arithmetic, with no continuous amplitudes and no inequality violated by number; structural resemblance, one origin, not a claim about the world |  | [yes](../samples/yes/H3-a-distant-signature-moves-a-cell-that-did-not-move.json) | The falsifier: one act that is both local and non-local — a change that travels a distance and also reaches a cell that does not rest on its bound. The two poles would then be one, and this would be a claim about the world rather than a hint. | not a derivation |\n| H4-enriched-semilattice | conjecture | lattice | T50-signatures-are-meets, T49-amplitude-has-sign-and-phase, T41-distance-is-the-walk-between-lookings | the cells under the order of rest, with signatures as meets and the phase each origin carries as what they are enriched over, look like an enriched meet-semilattice, and what this library proves would be its representation theory; identity, composition, associativity, commuting and naturality hold for the operations as vectors, while observation is not one of them — it moves the metric and can split a cell no signature could, so it is a functor out of the category and not an arrow inside it, and the join, which does not travel, is a second kind of arrow this conjecture does not name |  | [yes](../samples/yes/H4-enriched-semilattice.json) | The falsifier: a pair of the five operations whose composition the axioms do not determine. The join is already known to be absent from them: it does not travel, and naming what it is would be a second kind of arrow this conjecture does not have. | not a derivation |\n| H5-hilbert-is-the-obligatory-without-four-fields | conjecture | lattice | H3-a-distant-signature-moves-a-cell-that-did-not-move, T49-amplitude-has-sign-and-phase, T36-agreement-has-a-resolution, T45-cells-that-rest-on-one-bound-are-entangled | amplitudes with a phase and no origin, no signed ceiling, no resolution and no epoch band are what is left of a cell when those four fields are dropped, so the measurement problem is the absence of who, imposed symmetries the absence of a signed ceiling, the classical limit the absence of resolution and universal interference the absence of bands; restored, four claims in phase keep exactly the share of the whole their number leaves them against an environment spread over the period, and six origins around one value are six objects read finely and one read coarsely, so how many things there are is a reading; structural resemblance, one origin, not a claim about the world |  | [yes](../samples/yes/H5-hilbert-is-the-obligatory-without-four-fields.json) | The falsifier: a result of the usual formalism that cannot be stated with the four fields, or a prediction the restored fields make that the arithmetic forbids. This conjecture is about which structure is missing, not about numbers the arithmetic already gives. | not a derivation |\n| H6-structure-of-measurement | conjecture | lattice | D11-the-obligatory, A0-carrier, A1-encounter, A3-resolution | four instruments that share no mechanism fold the same way because a measurement is a cell: for a thermometer, a survey, an interferometer and a headcount the five fields are named without strain and the same five operations answer for all four; no number is claimed, only that the table can be filled in every row |  | [yes](../samples/yes/H6-structure-of-measurement.json) | The falsifier: an instrument whose reading cannot be written as a floor and a ceiling held by an origin at an epoch and a resolution, without straining any of the five. | not a derivation |\n| H7-the-medium-of-encounters | conjecture | field | A1-encounter, T41-distance-is-the-walk-between-lookings, D1-region | what carries an encounter is a medium, and a medium is what two origins can both reach: in each instrument the medium can be named and the encounter happens where two origins touch it and nowhere else, so two origins at one cell make information and the same two at two cells make none; structural resemblance, one origin, not a claim about the world |  | [yes](../samples/yes/H7-the-medium-of-encounters.json) | The falsifier: an encounter between two origins with nothing they both reach. Then information would appear where nothing was shared. | not a derivation |\n| H8-refereed-game | conjecture | lattice | T57-a-refereed-game-governs-itself, A1-encounter, T50-signatures-are-meets | a field of cells with origins and signatures is a game whose referee is inside it: a move is a claim, a claim alone is potential, two that meet are information, a signature moves every cell that rests on it, and no player may suspend the rule that makes the first four; the claim is that each of five readings of the field matches all five properties; structural resemblance, one origin, not a claim about the world |  | [yes](../samples/yes/H8-refereed-game.json) | The falsifier: a reading of the field where a player can suspend the rule that two origins are needed, or where one claim is information. Then the referee would be outside the game. | not a derivation |\n| H9-the-obligatory-is-maximal | conjecture | lattice | A8-four-states, A9-everything-is-a-bound, D11-the-obligatory | the obligatory is maximal: take away any one of its seven parts and what is left is an object already known, and anything added to it is either already inside it or breaks the four states |  | [yes](../samples/yes/H9-the-obligatory-is-maximal.json) | The falsifier: a part the obligatory lacks that keeps the four states and the six operations and is not a reading of one of them. | not a derivation |\n| T1-only-verdict-bits-travel | theorem | chain | D4-debit-in-bits, A2-meet | an origin that moves a verdict carries at least the bits by which it reduces the cell's debit, and exactly those suffice |  | [yes](../samples/yes/T1-only-verdict-bits-travel.json) | derived by the reader | yes |\n| T10-drift | corollary | quotient | A5-origin-is-extension | two definitions of one question diverge without either failing; the divergence is invisible until an encounter joins them |  | [yes](../samples/yes/T10-drift.json) | derived by the reader | yes |\n| T100-an-act-is-paid-with-attention | theorem | chain | T93-every-act-is-paid-by-its-origin, A6-vacuity, A7-freedom-conserved, T90-freedom-is-entropy | an origin has finite looks per epoch; what it looks at closes a bit of the cell and a bit of itself, and what it stops looking at reopens by vacuity. Freedom plus what is closed is conserved exactly; the field settles where closing equals forgetting, so what one attention can hold real is bounded by its looks over the forgetting rate. The price of an act is attention, and the entropy goes to the grey. |  | [yes](../samples/yes/T100-an-act-is-paid-with-attention.json) | What would break T100, run against the object. | yes |\n| T101-fragility-is-the-fewest-withdrawals | theorem | field | A5-origin-is-extension, T85-withdrawal-is-an-anti-claim, T67-withdrawal-is-reinterpretation | a cell is as fragile as the fewest origins whose withdrawal moves one of its poles: a pole one origin sets breaks with that origin, and a pole several origins set together holds until all of them withdraw | `frag(c) = min { \\|W\\| : W ⊆ O(c), ⊥(c∖W) ≠ ⊥(c) or ⊤(c∖W) ≠ ⊤(c) }` | [yes](../samples/yes/T101-fragility-is-the-fewest-withdrawals.json) | Two origins that each set one pole, called fragile at two: withdrawing either alone moves a pole. | yes |\n| T102-a-combine-meets-one-cell-and-coarsens-many | corollary | field | A2-meet, T5-coarsen-widens | a combine of bounds meets them where they land on one cell and coarsens them where they land on many: two readers of one field fold to their meet, which neither stands beside nor forks from, and readers of several fields keep their own cells and are read together only at the region they share |  | [yes](../samples/yes/T102-a-combine-meets-one-cell-and-coarsens-many.json) | Today's fold of the two validators: both lines stand and a fork is named, which the meet refuses. | yes |\n| T103-a-reading-holds-at-what-it-read | theorem | field | A4-epoch, T21-reading-is-an-origin, D8-present | a reading holds at what it read, as it was when it was read: its epoch is the state of the atoms it touched, so once those atoms move the reading stays an origin of that past and says nothing of the present until it is read again; a text, a table, a schema and a tree hold their readings the same way |  | [yes](../samples/yes/T103-a-reading-holds-at-what-it-read.json) | A reading taken before its atoms moved, called present. | yes |\n| T104-concurrent-withdrawal-cancels | corollary | field | T85-withdrawal-is-an-anti-claim, T49-amplitude-has-sign-and-phase | an observation and its withdrawal cancel iff they are concurrent, neither resting on the other: within one lock the same epoch is the concurrent case and across locks only rest decides, so a withdrawal that rests on its observation is turned by the epochs between them and leaves interference |  | [yes](../samples/yes/T104-concurrent-withdrawal-cancels.json) | What would break T104, run against the object: a withdrawal that rests on its observation, turned by the epochs between them, still cancelling it. | yes |\n| T11-size-is-questions | theorem | quotient | T10-drift, A6-vacuity | the size of a body of knowledge is its count of distinct questions with an encounter; the rest is drift |  | [yes](../samples/yes/T11-size-is-questions.json) | derived by the reader | yes |\n| T12-reliability-is-encounter-rate | theorem | valuation | A1-encounter, T10-drift | the fraction of a domain's claims that are facts equals its second-origin rate; drift and surviving falsehood are its complement |  | [yes](../samples/yes/T12-reliability-is-encounter-rate.json) | derived by the reader | yes |\n| T13-c_n | corollary | valuation | D6-premium | c_n = log2 n − 2 + 2/n bounds a pairwise audit's conservativeness in bits |  | [yes](../samples/yes/T13-c_n.json) | derived by the reader | yes |\n| T14-supermodular-degeneration | theorem | valuation | D6-premium, A1-encounter | premium is zero iff the valuation is supermodular there: a second origin adds nothing |  | none | derived by the reader | yes |\n| T15-premium-is-residual-information | theorem | valuation | D6-premium, thm-gap | the premium equals residual information exactly for valuations with the data-processing property |  | none | derived by the reader | yes |\n| T16-encounter-inequality | theorem | chain | A1-encounter, D4-debit-in-bits, T14-supermodular-degeneration | an independent second origin never raises a cell's debit: it falls by log2 of the states it leaves standing, a full bit exactly when it admits at most half of them, and nothing exactly when it