Native relational support events and synchronization¶
Sections 8–13 specify supplied attachments, reset budgets, occurrence-law nonselection and precontact locking. They retain native relational flow, operator maps and event premises as separate models; an admitted event or synchronized preparation does not select its occurrence.
Part of Native regional composition. Section numbers remain stable across this collection. Each result keeps its full hypotheses, implementation and checks; the execution plan alone assigns research work.
8. One supplied bridge: sufficient interface and endpoint-only obstruction¶
Full-state interface card¶
Take two disjoint, separately admitted connected simple unit graphs, with held nonnegative capacities and the same acute relational model \(e\geq0,\ w,\beta>0\). Add the supplied unit edge \((a,b)\), without changing form, phase or capacity. This is a support intervention, not an autonomous edge-creation law or a continuous integration step.
At each port \(i\), the following internal message suffices for its new instantaneous row:
The degree \(d_i\) counts internal neighbors, and the complex resultant retains both components, not just a mean phase. Keep the internal node state and the relative regional form/phase frames behind these messages. They reconstruct the port field at this snapshot, not the future of an autonomous coarse node. No history is required by this fully retained Markov law; discarding internal state reintroduces the existing memory obligation.
Let \(r=x_a-x_b\) and \(\delta=\operatorname{wrap}(\theta_b-\theta_a)\). Directly adding one term to each neighbor sum gives
Every nonport neighbor set is unchanged. With \(\alpha_i=\operatorname{Arg}z_i\) and \(\operatorname{sinc}(0)=1\), the existing law reads
Substitution of the primed data is the complete attachment identity; no extra force or pressure term is introduced. Old acute edges together with \(|\delta|<\pi/2\) suffice for the new acute graph: every real resultant part is positive and every \(H'_i>0\). The public observer keeps that domain; a rejected sufficient admission does not exclude all wider regular states. Zero capacity is allowed and still suppresses both continuous rows.
Only the port rows can change in the ideal local law. In particular,
Thus an unchanged cross-edge difference does not preserve the old internal row: both its degree normalization and phase metric depend on the new neighbor. Port resultant addition can also rotate the phase source even when the supplied phase gap is zero, unless the old resultant is real.
The event adds exactly
to \(S=E_D+\beta V\). Its outward form cut from the left component is \(r\). Neither quantity is the change in a local velocity. A support event need not inherit fixed-support storage dissipation; admit its budget separately.
Fixed C5 discriminator, derived before evaluation¶
Reuse the input owner prepared interaction: nodes \(0,\ldots,9\), two C5 components, bridge \((0,5)\), \(x=\epsilon e_1,\ \epsilon=1/256\), \(\theta_k=\theta_{k+5}=2\pi k/5\), unit capacities and \(e=w=1/2,\ \beta=1\). Regional phase offset is zero; the left and right form means remain \(\epsilon/5\) and zero. Set \(c=\cos(2\pi/5)\), \(h=1+2c\); the exact identity \(2ch=1\) is specific to this cycle and is not a universal geometric selection principle.
At donor port 0, \(q_0=-\epsilon\) is unchanged, while \(d_0:2\to3,\ z_0:2c\to h,\ H_0:2\pi c\to\pi h\). At recipient port 5 the same metric change occurs but \(q_5=0\). All ideal phase sources initially vanish. Therefore
| Quantity at donor port 0 | Separate component | Joined support | Joined minus separate |
|---|---|---|---|
| Form rate | \(\epsilon/4\) | \(\epsilon/6\) | \(-\epsilon/12=-1/3072\) |
| Phase rate | \(-\epsilon/(4\pi c)\) | \(-\epsilon/(2\pi h)\) | \(+\epsilon/(2\pi)=1/(512\pi)\) |
Every other ideal form/phase rate is unchanged, including the initially stationary recipient. The bridge differences, outward cut and event storage jump are all zero. Nevertheless the total continuous loss changes from \(3\epsilon^2/2\) to \(17\epsilon^2/12\): it decreases by \(\epsilon^2/12\), while the storage derivative increases by the same amount.
This is an exact counterexample to keeping the old component rows and adding only a term that vanishes when the endpoint form and phase agree. It rejects that explicit control, not every endpoint model with richer retained data. The sufficient interface above retains precisely the internal quantities that the control discards; no minimality theorem is claimed.
The existing interaction basin already admits this ideal preparation and eventual local recovery on this single-bridge support. Its proof is reused, not rederived. The separate two-adjacent-bridge capture API does not admit this graph.
Represented admission and shared implementation¶
The read-only attachment observer evaluates each connected component separately, rebuilds only their captured state on a detached graph, adds the supplied bridge, and evaluates the joined field through the same native owner. It never sends the disconnected union to the relational evaluator. The existing support-transport owner may read that union for cut and exact form-energy reset accounting.