narrows nothing |  | [yes](../samples/yes/T16-encounter-inequality.json) | The falsifier: a drop of a full bit where the second origin left fifteen of the sixteen states standing. | yes |\n| T17-coding-of-verdicts | theorem | chain | T1-only-verdict-bits-travel, D4-debit-in-bits | no claim shorter than the debit reduction moves a verdict; none longer is needed |  | none | derived by the reader | yes |\n| T18-freedom-is-a-currency | theorem | valuation | A7-freedom-conserved, D9-yield, A5-origin-is-extension | closed bits with a second origin are scarce, unforgeable, measurable, transferable and third-party verifiable; no issuer exists |  | [yes](../samples/yes/T18-freedom-is-a-currency.json) | derived by the reader | yes |\n| T19-signatures-fold | theorem | lattice | A2-meet, D7-fork | two signed bounds on one cell fold by form: narrower wins, equal is an encounter, incomparable is a fork; nothing is ever replaced |  | [yes](../samples/yes/T19-signatures-fold.json) | derived by the reader | yes |\n| T2-coordinate-where-disagree | theorem | field | A1-encounter, D0-coordinate | a new coordinate appears exactly where two origins disagree at the same resolution |  | [yes](../samples/yes/T2-coordinate-where-disagree.json) | derived by the reader | yes |\n| T20-age-is-encounters | theorem | field | A4-epoch, A1-encounter, D8-present | a region's age is the encounters since its last observation, not elapsed clock; the width of its present is the inverse of its encounter rate |  | [yes](../samples/yes/T20-age-is-encounters.json) | derived by the reader | yes |\n| T21-reading-is-an-origin | theorem | galois | A5-origin-is-extension, A1-encounter | a reading is a claim by the reader on the cell at its resolution — a weak encounter; undisputed density is agreement |  | [yes](../samples/yes/T21-reading-is-an-origin.json) | derived by the reader | yes |\n| T22-fold-is-superposition | theorem | field | A2-meet, A4-epoch, A1-encounter | two claims fold when their epochs meet and their intervals meet; epochs that meet with intervals that do not are a fork, and epochs that do not meet are no encounter at all |  | [yes](../samples/yes/T22-fold-is-superposition.json) | derived by the reader | yes |\n| T23-domain-has-a-spectrum | theorem | field | T22-fold-is-superposition, T20-age-is-encounters, A3-resolution | two origins meet at the common multiple of their periods, so a domain has a spectrum of resolutions where encounters happen; read coarser, a claim recovers the fine cells that lay under it |  | [yes](../samples/yes/T23-domain-has-a-spectrum.json) | derived by the reader | yes |\n| T24-ignorance-has-coordinates | theorem | valuation | D10-open-question, A7-freedom-conserved | what is not known has coordinates: the open questions rank by premium against dependents and cost, and a small head of that order holds most of the value |  | [yes](../samples/yes/T24-ignorance-has-coordinates.json) | derived by the reader | yes |\n| T25-rules-earn-their-place | theorem | quotient | A6-vacuity, T1-only-verdict-bits-travel, T2-coordinate-where-disagree | counted by mentions, a corpus of rules grows vacuous without bound; counted by the verdicts they move, vacuity converges: a rule earns its place by moving a verdict |  | [yes](../samples/yes/T25-rules-earn-their-place.json) | derived by the reader | yes |\n| T26-knowing-has-four-states | theorem | lattice | A9-everything-is-a-bound, A8-four-states, A1-encounter | knowing grows by disagreement: most cells are born where two origins split, and what stays grey is a small remainder |  | [yes](../samples/yes/T26-knowing-has-four-states.json) | derived by the reader | yes |\n| T27-disagreement-can-be-provoked | theorem | field | T2-coordinate-where-disagree, A7-freedom-conserved, A5-origin-is-extension | spending a budget where the premium is highest closes more than spending it blindly, and provoking the disagreement closes more still and names coordinates that were not there |  | [yes](../samples/yes/T27-disagreement-can-be-provoked.json) | derived by the reader | yes |\n| T3-fixed-point | theorem | product | A2-meet, A1-encounter | two heads folding the same claims converge to one fold; the fixed point is unique |  | none | derived by the reader | yes |\n| T36-agreement-has-a-resolution | theorem | chain | A3-resolution, T2-coordinate-where-disagree, T7-sufficient-resolution | n origins on a cell agree at a resolution; raising it opens agreements into forks, each a coordinate the coarser scale hid; the sufficient resolution of a region is where its residual figure stops changing |  | [yes](../samples/yes/T36-agreement-has-a-resolution.json) | derived by the reader | yes |\n| T37-a-lock-is-a-mould | theorem | dual | A0-carrier, A6-vacuity, T25-rules-earn-their-place, T36-agreement-has-a-resolution | stretching every ceiling by one factor separates what no width can hold, a cell whose origins do not meet, from a ceiling that was merely tight; pouring another world into a lock counts the places it fills; and a lock deforms with use, what was touched closing toward what was seen there and what was not opening, toward the shape the world gives it |  | [yes](../samples/yes/T37-a-lock-is-a-mould.json) | derived by the reader | yes |\n| T38-origins-are-a-polygon | theorem | product | A1-encounter, A2-meet, A5-origin-is-extension, T2-coordinate-where-disagree | n independent origins on a cell are the vertices of a polygon whose interior is their meet: two an edge, three an area, four a vertex the others do not enclose, five a verdict that holds without any one of them, and beyond that a meet that narrows toward consensus — a meet that never closes at any n is two cells wearing one name |  | [yes](../samples/yes/T38-origins-are-a-polygon.json) | derived by the reader | yes |\n| T39-cells-grow-until-stopped | theorem | product | A0-carrier, A1-encounter, A3-resolution, T7-sufficient-resolution | a cell grows until something stops it: a wall the lock names, or the place where the next origin is nearer than its own, so a cell alone between two walls is as wide as they leave it and one that lands beside another takes the midpoint; what stops a cell is never its own doing, and an origin landing beside it narrows it and widens nothing |  | [yes](../samples/yes/T39-cells-grow-until-stopped.json) | derived by the reader | yes |\n| T4-composition | remark | product | A2-meet | receipts compose in any order; the fold of a union is the meet of the folds |  | none | derived by the reader | not a derivation |\n| T40-cost-by-density | theorem | field | T38-origins-are-a-polygon, T27-disagreement-can-be-provoked, T18-freedom-is-a-currency | what knowing costs is not set by how many origins look but by where they land: one budget of two hundred gives a ratio near two placed blind and above twenty placed where reading already is, so the shape of what is known is a fact about placement, and it leaves cities and deserts in one field |  | [yes](../samples/yes/T40-cost-by-density.json) | derived by the reader | yes |\n| T41-distance-is-the-walk-between-lookings | theorem | field | T40-cost-by-density, T23-domain-has-a-spectrum, T2-coordinate-where-disagree | two cells are one step apart when an origin looked at both, so a world is the graph of its lookings and distance is the walk between them: zero on a cell against itself, symmetric, never shortened by a third cell, and across a desert no looking crosses there is no number at all, which is the horizon |  | [yes](../samples/yes/T41-distance-is-the-walk-between-lookings.json) | derived by the reader | yes |\n| T42-no-privileged-observer | theorem | field | A5-origin-is-extension, A9-everything-is-a-bound, A4-epoch, D8-present, T40-cost-by-density, T17-coding-of-verdicts | two observers of one world hold their own origins and read the same cells differently, and that is not a contradiction: wherever they hold the same origins at a cell they agree, every time, so the disagreement is the difference of their horizons and never of the world |  | [yes](../samples/yes/T42-no-privileged-observer.json) | derived by the reader | yes |\n| T43-each-origin-gives-one-fact-and-one-distance | theorem | galois | A1-encounter, A9-everything-is-a-bound, T41-distance-is-the-walk-between-lookings, A6-vacuity | information and distance are two readings of one thing: with no origin there is neither a fact nor a finite distance, and each origin that lands gives exactly one of each |  | [yes](../samples/yes/T43-each-origin-gives-one-fact-and-one-distance.json) | derived by the reader | yes |\n| T44-an-origin-is-worth-its-reach | theorem | field | T41-distance-is-the-walk-between-lookings, T27-disagreement-can-be-provoked, T18-freedom-is-a-currency, A5-origin-is-extension | one claim of one width is worth what it connects: nothing inside a city where every pair already had a walk, one pair between two lone cells, and the product of two components when it lands across a desert, so where to look is a question with an answer and the answer is read before the claim lands |  | [yes](../samples/yes/T44-an-origin-is-worth-its-reach.json) | derived by the reader | yes |\n| T45-cells-that-rest-on-one-bound-are-entangled | theorem | product | A0-carrier, A9-everything-is-a-bound, T19-signatures-fold, T41-distance-is-the-walk-between-lookings | cells that rest on one bound move together when it moves: narrowing a ceiling turns cells red with nobody looking at them, some past the horizon of whoever signed it, and the effect is the bound's and not the order's, since the same narrowing taken in two steps names exactly the cells it names in one |  | [yes](../samples/yes/T45-cells-that-rest-on-one-bound-are-entangled.json) | derived by the reader | yes |\n| T46-a-bound-moves-cells-no-walk-reaches | theorem | lattice | A0-carrier, T41-distance-is-the-walk-between-lookings, T45-cells-that-rest-on-one-bound-are-entangled | a world changes in two ways at once: looking travels and stops at the horizon, a bound does not travel and carries no step, so moving it changes cells no walk reaches; counting the bound as if it were a looking would make the whole world one neighbourhood, which is how the two are told apart |  | [yes](../samples/yes/T46-a-bound-moves-cells-no-walk-reaches.json) | derived by the reader | yes |\n| T47-a-two-pole-object | theorem | lattice | A0-carrier, A7-freedom-conserved, D4-debit-in-bits, T45-cells-that-rest-on-one-bound-are-entangled, T46-a-bound-moves-cells-no-walk-reaches | one coordinate on a scale of a hundred and five levels is superposed at six point seven bits, definite when two claims meet at one level, and forbidden rather than narrow when a ceiling falls under