The field retains its already computed real/imaginary relative resultant. Port cards, full component/joined fields and rational differences retain degree, gradient, metric, source and arithmetic evidence without reimplementing pressure or introducing evolution. The thin SDK delegate and detached report export share that observation; neither applies the edge to a live network.
Binary64 phase lifts are not exact multiples of ideal pi. Current native
execution captures each relative neighbor resultant once and uses those sums
for both the pressure source and phase metric. The field records
pressure_path="relative_resultant_canonical". Sharing the captured source
does not eliminate the retained pressure-split and rate rounding defects or
bound transcendental evaluation error. Static comparisons use explicit
absolute tolerances for implementation agreement, not a certified ideal-input
or ODE error bound. The fixed C5 controls use binary64, no randomness or
integration step, and absolute rate tolerance \(10^{-15}\) with zero relative
tolerance. Frozen response artifacts retain their recorded arithmetic paths.
The static controls exercise the rational zero-phase case, this fixed C5 witness, storage addition and domain/read-only boundaries. They add no trajectory, parameter scan or replacement historical response. Exact-real unchanged interior rows and the actual represented differences remain distinct observations.
Disposition: instantaneous interface admission is established in this scope and the named endpoint-only control is excluded. The existing transmission study already supplies the temporal response, so this result does not justify another F4 transmission campaign. It exposes a missing premise instead: which admissible support intervention occurs, if any. The support-event analysis separates inherited accounting from an additional passivity premise and an occurrence law. The single queue owns subsequent work; this report supplies no selector.
9. Support-event budgets and the missing occurrence law¶
State card and reusable event owners¶
Keep a fixed finite node set, signed form \(x\), circular phase \(\theta\), held capacities \(\nu_i\geq0\), and the same coefficients \(e\geq0,\ w,\beta>0\). Each support is simple, undirected and unit-weighted. Every consumed edge is strictly acute. Connected components have at least two nodes and are admitted separately when the support is disconnected; this does not broaden the connected-graph engine interface.
Retain relative component form and phase offsets. Common form shifts and common phase rotations preserve the accounting; independently recentering the components generally changes a candidate edge's cost.
Between events, use exactly the admitted unforced relational flow. At an event, only the edge set changes: \(x,\theta,\nu\) have identical left and right limits. Pressure, degree, resultant and phase metric are recomputed from the resulting support; they are not independent stored supplies of work. Both event endpoints must satisfy the stated model domain. This card does not admit node birth, continuous conductance evolution, a state reset, forcing or a capacity event.
The relevant owners already separate action from occurrence:
| Owner | Reusable content and boundary |
|---|---|
| Relational flow | Joint storage and fixed-support loss, with represented arithmetic defects kept separate from the exact-real identity |
| Attachment observation | Fresh component/joined fields for a supplied bridge; no event is executed or selected |
| Support transport | Exact same-node, same-form added/removed-edge Dirichlet reset; this accounts for form storage, not the full phase contribution |
| Pattern contact | A supplied attachment/removal schedule under a different sine phase law; neither that schedule nor that law follows from relational exchange |
| THOL birth and transport | Configured birth, coupling and dispatch contracts; a child construction or weighted coupling action is not this state-preserving unit-edge event |
The existing selector symmetry owner can test an independently declared finite candidate action. Symmetry may exclude a unique deterministic choice; it does not create an event clock.
Exact jump and finite hybrid balance¶
For an unordered pair \(a,b\), define its nonnegative edge storage
The declared storage on support \(E\) is \(S_E=\sum_{\{a,b\}\in E}c_{ab}\). For a state-preserving event \(E^-\to E^+\), let \(A=E^+\setminus E^-\) and \(R=E^-\setminus E^+\). Unchanged edges cancel, so the exact ideal jump is
No occurrence assumption is needed for this identity. In particular, pure addition has \(\Delta S\geq0\), even though the continuous flow satisfies
Suppose a declared trajectory is admitted on every continuous segment and has finitely many such events in \([0,T]\). Integrating each fixed-support identity and telescoping its endpoints gives
The endpoint values are before any included event at zero and after any included event at \(T\); the sum uses those same endpoint conventions. This is conditional accounting for an admitted hybrid trajectory, not a construction of one. It supplies no event schedule, non-Zeno theorem, global continuation, capture certificate on changed support or binary64 Euler-error bound. Zero capacities remain allowed; they suppress their continuous rows without removing an event's edge-storage cost.
An event budget such as \(\Delta S_k\leq W_k\) requires independently declared available work \(W_k\). Neither the fixed-support loss identity nor pressure refresh supplies that work. Past dissipation cannot be spent retrospectively as an undeclared reservoir: retaining a budget/history and its replenishment or imposing a different cumulative criterion would add premises to this state card.