that floor, while a second coordinate resting on the bound that moved narrows to five point four bits with nobody measuring it |  | [yes](../samples/yes/T47-a-two-pole-object.json) | derived by the reader | yes |\n| T48-epoch-is-phase | theorem | field | A4-epoch, T22-fold-is-superposition, T45-cells-that-rest-on-one-bound-are-entangled | an epoch is a phase: claims arriving at one turn of a period reinforce and claims half a period apart cancel, so sweeping the delay of half of them draws the period's own shape, and observing one route, which is fixing when those claims arrived, moves the pattern to its complement |  | [yes](../samples/yes/T48-epoch-is-phase.json) | derived by the reader | yes |\n| T49-amplitude-has-sign-and-phase | theorem | field | T48-epoch-is-phase, T19-signatures-fold, A4-epoch, A1-encounter | a claim carries a size and a direction, how much it narrows and the turn it landed on, and a withdrawal is the same size the other way: two hundred claims in phase make the square of their number and the same two hundred strewn make its order, while a write and its withdrawal at one epoch leave nothing and at different epochs do not cancel | `A(c) = \\|Σ_{o∈O} s_o w_o e^{iφ_o}\\|²` | [yes](../samples/yes/T49-amplitude-has-sign-and-phase.json) | A withdrawal two epochs after its write, called a cancellation: a write and its withdrawal cancel at one epoch only. | yes |\n| T5-coarsen-widens | theorem | chain | A3-resolution, A6-vacuity | a claim coarsened to a lower resolution fits at least as many regions |  | [yes](../samples/yes/T5-coarsen-widens.json) | derived by the reader | yes |\n| T50-signatures-are-meets | theorem | lattice | A2-meet, A9-everything-is-a-bound, T19-signatures-fold | a signature narrows a ceiling and travels to every cell that rests on it as a meet, and meets do not remember the order they were taken in, so no order of any number of signatures ever shows in the field they leave; widening is not a meet and does not travel, which is why a narrowing costs one signature and an error costs one at every cell, and folding the field and signing it commute even with a withdrawal present | `sign: ⊤(c′) ← ⊤(c′) ∧ s ∀ c′ ⊒ c` | [yes](../samples/yes/T50-signatures-are-meets.json) | derived by the reader | yes |\n| T51-the-shape-of-knowing-can-be-designed | theorem | field | T44-an-origin-is-worth-its-reach, T41-distance-is-the-walk-between-lookings, T45-cells-that-rest-on-one-bound-are-entangled, T49-amplitude-has-sign-and-phase | for a budget of origins there is a siting that minimises what a field cannot reach, and it is found by bridging the two largest islands in turn, because what an origin is worth is the product of the two sides it joins; in three hundred cells left with four thousand three hundred and eighty unreachable pairs, twelve origins dropped blindly closed between none and a fifth of them and the same twelve chosen by reach closed four fifths |  | [yes](../samples/yes/T51-the-shape-of-knowing-can-be-designed.json) | derived by the reader | yes |\n| T52-directing-where-origins-land-closes-the-horizon | theorem | field | T51-the-shape-of-knowing-can-be-designed, A6-vacuity, T44-an-origin-is-worth-its-reach, A9-everything-is-a-bound | a field left to itself concentrates where it already is and keeps its deserts: under attachment its horizon falls slowly and does not close, in sixty epochs and in either draw, while the same budget spent each epoch where reach is greatest closes it in four; the machine that does it holds nothing and asserts nothing, reads the field as it stands, answers only which two islands to bridge, and is itself a cell of the field it weaves |  | [yes](../samples/yes/T52-directing-where-origins-land-closes-the-horizon.json) | derived by the reader | yes |\n| T53-a-decision-is-what-the-form-leaves-open | theorem | lattice | A0-carrier, A8-four-states, D3-state, T26-knowing-has-four-states | a lock is mostly settled: the form of most claims leaves them one legal state and they are applied, what the form leaves wider splits into those a reading has closed and those standing in force with nothing read against them, and the few that say of themselves that they are decisions are the only ones open; the four classes are a partition, every claim in exactly one, and a decision is not a claim with a wide ceiling but a claim that declares itself unanswered |  | [yes](../samples/yes/T53-a-decision-is-what-the-form-leaves-open.json) | derived by the reader | yes |\n| T55-a-policy-is-four-weights | theorem | field | A1-encounter, T44-an-origin-is-worth-its-reach, T51-the-shape-of-knowing-can-be-designed, T49-amplitude-has-sign-and-phase | a policy is four weights and nothing else, over the reach of a cell, how together its claims arrived, what stands beside it and what it still leaves free; the law is the relation, that a policy carrying all four closes at least as much as any policy carrying one and forks no more than those that close at all, and the counts belong to whoever reproduces it, here fifty-two closed against fifty-one and forty-eight with the fewest forks of the three |  | [yes](../samples/yes/T55-a-policy-is-four-weights.json) | derived by the reader | yes |\n| T56-forks-over-closes-separates-regimes | theorem | field | T55-a-policy-is-four-weights, D7-fork, T27-disagreement-can-be-provoked, T36-agreement-has-a-resolution | how knowing moves through a field is one number, the forks it opens against the closes it makes, and that number tells regimes apart that no count of looks can: one policy over one field, with claims drawn near the meet, across it, or from two origins alone, separates into three regimes two orders apart |  | [yes](../samples/yes/T56-forks-over-closes-separates-regimes.json) | derived by the reader | yes |\n| T57-a-refereed-game-governs-itself | theorem | field | A1-encounter, A5-origin-is-extension, T55-a-policy-is-four-weights, T21-reading-is-an-origin | the rule that makes the game is in the field and not above it: a cell one origin has claimed is grey and nothing that origin does alone will close it, so what an origin is worth is asymmetric, nothing over its own claims and everything over another-s, and the referee is a rule no player can suspend because it is the rule that makes a claim information |  | [yes](../samples/yes/T57-a-refereed-game-governs-itself.json) | derived by the reader | yes |\n| T58-observation-is-over-regions-by-reach | theorem | field | T41-distance-is-the-walk-between-lookings, A3-resolution, D1-region, T44-an-origin-is-worth-its-reach | a domain is read by regions and not by its atoms: what the lock names are regions, what the readers touch are the atoms those regions happen to hold, and a domain holds several atoms to every region, so one that grows a thousand atoms does not grow a thousand questions, and the cost of knowing it follows the regions and never the count of its atoms |  | [yes](../samples/yes/T58-observation-is-over-regions-by-reach.json) | derived by the reader | yes |\n| T6-present-converges | theorem | field | D8-present, A6-vacuity | the width of the present is maturation rate minus forgetting rate; a system with only past or only future has no present |  | [yes](../samples/yes/T6-present-converges.json) | derived by the reader | yes |\n| T63-author-time-memory-are-relative | theorem | galois | A4-epoch, A5-origin-is-extension, D0-coordinate, T2-coordinate-where-disagree | who said it, when they said it and what is remembered are three coordinates of one claim and none of them is the claim: two origins saying one span are two claims and one origin saying it twice is one, so what a cell holds together depends on the spans alone while what it is depends on all three |  | [yes](../samples/yes/T63-author-time-memory-are-relative.json) | derived by the reader | yes |\n| T65-ownership-is-being-an-origin | theorem | field | A5-origin-is-extension, A1-encounter, D5-declaration, T8-exact-withdrawal | to own a claim is to be the origin of it and nothing else: a cell holds what each origin said, taking an origin away takes exactly what it said and no more, and no narrowing survives the origin that made it |  | [yes](../samples/yes/T65-ownership-is-being-an-origin.json) | derived by the reader | yes |\n| T66-there-is-no-after | theorem | field | A4-epoch, T22-fold-is-superposition, T50-signatures-are-meets, D8-present | there is no after: what stands at a cell is what the latest epoch says and the order the claims arrived in never decides, so every order of one set of claims leaves one reading and a clock is not a fact about the field |  | [yes](../samples/yes/T66-there-is-no-after.json) | derived by the reader | yes |\n| T67-withdrawal-is-reinterpretation | theorem | field | T8-exact-withdrawal, A4-epoch, A1-encounter, T19-signatures-fold | a withdrawal takes back what a claim did and never that it was made: the claim stays on the cell, the origin still spoke, and what changed is what the cell now means, which is why a withdrawal is a reinterpretation and not an erasure |  | [yes](../samples/yes/T67-withdrawal-is-reinterpretation.json) | derived by the reader | yes |\n| T7-sufficient-resolution | corollary | chain | A3-resolution | for every verdict there is a coarsest resolution that still moves it, strictly coarser than the finest available |  | [yes](../samples/yes/T7-sufficient-resolution.json) | derived by the reader | yes |\n| T70-a-gap-is-a-coarse-claim | theorem | field | A6-vacuity, A3-resolution, D1-region, T5-coarsen-widens | a gap is not the absence of a claim but a claim at the coarsest resolution: a cell nobody has narrowed holds the whole scale, which is exactly what saying nothing says, so a reading that is missing and a reading that allows everything are one line and a tree with gaps is a tree read coarsely |  | [yes](../samples/yes/T70-a-gap-is-a-coarse-claim.json) | derived by the reader | yes |\n| T73-a-cell-that-folds-itself-converges | theorem | galois | A9-everything-is-a-bound, A7-freedom-conserved, A1-encounter | a cell that folds itself converges: observing each epoch a claim of width ten centred on its own meet and off by at most s, it narrows only when the offset cuts inside what it holds, so its freedom settles toward log2(11 − 2s) from above and it never conflicts; at s = 3, sixty epochs leave it between 2.38 and 2.43 bits on three seeds |  | [yes](../samples/yes/T73-a-cell-that-folds-itself-converges.json) | derived by the reader | yes |\n| T8-exact-withdrawal | corollary | complement | D2-ideal | a withdrawal is exact iff the scale is distributive |  | none | derived by the reader | yes |\n| T85-withdrawal-is-an-anti-claim | corollary | field | T49-amplitude-has-sign-and-phase | a withdrawal is the claim it takes back with its sign turned: landed in phase with it, the two leave nothing, the cell goes back to grey and every bit the claim closed is open again |  | [yes](../samples/yes/T85-withdrawal-is-an-anti-claim.json) | What would break T85, run against the object. | yes |\n| T87-the-quarter-turn | theorem | field | T48-epoch-is-phase, T49-amplitude-has-sign-and-phase, A8-four-states | two claims reinforce within a quarter turn of each other and cancel beyond it, since \\|e^{ia} + e^{ib}\\|² < 2 exactly when \\|a − b\\| > π/2: a fork is destructive interference, and a field is laminar while fewer than half its pairs lie beyond a quarter turn, forks over closes below one | `\\|e^{ia} + e^{ib}\\|² < 2 ⟺ \\|a − b\\| > π/2` | [yes](../samples/yes/T87-the-quarter-turn.json) | What would break T87, run against the object. | yes |\n| T88-turbulence-is-factions | corollary | field | T87-the-quarter-turn | phases spread evenly give forks over closes of exactly one; above one is several coherences facing each other, and only a third coherence lowers it |  | [yes](../samples/yes/T88-turbulence-is-factions.json) | What would break T88, run against the object. | yes |\n| T89-disagreement-is-where-to-look | corollary | field | T87-the-quarter-turn | disagreement is where to look: a policy that weights the cells whose origins lie beyond a quarter turn of each other closes more than one blind to them |  | [yes](../samples/yes/T89-disagreement-is-where-to-look.json) | What would break T89, run against the object. | yes |\n| T9-self-application-converges | theorem | quotient | A6-vacuity, A1-encounter, A9-everything-is-a-bound | a system folded by its own ceilings converges to what its encounters recognize; the residue is unique |  | [yes](../samples/yes/T9-self-application.json) | derived by the reader | yes |\n| T90-freedom-is-entropy | theorem | chain | D4-debit-in-bits, A7-freedom-conserved, T48-epoch-is-phase | freedom is log₂ of the states between the poles, which is the cell's entropy: encounters lower it, vacuity raises it, and nothing else moves it; two paths interfere from phase alone, and that is lost once which path was taken is observed |  | [yes](../samples/yes/T90-freedom-is-entropy.json) | What would break T90, run against the object. | yes |\n| T91-many-objects-make-a-field | corollary | product | T92-a-cell-is-made-of-cells | many objects make a field: cells read pairwise meet or fork, the meets join them into components, and the whole can have no meet although most of its pairs do |  | [yes](../samples/yes/T91-many-objects-make-a-field.json) | What would break T91, run against the object. | yes |\n| T92-a-cell-is-made-of-cells | theorem | product | A9-everything-is-a-bound, A3-resolution, A2-meet | origins, bounds, claims and rests are cells, and the state a cell shows is the meet of its parts' states; reading deeper is reading at a finer resolution, and it converges |  | [yes](../samples/yes/T92-a-cell-is-made-of-cells.json) | What would break T92, run against the object. | yes |\n| T93-every-act-is-paid-by-its-origin | theorem | field | A2-meet, A5-origin-is-extension, T44-an-origin-is-worth-its-reach | an act is a meet on its origin's own ceiling: the medium and the policy are cells too, reach is the meet of regions, being in phase the meet of bands and choosing what fits, and there is no third object |  | [yes](../samples/yes/T93-every-act-is-paid-by-its-origin.json) | What would break T93, run against the object. | yes |\n| T94-a-signature-stops-where-the-meet-is-identity | theorem | lattice | A2-meet, T45-cells-that-rest-on-one-bound-are-entangled, T50-signatures-are-meets | propagation of a sign or require halts at any cell where the new bound is absorbed (meet(a,b) = a), because every cell resting on it already carries a bound at least as tight; a region's bounds summarise as an interval (min ceiling, max floor) and a signature outside the interval never enters the region. Signing costs what it bites, never what lies beneath. |  | [yes](../samples/yes/T94-a-signature-stops-where-the-meet-is-identity.json) | What would break T94, run against the object. | yes |\n| T95-signatures-live-where-they-are-signed | theorem | lattice | A2-meet, T3-fixed-point, T45-cells-that-rest-on-one-bound-are-entangled, T94-a-signature-stops-where-the-meet-is-identity | a signature or requirement is a line at the bound, never a value copied into resting cells; a cell's poles are read up its chain of rest as the meet of the live signatures above it; an absorbed signature stays live so withdrawing a tighter one reveals it. Signing and withdrawing cost one line; reading costs the depth of rest; nothing propagates, and A2 with T3 makes the reading equal to any replay. | `⊤(c) = ⋀_{b ⊑ c} signed(b) ∧ ⋀_{o∈O} hi_o, one line per act` | [yes](../samples/yes/T95-signatures-live-where-they-are-signed.json) | What would break T95, run against the object. | yes |\n| T99-the-fold-is-total | theorem | lattice | A2-meet, T9-self-application-converges, A9-everything-is-a-bound | the fold is a monotone map on a lattice over a rest order with no cycles: every fold reaches its fixed point in finitely many steps and never fails to answer. The lock is not Turing complete by construction — it cannot loop — so verify can check it, rank can read it before landing, and accept can refuse before running. Arbitrary computation lives only in capsules, outside the fold, admitted by vectors and heard as claims. |  | [yes](../samples/yes/T99-the-fold-is-total.json) | What would break T99, run against the object. | yes |\n| lem-identity | theorem | spectrum | A0-carrier, A2-meet, D5-declaration | The graph identity — If $x\\wedge y=\\bot$ then \\[ \\RI_H(x,y)\\;=\\;\\log_2 n-\\frac{\\Phi(x,y)}{n}, \\qquad \\Phi(x,y)=\\sum_{e}c_e\\log_2\\frac{a_e\\,b_e}{c_e}, \\] the sum running over the edges of $G(x,y)$, where $c_e$ is the weight of $e$ and $a_e,b_e$ are the sizes of the two blocks it joins. |  | [yes](../samples/yes/lem-identity.json) | derived by the reader | yes |\n| lem-join | corollary | spectrum | lem-identity | Reduction to a discrete join — Every pair with trivial meet can be transformed, without decreasing $\\RI_H$ and in finitely many steps, into a pair on the same $n$ states with trivial meet and discrete join. |  | [yes](../samples/yes/lem-join.json) | derived by the reader | yes |\n| lem-mediant | corollary | spectrum | lem-merge | Components, and the mediant — Both $\\Sigma(m)-\\Sigma(u)$ and $\\Sigma(w)-\\Sigma(j)$ decompose as sums over the blocks of $m$, which are the connected components of the block graph $G(u,w)$; and a component contributing $0$ to the denominator contributes $0$ to the numerator. Hence the ratio is at most the largest ratio of a single component, each of which is a declaration on $n_t\\le n$ states with trivial meet. |  | none | derived by the reader | yes |\n| lem-meet | theorem | spectrum | lem-identity, prop-cmi | Reduction to a trivial meet — Let $\\tau(N)$ denote the maximum of $\\RI_H$ over pairs on $N$ states with trivial meet. Then for any $x,y\\in\\Pin$ with $x\\wedge y$ having blocks of sizes $n_1,\\dots,n_r$, \\[ \\RI_H(x,y)=\\sum_{t=1}^{r}\\frac{n_t}{n}\\,\\RI_H\\bigl(x\\|_{B_t},y\\|_{B_t}\\bigr) \\;\\le\\;\\sum_t \\frac{n_t}{n}\\,\\tau(n_t). \\] Each restricted pair has trivial meet. If $\\tau$ is non-decreasing the bound is $\\tau(n)$; and if $\\tau$ is moreover strictly increasing above $N=2$, equality for $n\\ge3$ forces $r=1$. |  | [yes](../samples/yes/lem-meet.json) | derived by the reader | yes |\n| lem-merge | theorem | spectrum | A0-carrier, A2-meet, D5-declaration | Every coarsening step costs two — For integers $p,q\\ge1$, $\\varphi(p+q)-\\varphi(p)-\\varphi(q)\\ge2$, with equality exactly at $p=q=1$. Consequently, if $P$ coarsens $Q$ from $\\|Q\\|$ blocks to $\\|P\\|$ blocks then $\\Sigma(P)-\\Sigma(Q)\\ge2\\bigl(\\|Q\\|-\\|P\\|\\bigr)$. |  | [yes](../samples/yes/lem-merge.json) | derived by the reader | yes |\n| lem-phi | theorem | spectrum | lem-join, lem-meet | The integer inequality — For every integer $t\\ge1$, $\\varphi(t)\\ge 2(t-1)$, with equality exactly at $t\\in\\{1,2\\}$. |  | [yes](../samples/yes/lem-phi.json) | derived by the reader | yes |\n| prop-additive | theorem | valuation | D2-ideal, D3-state | What is additive under independence, and what is not — Let $L=L_1\\times L_2$ with each factor of at least two elements, and let $f=f_1\\oplus f_2$. Then $\\viol_f$ and $\\viols_f$ are exactly additive on product declarations: two-sided constraints admit no economy of scale under independence. The vocabulary is not: \\[ \\delta(L_1\\times L_2)\\;=\\;\\|L_1\\|\\|L_2\\|-\\|MI(L_1)\\|-\\|MI(L_2)\\|, \\] which exceeds $\\delta(L_1)+\\delta(L_2)$ by exactly $(\\|L_1\\|-1)(\\|L_2\\|-1)-1$, strictly unless both factors are the two-element lattice. |  | [yes](../samples/yes/prop-additive.json) | derived by the reader | yes |\n| prop-cmi | theorem | valuation | A0-carrier, A2-meet, D5-declaration | cf. — $\\RI_H(x,y)=I(x;y\\mid x\\wedge y)$. |  | [yes](../samples/yes/prop-cmi.json) | derived by the reader | yes |\n| prop-dep | corollary | valuation | thm-collapse | Both costs, decomposed — For every declaration on $L$, with $u,w,m,j$ as usual and $\\viol_i,g_i$ computed from $f_i$ on coordinate $i$, \\[ \\viol_f \\;=\\; \\viol_1+\\viol_2+\\bigl[D(u)-D(m)\\bigr], \\qquad g_f \\;=\\; g_1+g_2+\\RI_D(u,w), \\] the second under the data-processing property for $f,f_1,f_2$. |  | [yes](../samples/yes/prop-dep.json) | derived by the reader | yes |\n| prop-halfboolean | theorem | complement | D2-ideal, D3-state | Half Boolean — For every finite lattice and every debit $j$, the limit side is truth-functional: $j\\le w_1\\wedge w_2$ iff $j\\le w_1$ and $j\\le w_2$. The demand side --- whether $j\\le u_1\\vee u_2$ is determined by $(j\\le u_1,\\ j\\le u_2)$ --- is truth-functional if and only if $L$ is distributive. |  | none | derived by the reader | yes |\n| thm-collapse | theorem | valuation | D6-premium, thm-gap | Collapse, and its converse — The following are equivalent for a monotone valuation $f$. · $f$ has the data-processing property. · For every $u,w$: $\\viols_f(u,w)=f(u\\vee w)-f(w)$, attained at $S=u\\vee w$. · For every $u,w$: $g(u,w)=\\RI_f(u,w)$. |  | none | derived by the reader | yes |\n| thm-decomp | theorem | ideal | D2-ideal, D3-state | Exact decomposition — On every finite lattice $u=\\bigvee\\{j\\in\\JI(L):j\\le u\\}$ and $w=\\bigwedge\\{m\\in\\MI(L):w\\le m\\}$. Hence $\\reg$ is injective: a declaration is its register, with nothing lost. |  | none | derived by the reader | yes |\n| thm-dep | theorem | valuation | prop-dep, thm-mi | Additivity up to the mutual information — For every declaration on a product of two subsystems under a joint measure, \\[ \\viol_1+\\viol_2-I_\\mu\\;\\le\\;\\viol_f\\;\\le\\;\\viol_1+\\viol_2, \\qquad \\bigl\\|\\,g_f-(g_1+g_2)\\,\\bigr\\|\\;\\le\\;I_\\mu . \\] The conflict is therefore always discounted by dependence and never inflated, while the premium is always inflated and never discounted by more than the same amount. Both ends that are reached are reached by different declarations: the left inequality on the conflict at $u=\\top$, $w=\\bot$, and $g_f=g_1+g_2+I_\\mu$ at the crossed declaration of Theorem . The lower end of the band on the premium is not known to be attained, and Remark says what reaching it would require. |  | [yes](../samples/yes/thm-dep.json) | derived by the reader | yes |\n| thm-extremal | theorem | spectrum | lem-identity, lem-join, lem-meet, lem-phi | The extremal — For every $n\\ge1$, \\[ c_n\\;=\\;\\log_2 n-2+\\frac{2}{n}, \\] and for $n\\ge3$ the maximum is attained exactly at the following configuration and its relabellings (for $n\\le2$ the maximum is $0$ and every pair attains it): identify the $n$ states with the $n$ edges of a path on $n+1$ vertices; let $x$ group the edges sharing an even-numbered endpoint and $y$ those sharing an odd-numbered endpoint. These have $\\lceil (n+1)/2\\rceil$ and $\\lfloor (n+1)/2\\rfloor$ blocks, every block of size $1$ or $2$, trivial meet and discrete join. |  | [yes](../samples/yes/thm-extremal.json) | derived by the reader | yes |\n| thm-gap | theorem | valuation | A0-carrier, A2-meet, D5-declaration | The gap, exactly — For every lattice and every monotone $f$, \\[ g(u,w)\\;=\\;\\max\\bigl\\{\\,f(p)+f(q)-f(p\\vee q)\\;:\\;p\\le u,\\ q\\le w\\,\\bigr\\}\\;-\\;f(m), \\] the maximum being at least $f(m)$, as $p=q=m$ shows. |  | none | derived by the reader | yes |\n| thm-local | theorem | valuation | A0-carrier, A2-meet, D5-declaration | Retreat is locally optimal — For every lattice and every monotone $f$, the retreat $S_0=m$ is a weak local minimum of $C$ with respect to cover moves. |  | [yes](../samples/yes/thm-local.json) | derived by the reader | yes |\n| thm-mi | corollary | valuation | thm-collapse | Mutual information is a planning premium — Take the declaration that demands the first subsystem be answerable and forbids the state to exceed the second: $u=(\\top,\\bot)$, $w=(\\bot,\\top)$. Then \\[ \\viol_f=H(X),\\qquad \\viols_f=H(X\\mid Y),\\qquad g_f=I_\\mu(X;Y), \\] while $g_1=g_2=0$. |  | [yes](../samples/yes/thm-mi.json) | derived by the reader | yes |\n| thm-modular | corollary | valuation | thm-gap | Degeneration — $g\\equiv0$ --- retreat is optimal for every declaration --- if and only if $f$ is supermodular, $f(x)+f(y)\\le f(x\\vee y)+f(x\\wedge y)$ for all $x,y$. Modular valuations are the special case of equality. |  | none | derived by the reader | yes |\n| thm-ratio | theorem | spectrum | lem-mediant, lem-merge | The retreat ratio — For every $n\\ge2$, over declarations in conflict on $\\Pin$ under Shannon entropy, \\[ \\max\\frac{\\viol_H}{\\viols_H}\\;=\\;\\rho(n)\\;:=\\;\\frac{n}{2}\\,H_{bal}(n,2) \\;=\\;\\frac{\\varphi(n)-\\varphi\\bigl(\\lceil\\tfrac n2\\rceil\\bigr) -\\varphi\\bigl(\\lfloor\\tfrac n2\\rfloor\\bigr)}{2}, \\] attained exactly when the demand $u$ is a balanced two-block partition, the block graph is connected, and the limit merges two singletons of $u\\vee w$. In particular the maximum equals $n/2$ exactly for even $n$, and $n/2-\\Theta(1/n)$ for odd $n$. |  | [yes](../samples/yes/thm-ratio.json) | derived by the reader | yes |\n| thm-witness | theorem | valuation | D2-ideal, D3-state | The witness — Let $F$ be in conflict. Then $F$ is sound for every monotone valuation if and only if \\[ \\textrm{(W)}\\qquad u\\in A\\quadand\\quad \\exists\\,b\\in B:\\ u\\wedge b=m . \\] |  | [yes](../samples/yes/thm-witness.json) | Every conflicting pair of families in b2 called witnessed: condition (W) holds for 144 of the 188, and the claim of 188 is refused. | yes |\n\n## Forms\n\n- **alphabetForm**, a form of WideMask\n- **continuousForm**, a form of { lo: number; hi: number }\n- **ladderForm**, a form of Unit\n- **latticeForm**, a form of Unit\n\n## Units\n\n| unit | signature |\n|---|---|\n| asUnit | `(c: Scale, d: Declaration, subject = '')` |\n| compose | `(c: Scale, units: readonly Unit[])` |\n| fuse | `(c: Scale, units: readonly Unit[])` |\n| rebuild | `(c: Scale, reg: Register, subject = '')` |\n| unit | `(subject: string, bounds: { floor: string; ceiling: string })` |\n| unitOf | `(c: Scale, place: Point)` |\n| widen | `(c: Scale, units: readonly Unit[])` |\n\n## Alphabet\n\n| symbol | names |\n|---|---|\n| ⊥ | floor |\n| ⊤ | ceiling |\n| O | origins |\n| ⊑ | rest |\n| (o, φ) | Origin |\n| s | Act |\n| state | State |\n" at=witness:041/retain-declared-documentation by=target form=alphabet measure=text role=writes scope=prose/en/reference shape=docs/reference.md value=lock
 bound-lock/1 about="# Signature\n\n```mermaid\nflowchart LR\n  c0[\"cell\"]\n  c1[\"meet\"]\n  c2[\"sign\"]\n  l0([\"the obligatory\"])\n  c0 --> l0\n  c1 --> l0\n  classDef shown fill:#2da44e,color:#ffffff,stroke:#2da44e\n  class l0 shown\n```\n\nThe object is c = ⟨⊥, ⊤, O, ⊑⟩ over L at r: four parts in two pairs, the value (⊥, ⊤) and the relations (O, ⊑), with the lattice L it is given and the resolution r it is read at belonging to the reading and the sign to the act; an origin is (o, φ), and the poles are read, never stored: ⊥ is what is required joined with what its live claims hold, ⊤ what is signed met with what they allow; eight acts in four pairs and seven names answer for it: meet and join on the ceiling, local, a join only with a witness; observe and withdraw on the floor, local, a withdrawal being an observation at −1 that restores exactly; sign and require on both poles, travelling by rest, commuting and both closing freedom; encounter and refine with no pole; and no widening travels.\n\n## Run it\n\n```ts\nimport { intervals } from '@lapxo/obligations/forms';\nimport { cell, meet, sign } from '@lapxo/obligations/views/field';\n\nconst scale = intervals(0, 100);\nconst world = [0, 1, 2, 3].map((i) => cell(`c${i}`, 0, [{ origin: `own${i}`, span: { lo: 20 * i, hi: 100 } }], ['ceiling']));\nconst before = new Map([['ceiling', cell<{ lo: number; hi: number }>('ceiling')]]);\nconst after = sign(scale, before, 'ceiling', { lo: 0, hi: 45 });\n\nconsole.log('cells resting on the ceiling', world.filter((one) => one.restsOn.includes('ceiling')).length);\nfor (const [i, one] of world.entries()) {\n  console.log(`c${i} before`, JSON.stringify(meet(scale, one, before)), '· after', JSON.stringify(meet(scale, one, after)));\n}\nconsole.log('claims edited by the signature', world.flatMap((one) => one.seen).filter((claim) => claim.span.hi !== 100).length);\nconsole.log('why       one signature narrows a bound and every cell resting on it reads the narrower one, without touching a single claim any origin made');\n```\n\n```\ncells resting on the ceiling 4\nc0 before {\"lo\":0,\"hi\":100} · after {\"lo\":0,\"hi\":45}\nc1 before {\"lo\":20,\"hi\":100} · after {\"lo\":20,\"hi\":45}\nc2 before {\"lo\":40,\"hi\":100} · after {\"lo\":40,\"hi\":45}\nc3 before {\"lo\":60,\"hi\":100} · after {\"lo\":60,\"hi\":45}\nclaims edited by the signature 0\nwhy       one signature narrows a bound and every cell resting on it reads the narrower one, without touching a single claim any origin made\n```\n\nIts origins meet.\n\n## Source\n\n[Its source](../../examples/release/signature.ts) is one of the examples of obligations, and it runs as it is written, against the built package, with nothing compiled for it.\n" at=witness:041/retain-declared-documentation by=target form=alphabet measure=text role=writes scope=prose/en/learn-signature shape=docs/learn/signature.md value=lock
@@ -71,6 +71,9 @@ bound-lock/1 about="For two unit complex amplitudes with real phases, let delta=
 bound-lock/1 about="How to read it" at=witness:the-root-spoke-for-the-place by=target form=alphabet measure=text role=writes scope=prose/en/reading/heading value=lock
 bound-lock/1 about="In a common probability space with justified declared failure upper bounds p_i, the probability of all declared events succeeding is at least max(0,1-sum(p_i)); sum(p_i)>=1 is trivial. Agreement and copy counts do not estimate probabilities. Historical alias: freedom-is-a-currency; no transferability or unforgeability claim." at=policy:law/T18 by=target form=alphabet kind=theorem measure=status needs=samples/measures/confidence/union-bound.json|vectors/measures/confidence/expected.json role=demands scope=law/T18-freedom-is-a-currency value=present
 bound-lock/1 about="In a finite undirected graph with distinct named vertices, positive component sizes and unrestricted unit-cost fresh bridges that join only their two chosen components, a nonnegative integer budget b minimises unreachable unordered pairs by connecting the min(b+1,k) largest of k components into one. The objective is (N^2-sum(componentSize^2))/2; arbitrary site restrictions, nonuniform costs and origin-name collisions are outside this theorem. design returns a star on representatives of those components. Historical comparisons against particular blind seeds remain experiments, not universal superiority over every competing draw." at=policy:law/T51 by=target form=alphabet kind=theorem measure=status needs=samples/yes/T51-the-shape-of-knowing-can-be-designed.json|src/views/product/design.ts|src/views/product/reach.ts|test/obligatory.test.ts restsOn=law/A0-carrier|law/A5-origin-is-extension role=demands scope=law/T51-the-shape-of-knowing-can-be-designed value=present
+bound-lock/1 about="L7: conditional bound on one declared common probability space: P(N>k)<=min(1,sum over |S|=k+1 of min p_i), with rational declared marginal failure upper bounds and 0<=k<=m. Trivial iff computed upper bound is one. Counterexamples refute the former top-k+1 success guarantee. Joint-independent Bernoulli convolution is a separate declared hypothesis; rho=0 and pairwise independence do not supply it. Existing stronger 0.4.2 count bound is unchanged and needs no joint independence." at=policy:law/L7 by=target form=alphabet kind=definition measure=status needs=samples/measures/l7-tolerant-k.json|vectors/laws/l7-tolerant-k.json|vectors/laws/policy.json restsOn=law/T18-freedom-is-a-currency role=demands scope=law/L7-tolerant-k value=present view=valuation