Passive pure addition requires exact coincidence¶
Impose the additional event premise that no work is supplied and storage cannot increase at each event: \(\Delta S\leq0\). For pure addition, every summand is nonnegative. Consequently
The right side gives zero jump, not strict decrease. It is sufficient for the new edge's acute admission; the existing support/state checks remain required. Capacities at its two endpoints need not coincide. This restriction is derived from the selected storage plus event passivity, not from the nodal equation alone. It does not preclude work-funded attachment, simultaneous removal or an admitted nodal reset.
A zero-cost bridge can nevertheless change both port rates, as the exact C5 witness shows. More generally, at a zero-cost bridge the two port gradients \(q_i\) are unchanged and their degrees grow by one. Thus
Equality holds when each affected \(e\nu_i q_i^2\) vanishes. Otherwise the new connection decreases instantaneous dissipation despite having no storage jump. This is not a prediction that its entire future dissipates less.
Continuous-intensity obstruction on an open full-state domain¶
Fix one candidate missing edge \(a,b\) and an open admitted full-state domain \(U\) for the pre-event components. Forms and phases are independent coordinates there; in particular, a sufficiently small change of \(x_a\) is allowed. Capacities may be fixed, including zero. Consider a continuous nonnegative occurrence intensity \(\lambda_{ab}:U\to[0,\infty)\) with the requirement that whenever \(\lambda_{ab}(X)>0\), its state-preserving pure addition satisfies the passive-event premise above. Equivalently, the event sector requires \(\lambda_{ab}(X)c_{ab}(X)=0\) at each state.
Then \(\lambda_{ab}\equiv0\) on \(U\). Indeed, the admissible coincidence set has empty interior: perturbing \(x_a\) at a coincident state makes \(c_{ab}>0\). If the intensity were positive at that state, continuity would keep it positive in a neighborhood containing such an inadmissible perturbation, a contradiction. Away from coincidence passivity already forces the intensity to vanish. The same argument applies to each member of a finite family of candidate additions.
This statement concerns a continuous intensity for discrete unit-edge jumps. It is not a theorem about continuously changing conductance, a state-reset process, stochastic expected-budget cancellation, a discontinuous guard, or a domain restricted in advance to coincident endpoints. It does not forbid an event on a discrete coincidence guard; it shows that a nontrivial such guard is additional structure, not a continuous extension selected by the existing flow. In particular, the weaker expected-generator condition \(-D+\sum_{ab}\lambda_{ab}c_{ab}\leq0\) could offset positive-cost jumps with continuous loss. It is a different stochastic premise, not the pathwise event passivity assumed here.
Exact nonselection at an admitted zero-cost event¶
Keep the supplied ordered candidate \((0,5)\), clock origin and the entire C5 preparation of section 8, with \(\epsilon=1/256\) and unit capacities. Two declarations satisfy the same nodal laws and passive-event requirement:
- Execute no attachment and continue the separately admitted components.
- Execute the supplied bridge once at \(t=0\), without a nodal reset, and continue the admitted joined model.
Both have event jump zero. Both admit a local continuous continuation by smoothness in their respective strictly acute domains. Yet their donor form and phase rates immediately differ by \(-\epsilon/12\) and \(+\epsilon/(2\pi)\), respectively. This is the existing static witness, not a repeated transmission experiment. Specifying the same candidate in both cases prevents a port-selection ambiguity from hiding the distinct occurrence decision.
The two declarations are compatible conditional interventions, not two derived autonomous laws. They prove that the flow, storage and event passivity do not require the admissible attachment. Neither the event time nor an occurrence rate is selected by that accounting.
Deletion and atomic exchange are different accounting cases¶
Pure deletion has \(\Delta S=-\sum_R c_{ab}\leq0\), provided the resulting support is still admitted. Energy decrease alone does not ensure connected support, retain a named cycle or preserve a pattern's identity. A deleted cycle edge can remove the very cycle on which winding was defined without any continuous phase crossing.
A simultaneous addition/removal event is passive exactly when \(\sum_A c_{ab}\leq\sum_R c_{ab}\). For an exact scalar example, take the path \(0-1-2-3\), phase zero, unit capacities and \(x=(0,2,1,1)\). Remove \((0,1)\) and add \((0,2)\). Both supports are connected simple unit graphs with acute phases. The removed edge costs \(2\), the added edge costs \(1/2\), and hence \(\Delta S=-3/2\), despite unequal form at the new endpoints. The strict inequality persists on a sufficiently small open neighborhood of this state; the coincidence obstruction for pure addition does not apply.
This is a single atomic exchange budget. If deletion and addition were separate events, each required to be passive, the positive-cost addition would still fail. Edge-count preservation, candidate selection, identity preservation and event timing are independent premises. No exchange law or active campaign follows from this accounting example.