+bound-lock/1 about="L8: law by enumeration only for 3<=n<=64, positive integer divisors p|n and integer 0<=s<p. D=log2(n/p)-h(s/p), explicit binary64-log2 approximation. Odd-period minimum uses h(floor(p/2)/p), even-period minimum uses one. Complete maximum at zero; nonzero domain has endpoints 1,p-1, empty when p=1. No information or worst-case interpretation follows from nonnegativity. T87 retains its separate wrapped-phase amplitude identity." at=policy:law/L8 by=target form=alphabet kind=definition measure=status needs=samples/measures/l8-phase-defect.json|vectors/laws/l8-phase-defect.json|vectors/laws/policy.json role=demands scope=law/L8-F4-integer-phase-defect value=present view=valuation
+bound-lock/1 about="L9: law by counterexample of existence of noninjective weighted mix, and measure with a declared prior on a finite explicitly supplied set of named-factor configurations. Equal-weight ab and ba give the same bottom tie and different named forks. Loss is H(configuration | mix)=H(configuration)-H(mix); a uniform prior must be explicitly selected, absent prior is unmeasured. The four-state uniform example has mean loss 0.5 bits and one bit conditioned on tie; the injective named fork has zero loss. Defines no new algebraic act, governance selection, M/3 or M/4 law, or destroyed field." at=policy:law/L9 by=target form=alphabet kind=definition measure=status needs=samples/measures/l9-mix-loss.json|vectors/laws/l9-mix-loss.json|vectors/laws/policy.json role=demands scope=law/L9-mix-projection-loss value=present view=valuation
 bound-lock/1 about="Learn from an example: {pages}" at=witness:the-root-spoke-for-the-place by=target form=alphabet measure=text role=writes scope=prose/en/pointers/learn value=lock
 bound-lock/1 about="M/3 is the only admitted conditional model. Superseded descriptive D/commonMass claims are retired; observed commonMass remains descriptive, never strength or independence." at=witness:descriptive-measures-beside-states by=target form=alphabet measure=text role=writes scope=rationale/040/model-text value=lock
 bound-lock/1 about="MIT License\n\nCopyright (c) 2026 Lapxo\n\nPermission is hereby granted, free of charge, to any person obtaining a copy\nof this software and associated documentation files (the \"Software\"), to deal\nin the Software without restriction, including without limitation the rights\nto use, copy, modify, merge, publish, distribute, sublicense, and/or sell\ncopies of the Software, and to permit persons to whom the Software is\nfurnished to do so, subject to the following conditions:\n\nThe above copyright notice and this permission notice shall be included in all\ncopies or substantial portions of the Software.\n\nTHE SOFTWARE IS PROVIDED \"AS IS\", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR\nIMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,\nFITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE\nAUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER\nLIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,\nOUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE\nSOFTWARE." at=policy:license by=target form=alphabet measure=id role=demands scope=license value=MIT
@@ -87,7 +90,7 @@ bound-lock/1 about="Open research demand: A separately declared reported-value m
 bound-lock/1 about="Open research demand: An explicit observation channel and identification conditions before a probe can stand for truth." at=witness:041/cleanup-models-and-contracts by=target form=interval measure=count needs=publication/models/imperfect-probes role=demands scope=model/0.4.1/imperfect-probes value=0..0
 bound-lock/1 about="Open research hypothesis; the unit-amplitude identity in T87 proves no policy advantage. Comparing disagreement-weighted and blind policies requires a declared objective, population and sampling model. Historical alias: disagreement-is-where-to-look." at=policy:law/T89 by=target form=alphabet kind=conjecture measure=status needs=samples/yes/T89-disagreement-is-where-to-look.json|samples/no/T89-disagreement-is-where-to-look.json restsOn=law/T87-the-quarter-turn role=demands scope=law/T89-disagreement-is-where-to-look value=present
 bound-lock/1 about="Open research interpretation; not a consequence of T87. Evenly spaced phases do not always give a forks/closes ratio of one: phases 0, 2pi/3, 4pi/3 give three cancelling pairs and zero closing pairs. General faction interpretation remains open. Historical alias: turbulence-is-factions." at=policy:law/T88 by=target form=alphabet kind=conjecture measure=status needs=samples/yes/T88-turbulence-is-factions.json|samples/no/T88-turbulence-is-factions.json restsOn=law/T87-the-quarter-turn role=demands scope=law/T88-turbulence-is-factions value=present
-bound-lock/1 about="Public API, rendered contracts, compact data and native captures, including the existing files the README links. Validation worlds and their modular dependencies are separate digest-addressed assets; no private executor is shipped in the library." at=witness:041/retain-declared-illustrations by=target form=alphabet measure=id role=reads scope=audit/wire/ships value=dist|README.md|LICENSE|TARGET.bound|receipts.bound|docs/measures.md|samples/measures|vectors/measures|samples/yes/algebra-is-closed.json|samples/yes/field-uses-price-and-debit.json|samples/yes/three-theories-one-algebra.json|src/lattice/index.ts|src/field/motion.ts|samples/yes/installed-artifact.json|samples/no/installed-artifact.json|vectors/measures/installed-artifact.cli.json|docs/img|docs/paper.md|docs/reference.md|docs/learn/in-phase.md|docs/learn/price.md|docs/learn/signature.md|docs/learn/thermometer.md|docs/learn/withdrawal.md
+bound-lock/1 about="Public API, rendered contracts, compact data and native captures, including the existing files the README links. Validation worlds and their modular dependencies are separate digest-addressed assets; no private executor is shipped in the library." at=witness:043/native-closure by=target form=alphabet measure=id role=reads scope=audit/wire/ships value=dist|README.md|LICENSE|TARGET.bound|receipts.bound|docs/measures.md|samples/measures|vectors/measures|samples/yes/algebra-is-closed.json|samples/yes/field-uses-price-and-debit.json|samples/yes/three-theories-one-algebra.json|src/lattice/index.ts|src/field/motion.ts|samples/yes/installed-artifact.json|samples/no/installed-artifact.json|vectors/measures/installed-artifact.cli.json|docs/img|docs/paper.md|docs/reference.md|docs/learn/in-phase.md|docs/learn/price.md|docs/learn/signature.md|docs/learn/thermometer.md|docs/learn/withdrawal.md|vectors/laws/policy.json|vectors/laws/l7-tolerant-k.json|vectors/laws/l8-phase-defect.json|vectors/laws/l9-mix-loss.json|vectors/laws/implementation-043.json
 bound-lock/1 about="Read admitted obligations and their native evidence; unit-test success alone cannot make a release ready." at=witness:descriptive-measures-beside-states by=target form=alphabet measure=id role=writes scope=release/blocks value=algebra|contracts|artifact
 bound-lock/1 about="Read the existing instrument programs region and its real release-policy verdict; this declares no new operation or library export." at=policy:041/release-verdict-visible by=target form=alphabet measure=id role=demands scope=view/programs value=programs
 bound-lock/1 about="Reduction to a discrete join — Every pair with trivial meet can be transformed, without decreasing $\\RI_H$ and in finitely many steps, into a pair on the same $n$ states with trivial meet and discrete join." at=policy:law/lem-join by=target form=alphabet kind=corollary measure=status needs=samples/yes/lem-join.json|test/f1-boundary-042.test.ts|vectors/laws/f1/path-witness.json restsOn=law/lem-identity role=demands scope=law/lem-join value=present
@@ -120,7 +123,7 @@ bound-lock/1 about="Research status: experiment, one-origin structural interpret
 bound-lock/1 about="Research status: experiment, one-origin structural interpretation; general/physical-law reading open. stated as a hint and never as a claim about the world: if what is is a field of floors and ceilings, a distant signature changes a cell whose own record never moved, and in two hundred trials of one cell holding one value the verdict came out both ways, so nothing stored at the cell explains the difference and what explains it is the bound the cell rests on; the resemblance is structural and never arithmetic, with no continuous amplitudes and no inequality violated by number; structural resemblance, one origin, not a claim about the world" at=policy:law/H3 by=target form=alphabet kind=conjecture measure=status needs=samples/yes/H3-a-distant-signature-moves-a-cell-that-did-not-move.json|samples/no/H3-a-distant-signature-moves-a-cell-that-did-not-move.json restsOn=law/T46-a-bound-moves-cells-no-walk-reaches|law/T47-a-two-pole-object role=demands scope=law/H3-a-distant-signature-moves-a-cell-that-did-not-move value=present
 bound-lock/1 about="Research status: open beyond the supplied sufficient-resolution fixtures. In a finite declared chain of resolutions, a coarsest resolution preserving the finest verdict exists because the finest itself qualifies; it need not be strictly coarser than the finest. The historical unconditional strict-coarsening claim is not a theorem." at=policy:law/T7 by=target form=alphabet kind=conjecture measure=status needs=samples/yes/T7-sufficient-resolution.json|samples/no/A3-resolution.json|vectors/info/sufficient-resolution.json restsOn=law/A3-resolution role=demands scope=law/T7-sufficient-resolution value=present
 bound-lock/1 about="Run it" at=witness:the-root-spoke-for-the-place by=target form=alphabet measure=text role=writes scope=prose/en/learn/run value=lock
-bound-lock/1 about="Scoped conditional measures and law boundaries beside the unchanged algebra." at=witness:042/release-metadata by=target form=alphabet measure=id role=writes scope=version shape=package.json value=0.4.2
+bound-lock/1 about="Scoped conditional measures and law boundaries beside the unchanged algebra." at=witness:043/native-closure by=target form=alphabet measure=id role=writes scope=version shape=package.json value=0.4.3
 bound-lock/1 about="T58 counted one historical tree. Retire the universal leaves>regions assertion; preserve the historical observation and the expected-refusal counterexample with a legal empty region." at=witness:descriptive-measures-beside-states by=target form=alphabet measure=text role=writes scope=rationale/040/T58 value=lock
 bound-lock/1 about="The 43 expected-refusal failures came from a borrowed adapter that ignored expected.ok=false; restore the original Obligations comparison. The field case loop now checks its returned boolean. No algebra or expected vector answer changes." at=witness:descriptive-measures-beside-states by=target form=alphabet measure=text role=writes scope=rationale/040/test-adapter value=lock
 bound-lock/1 about="The complete descriptive and M/3 contract, controlled procedures and captured vectors, rendered from own prose." at=witness:040/presentation-two-pages by=target form=alphabet measure=id role=demands scope=view/measures shape=docs/measures.md value=text
@@ -216,7 +219,8 @@ bound-lock/1 about="lane C signs the object and its second head: obligations, ve
 bound-lock/1 about="lane C signs the object, its second head and the sources, the leaves a rebirth of them writes, topos-doc while it waits, and the registry's word that its two places are published" at=witness:a-published-demand-its-lane-could-not-pay by=target form=alphabet measure=coverage role=reads scope=keys/lane-C value=obligations/**|verify/**|sources/**|topos-doc/**|region/obligations/**|region/verify/**|region/sources/**|write/obligations/**|write/verify/**|published/obligations/**|published/verify/**
 bound-lock/1 about="lane D's window widened to 380 for the base: receipts at a resolution through the view that ships them at @0 and @1, the agnosticism test over the measurements world, and laying the base's releases by digest as they are re-pinned; 188 kept as the start so every line this key already signed stays inside its window" at=policy:keys/lane-D/the-base by=target form=interval measure=epoch role=reads scope=keys/lane-D value=188..380
 bound-lock/1 about="lines that stand on one origin and are therefore not yet information" at=policy:region/grey by=target form=alphabet measure=text role=writes scope=region/grey value=lock
-bound-lock/1 about="name: release\non:\n  workflow_run:\n    workflows: [tag]\n    types: [completed]\n  workflow_dispatch:\njobs:\n  stage:\n    if: github.event_name == 'workflow_dispatch' || (github.event.workflow_run.conclusion == 'success' && github.event.workflow_run.head_branch == 'main')\n    concurrency: release-delivery\n    environment: release\n    runs-on: ubuntu-latest\n    permissions:\n      contents: write\n      id-token: write\n    steps:\n      - uses: actions/checkout@v4\n        with:\n          ref: ${{ github.event.workflow_run.head_sha || github.sha }}\n      - uses: actions/setup-node@v4\n        with:\n          node-version: 24\n      - run: npm install --prefix .instrument --no-save --ignore-scripts @lapxo/bound@0.1.8\n      - run: npm install --global npm@'12.1.0' --ignore-scripts\n      - name: Prepare release\n        env:\n          GH_TOKEN: ${{ github.token }}\n        run: |\n          set -euo pipefail\n          TAG=\"v$(node -p \"require('./package.json').version\")\"\n          printf 'TAG=%s\\n' \"$TAG\" >> \"$GITHUB_ENV\"\n          git fetch origin \"refs/tags/$TAG:refs/tags/$TAG\"\n          gh api --paginate \"repos/$GITHUB_REPOSITORY/releases\" --jq '.[].tag_name' > \"$RUNNER_TEMP/release-tags\"\n          if grep -Fxq \"$TAG\" \"$RUNNER_TEMP/release-tags\"; then\n            printf 'RELEASE %s · exists\\n' \"$TAG\"\n          else\n            gh release create \"$TAG\" --verify-tag --draft --title \"$TAG\" --generate-notes\n            printf 'RELEASE %s · draft created · awaiting signed assets\\n' \"$TAG\"\n          fi\n      - name: Read owner approval\n        env:\n          GH_TOKEN: ${{ github.token }}\n        run: |\n          set -euo pipefail\n          mkdir .release\n          gh release download \"$TAG\" --pattern approval-signed --dir .release\n          node .instrument/node_modules/@lapxo/bound/dist/cli/verb.js fold --as release .release/approval-signed > .release/verdict\n          cat .release/verdict\n      - name: Stage the approved release archive\n        env:\n          GH_TOKEN: ${{ github.token }}\n        run: |\n          set -euo pipefail\n          read -r status digest archive source extra < .release/verdict\n          test \"$status\" = READY\n          test -z \"$extra\"\n          gh release download \"$TAG\" --pattern \"$archive\" --dir .release\n          printf '%s  %s\\n' \"${digest#sha256:}\" \".release/$archive\" | sha256sum -c -\n          expected=\"$(node -p \"require('./package.json').name+'@'+require('./package.json').version\")\"\n          actual=\"$(tar -xOf \".release/$archive\" package/package.json | node -p \"const p=JSON.parse(require('fs').readFileSync(0,'utf8'));p.name+'@'+p.version\")\"\n          test \"$actual\" = \"$expected\"\n          printf 'SOURCE release/%s/%s · %s · declared %s\\n' \"$TAG\" \"$archive\" \"$digest\" \"$source\"\n          gh release edit \"$TAG\" --draft=false\n          printf 'RELEASE %s · published · %s\\n' \"$TAG\" \"$digest\"\n          npm stage publish \".release/$archive\" --access public --ignore-scripts\n          printf 'STAGED %s · %s · owner approval pending\\n' \"$actual\" \"$digest\"\n" at=witness:041/published-bound-018 by=target form=alphabet measure=text role=writes scope=prose/en/pipeline/approved-delivery shape=.github/workflows/release.yml value=lock
+bound-lock/1 about="name: release\non:\n  workflow_run:\n    workflows: [tag]\n    types: [completed]\n  workflow_dispatch:\njobs:\n  stage:\n    if: github.event_name == 'workflow_dispatch' || (github.event.workflow_run.conclusion == 'success' && github.event.workflow_run.head_branch == 'main')\n    concurrency: release-delivery\n    environment: release\n    runs-on: ubuntu-latest\n    permissions:\n      contents: write\n      id-token: write\n    steps:\n      - uses: actions/checkout@v4\n        with:\n          ref: ${{ github.event.workflow_run.head_sha || github.sha }}\n      - uses: actions/setup-node@v4\n        with:\n          node-version: 24\n      - run: npm install --prefix .instrument --no-save --ignore-scripts @lapxo/bound@0.2.1\n      - run: npm install --global npm@'12.1.0' --ignore-scripts\n      - name: Prepare release\n        env:\n          GH_TOKEN: ${{ github.token }}\n        run: |\n          set -euo pipefail\n          TAG=\"v$(node -p \"require('./package.json').version\")\"\n          printf 'TAG=%s\\n' \"$TAG\" >> \"$GITHUB_ENV\"\n          git fetch origin \"refs/tags/$TAG:refs/tags/$TAG\"\n          gh api --paginate \"repos/$GITHUB_REPOSITORY/releases\" --jq '.[].tag_name' > \"$RUNNER_TEMP/release-tags\"\n          if grep -Fxq \"$TAG\" \"$RUNNER_TEMP/release-tags\"; then\n            printf 'RELEASE %s · exists\\n' \"$TAG\"\n          else\n            gh release create \"$TAG\" --verify-tag --draft --title \"$TAG\" --generate-notes\n            printf 'RELEASE %s · draft created · awaiting signed assets\\n' \"$TAG\"\n          fi\n      - name: Read owner approval\n        env:\n          GH_TOKEN: ${{ github.token }}\n        run: |\n          set -euo pipefail\n          mkdir .release\n          gh release download \"$TAG\" --pattern approval-signed --dir .release\n          node .instrument/node_modules/@lapxo/bound/dist/cli/verb.js fold --as release .release/approval-signed > .release/verdict\n          cat .release/verdict\n      - name: Stage the approved release archive\n        env:\n          GH_TOKEN: ${{ github.token }}\n        run: |\n          set -euo pipefail\n          read -r status digest archive source extra < .release/verdict\n          test \"$status\" = READY\n          test -z \"$extra\"\n          gh release download \"$TAG\" --pattern \"$archive\" --dir .release\n          printf '%s  %s\\n' \"${digest#sha256:}\" \".release/$archive\" | sha256sum -c -\n          expected=\"$(node -p \"require('./package.json').name+'@'+require('./package.json').version\")\"\n          actual=\"$(tar -xOf \".release/$archive\" package/package.json | node -p \"const p=JSON.parse(require('fs').readFileSync(0,'utf8'));p.name+'@'+p.version\")\"\n          test \"$actual\" = \"$expected\"\n          printf 'SOURCE release/%s/%s · %s · declared %s\\n' \"$TAG\" \"$archive\" \"$digest\" \"$source\"\n          gh release edit \"$TAG\" --draft=false\n          printf 'RELEASE %s · published · %s\\n' \"$TAG\" \"$digest\"\n          npm stage publish \".release/$archive\" --access public --ignore-scripts\n          printf 'STAGED %s · %s · owner approval pending\\n' \"$actual\" \"$digest\"\n" at=witness:043/native-closure by=target form=alphabet measure=text role=writes scope=prose/en/pipeline/approved-delivery shape=.github/workflows/release.yml value=lock
+bound-lock/1 about="name: tag\non:\n  workflow_run:\n    workflows: [test]\n    types: [completed]\njobs:\n  tag:\n    if: github.event.workflow_run.conclusion == 'success' && github.event.workflow_run.event == 'push' && github.event.workflow_run.head_branch == 'main'\n    runs-on: ubuntu-latest\n    permissions:\n      contents: write\n    steps:\n      - uses: actions/checkout@v4\n        with:\n          ref: ${{ github.event.workflow_run.head_sha }}\n          fetch-depth: 0\n      - run: |\n          version=\"v$(node -p \"require('./package.json').version\")\"\n          git tag --list \"$version\" | grep -q . && exit 0\n          git tag \"$version\" && git push origin \"$version\"\n" at=witness:043/native-closure by=target form=alphabet measure=text role=writes scope=prose/en/pipeline/checked-tag shape=.github/workflows/tag.yml value=lock
 bound-lock/1 about="no law" at=witness:the-root-spoke-for-the-place by=target form=alphabet measure=text role=writes scope=prose/en/reference/unrested value=lock
 bound-lock/1 about="not a derivation" at=witness:the-root-spoke-for-the-place by=target form=alphabet measure=text role=writes scope=prose/en/reference/underived value=lock
 bound-lock/1 about="obligations knows no instrument: its name, @lapxo/bound and bound/, never the noun a bound, which is the theory-s" at=witness:the-noun-bound-in-an-axiom by=target form=alphabet measure=id role=reads scope=literal/src/** value=not:fs|path|process|Date|fetch|card|package|placement|judge|receipt|topos|@lapxo/bound|bound/|render|npm|github|symbol|generation|Infinity|parseFloat|toFixed|grey|agreed|split|free|required|forbidden|conflict
@@ -321,7 +325,7 @@ bound-lock/1 about="the second head's command in the job that runs both heads: i
 bound-lock/1 about="the shape of line each act is: the judge reads which act a line is here and names none in its own code" at=witness:the-judge-named-acts-in-its-own-code by=target form=alphabet measure=id role=reads scope=audit/wire/acts value="observe:role=writes|withdraw:value=withdraw|sign:reads,interval,hi|require:reads,interval,lo|join:widening,at=witness|refine:@res"
 bound-lock/1 about="the shortest example this package ships and what it prints, with the digest the lock pins for it" at=witness:the-door-said-nothing-of-what-it-owed by=target form=alphabet measure=text role=writes scope=region/first-use value=lock
 bound-lock/1 about="the signatures a line may carry, each with its era after its name: ed25519 in era1 signs what lands now, and the two before it are history, read and never written again" at=witness:an-era-named-by-a-digest-nobody-could-read by=target form=alphabet measure=id role=reads scope=audit/wire/signature-algorithms value=ed25519|sha256:58f1f0983fe65f062bebefd0a6ffc901049c67525ef12dd86ed4dbb7a184b834|ed25519:era1
-bound-lock/1 about="the tag of the version the manifest names, made once" at=witness:a-provider-chosen-inside-three-regions by=target form=alphabet measure=id needs=.github/workflows/tag.yml role=demands scope=view/pipeline-tag shape=.github/workflows/tag.yml value=github-actions
+bound-lock/1 about="the tag of the version the manifest names, made once" at=witness:043/native-closure by=target form=alphabet measure=id needs=.github/workflows/tag.yml role=demands scope=view/pipeline-tag shape=.github/workflows/tag.yml value=text
 bound-lock/1 about="the tag the forge makes for the version the manifest names, once" at=witness:the-place-carried-the-pipeline-of-another-registry by=target form=alphabet measure=text role=writes scope=region/pipeline-tag value=lock
 bound-lock/1 about="the theory of a place as its lists, each law with what met it and what would break it" at=policy:region/laws by=target form=alphabet measure=text role=writes scope=region/laws value=readings
 bound-lock/1 about="the three axioms the rest leans on hardest by their sentence, the others by name, and one theorem meeting its measurement" at=witness:the-door-said-nothing-of-what-it-owed by=target form=alphabet measure=text role=writes scope=region/what-holds value=lock
@@ -527,7 +531,6 @@ bound-lock/1 at=witness:041/installed-closure-reading by=target form=alphabet me
 bound-lock/1 at=witness:041/installed-closure-reading by=target form=alphabet measure=id role=writes scope=dep/runtime-installed-acceptance shape=test/support/installed-reader.js value=sha256:19b54c86bfb460b9068df95449a034f83e8876703366909179c95ebb94869135
 bound-lock/1 at=witness:041/installed-closure-reading by=target form=alphabet measure=id role=writes scope=sources/installed-acceptance value=https://github.com/Lapxo/obligations/releases/download/v0.4.1/a1aef1e16df8e2b1959403b9241a669e1d6342ba772569123e72b294acdbd4d2
 bound-lock/1 at=witness:041/installed-closure-reading by=target form=alphabet measure=id role=writes scope=sources/runtime-installed-acceptance value=https://github.com/Lapxo/obligations/releases/download/v0.4.1/19b54c86bfb460b9068df95449a034f83e8876703366909179c95ebb94869135
-bound-lock/1 at=witness:041/published-bound-018 by=target form=alphabet measure=id role=writes scope=pipeline/check value="npm ci && npm run build && npm test && npx @lapxo/bound@0.1.8 fold --check"
 bound-lock/1 at=witness:041/retain-declared-documentation by=target form=alphabet measure=reads role=writes scope=region/figure value=prose/en/figure
 bound-lock/1 at=witness:041/retain-declared-documentation by=target form=alphabet measure=reads role=writes scope=region/guide value=prose/en/guide
 bound-lock/1 at=witness:041/retain-declared-documentation by=target form=alphabet measure=reads role=writes scope=region/learn-in-phase value=prose/en/learn-in-phase
@@ -548,6 +551,18 @@ bound-lock/1 at=witness:042/native-publication-quartets by=target form=alphabet
 bound-lock/1 at=witness:042/native-publication-quartets by=target form=alphabet measure=id role=writes scope=dep/runtime-publication-042 shape=test/support/publication-reader.js value=sha256:4428e457e927142dc5867204d689a6575c43295583bb5789527ecdabf3c941f6
 bound-lock/1 at=witness:042/public-reader-delivery by=target form=alphabet measure=url needs=sha256:4428e457e927142dc5867204d689a6575c43295583bb5789527ecdabf3c941f6 role=writes scope=sources/runtime-publication-042 value=https://github.com/Lapxo/obligations/releases/download/v0.4.2/4428e457e927142dc5867204d689a6575c43295583bb5789527ecdabf3c941f6
 bound-lock/1 at=witness:042/public-reader-delivery by=target form=alphabet measure=url needs=sha256:ebd8b0d3cca22a3c66a6daf5c035326d2a8685b3ef1bb836e06f72a4f8ad658d role=writes scope=sources/publication-042 value=https://github.com/Lapxo/obligations/releases/download/v0.4.2/ebd8b0d3cca22a3c66a6daf5c035326d2a8685b3ef1bb836e06f72a4f8ad658d
+bound-lock/1 at=witness:043/native-closure by=target form=alphabet measure=id role=writes scope=pipeline/check value="npm ci && npm run build && npm test && npx @lapxo/bound@0.2.1 fold --check"
+bound-lock/1 at=witness:043/native-closure by=target form=alphabet measure=id role=writes scope=release/category/contracts value=publication/measures/**|publication/laws/**
+bound-lock/1 at=witness:043/native-closure by=target form=alphabet measure=url role=writes scope=sources/publication-043 value=https://github.com/Lapxo/obligations/releases/download/v0.4.3/publication-043.standing
+bound-lock/1 at=witness:043/native-closure by=target form=alphabet measure=url role=writes scope=sources/runtime-publication-043 value=https://github.com/Lapxo/obligations/releases/download/v0.4.3/publication-043.tgz
+bound-lock/1 at=witness:043/native-publication by=target form=alphabet measure=coordinates role=writes scope=region/publication-043 value=docs/measures.md|vectors/laws/implementation-043.json|samples/measures/l7-tolerant-k.json|vectors/laws/l7-tolerant-k.json|samples/measures/l8-phase-defect.json|vectors/laws/l8-phase-defect.json|samples/measures/l9-mix-loss.json|vectors/laws/l9-mix-loss.json
+bound-lock/1 at=witness:043/native-publication by=target form=alphabet measure=digest role=writes scope=uses/publication-043 value=sha256:234d6b19c1fe7bc5a03937dcbb1a9196281b67170fb0f9317ed888b1bd890471
+bound-lock/1 at=witness:043/native-publication by=target form=alphabet measure=id role=writes scope=dep/runtime-publication-043 shape=test/support/publication-reader.js value=sha256:252128112ac91f20451fdeb75a04a432403c4f718c99d6d98cd84b7e1fccc731
+bound-lock/1 at=witness:043/native-publication by=target form=interval measure=bytes role=writes scope=reader/capsule/publication-043/response-bytes value=0..65536
+bound-lock/1 at=witness:043/native-publication by=target form=interval measure=count needs=docs/measures.md|samples/measures/l7-tolerant-k.json|vectors/laws/l7-tolerant-k.json|vectors/laws/implementation-043.json role=demands scope=publication/laws/l7-tolerant-k value=0..0
+bound-lock/1 at=witness:043/native-publication by=target form=interval measure=count needs=docs/measures.md|samples/measures/l8-phase-defect.json|vectors/laws/l8-phase-defect.json|vectors/laws/implementation-043.json role=demands scope=publication/laws/l8-phase-defect value=0..0
+bound-lock/1 at=witness:043/native-publication by=target form=interval measure=count needs=docs/measures.md|samples/measures/l9-mix-loss.json|vectors/laws/l9-mix-loss.json|vectors/laws/implementation-043.json role=demands scope=publication/laws/l9-mix-loss value=0..0
+bound-lock/1 at=witness:043/native-publication by=target form=interval measure=milliseconds role=writes scope=reader/capsule/publication-043/timeout value=0..5000
 bound-lock/1 at=witness:a-need-that-names-a-leaf-that-left by=target form=alphabet measure=status needs=src/|vectors/fold.json restsOn=audit/obligations/replay-is-the-proof role=demands scope=audit/close/fold-is-derivable shape=decision value=present
 bound-lock/1 at=witness:a-need-that-names-a-leaf-that-left by=target form=alphabet measure=status needs=vectors/atom/derived.json|src/laws/laws.bound role=demands scope=audit/derived-is-not-an-origin shape=vector value=present
 bound-lock/1 at=witness:a-need-that-names-a-leaf-that-left by=target form=alphabet measure=status needs=vectors/field.json|samples/yes/|samples/no/ role=demands scope=audit/the-structure-is-a-field shape=vector value=present
@@ -632,7 +647,6 @@ bound-lock/1 at=witness:descriptive-measures-beside-states by=target form=alphab
 bound-lock/1 at=witness:descriptive-measures-beside-states by=target form=alphabet measure=id role=demands scope=view/receipts shape=receipts.bound value=receipts@8|receipts@0|receipts@1
 bound-lock/1 at=witness:descriptive-measures-beside-states by=target form=alphabet measure=id role=writes scope=release/category/algebra value=algebra-is-closed|field-uses-price-and-debit|three-theories-one-algebra
 bound-lock/1 at=witness:descriptive-measures-beside-states by=target form=alphabet measure=id role=writes scope=release/category/artifact value=release/installed-artifact
-bound-lock/1 at=witness:descriptive-measures-beside-states by=target form=alphabet measure=id role=writes scope=release/category/contracts value=publication/measures/**
 bound-lock/1 at=witness:descriptive-measures-beside-states by=target form=alphabet measure=id role=writes scope=scope/minimum value=algebra|contracts|artifact
 bound-lock/1 at=witness:descriptive-measures-beside-states by=target form=alphabet measure=signer role=writes scope=keys/owner value=file
 bound-lock/1 at=witness:descriptive-measures-beside-states by=target form=alphabet measure=url needs=sha256:0f5b05b2bd1b1f5e3641e6539e61fae87d791c4d51aa5d3d304ffdc064e7fe4d role=writes scope=sources/topos-github/standing value=https://github.com/Lapxo/topos-github/releases/download/v0.1.4/0f5b05b2bd1b1f5e3641e6539e61fae87d791c4d51aa5d3d304ffdc064e7fe4d

package/dist/measures/index.d.ts

@@ -24,3 +24,9 @@ export { coincidenceExcess } from './coincidence.ts';
 export type { ValueReading, TruthReading, CoincidenceReference, CoincidencePolicy } from './coincidence.ts';
 export { unionBoundConfidence, tolerantCountConfidence } from './confidence.ts';
 export type { DeclaredFailure } from './confidence.ts';
+export { tolerantSubsetConfidence, independentCountConfidence } from './tolerant-subsets.ts';
+export type { JointIndependenceDeclaration } from './tolerant-subsets.ts';
+export { phaseDefect, phaseDefectExtrema } from './phase-defect.ts';
+export type { PhaseDefectInterpretation } from './phase-defect.ts';
+export { mixProjectionLoss } from './mix-loss.ts';
+export type { NamedConfiguration, ConfigurationPrior, MixLossModel, MixLossPolicy } from './mix-loss.ts';

package/dist/measures/index.js

@@ -11,3 +11,6 @@ export { commonCauseCapacityBound, capacityBound, capacityP0, mutualInformation,
 export { fitUnanimityModel } from "./unanimity-fit.js";
 export { coincidenceExcess } from "./coincidence.js";
 export { unionBoundConfidence, tolerantCountConfidence } from "./confidence.js";
+export { tolerantSubsetConfidence, independentCountConfidence } from "./tolerant-subsets.js";
+export { phaseDefect, phaseDefectExtrema } from "./phase-defect.js";
+export { mixProjectionLoss } from "./mix-loss.js";