Shared represented accounting and static controls¶
The existing attachment report now exposes continuous_loss_change and
represented_zero_supply_passive as derived properties, without storing a
second energy formula. assess_supply(supplied_work) compares signed
caller-declared work with its captured storage_change. The returned exact
rational margin and balance flag concern the represented snapshot, not an
ideal trigonometric certificate or authentication of an available resource.
Continuous loss is a rate and is never credited as event work. The common
SDK exporter retains this assessment; no operation modifies a live graph.
The static controls check an independently known equal-phase bridge cost, exact budget boundaries including sub-binary64 rational margins, the retained zero-cost C5 discriminator, and the P4 atomic-exchange witness through the native field and shared transport reset. They do not implement an exchange selector or rerun a trajectory. The API contract owns represented-input and export details. The exact-real jump, passivity and continuous-intensity conclusions above are analytic results under their stated premises; a test of finite arithmetic is not their proof.
Disposition: the inherited storage identity, the additional passive-event restriction and the unselected occurrence decision are now distinct. Nontrivial support dynamics requires an explicitly admitted continuation of this state/event card; the execution plan owns that choice. These results introduce no selector, reservoir or runtime support mutation.
10. Passive bridge relocation with a shared recovery domain¶
Atomic event and signed local interface¶
Keep the two supplied ordered cycles \((0,1,2,3,4)\) and \((5,6,7,8,9)\), their internal unit edges, full nodal state and relative frames. Replace the single bridge \((0,5)\) by \((1,6)\) in one event. The pre-event and post-event graphs are connected simple unit graphs; no intermediate disconnected flow is part of this event. Holding edge count, preserving these cycles and supplying this candidate are additional event premises, not consequences of the nodal identity.
The interface extends section 8 without a second pressure law. For any supplied atomic replacement, let \(\sigma_{ij}=1\) on added edges, \(-1\) on removed edges, and zero elsewhere. From the pre-event full graph's port quantities, the exact ideal updates are
They also cover shared endpoints, where degree changes can cancel while gradient and resultant changes do not. Substituting these quantities into the same \(g,H,p,\dot x,\dot\theta\) rows gives the post-event field. Only nodes incident to a changed edge can change their ideal local rows. Both graph/state domains still require independent admission. In particular, an edge-cost inequality alone cannot admit a nonacute proposed edge.
The exact storage jump is
All internal cycle edges and nodal phases are retained, so both snapshot windings are unchanged. Future identity requires the continuous recovery argument below; it does not follow from that snapshot observation alone.
One equilibrium and one explicit bound for both supports¶
Set \(\kappa=2\pi/5\), \(c=\cos\kappa\), and \(\theta_{*,k}=\theta_{*,k+5}=k\kappa\) for \(k=0,\ldots,4\). Both possible bridge gaps vanish at this phase state. The cycle sine sums cancel, so uniform form and \(\theta_*\) are an equilibrium for either complete relational law. Their common equilibrium storage is
Assume \(e,w,\beta>0\) and strictly positive held capacities at every node. Equal capacities are unnecessary. Apply the existing local recovery theorem and explicit barrier separately to the two graphs, with no change to its flow or storage premises. Choose consistent local phase lifts and put
Both graphs have ten nodes and diameter five. The same pathwise Cauchy bound used in the interaction proof gives \(\lambda_2(B)\geq2/45\). In the radius-\(r\) ball, every reference edge gap changes by less than \(\sqrt2r=\pi/20\); hence all edges in either graph remain acute and the common Hessian lower bound is \(c_r=\sin(\pi/20)>1/10\). The strict sine inequality follows from concavity above the chord on \([0,\pi/2]\).
For either graph the theorem's barrier consequently satisfies
Write \(E_{\rm rel}^\pm=S_{E^\pm}(X)-S_*\). The following strict conditions therefore define a sufficient open domain of full nodal states:
Because the equilibrium storage is the same on both supports, \(E_{\rm rel}^+=E_{\rm rel}^-+\Delta S<E_{\rm rel}^-\). The event preserves \(z\), so both pre-event and post-event preparations meet the existing basin conditions. Holding the respective support fixed, each continuous solution remains acute and converges to the same phase shape and uniform form modulo its limiting common offsets. Those offsets need not agree between the two evolutions. Internal winding one persists. No repeated-event stability or universal decay time is established.
Exact strict preparation and independent controls¶
Take \(x=\epsilon e_0\) with the phase state \(\theta_*\) above and \(\epsilon\ne0\). The two original internal edges incident to node 0 contribute \(\epsilon^2\) in total. The old bridge contributes \(\epsilon^2/2\); the new bridge contributes zero. Therefore
All ideal phase sources vanish at this preparation. Put \(h=1+2c\). The changed port data, inserted into the unchanged rate formulas of section 8, are
| Node | Before \((d,q,H)\) | After \((d,q,H)\) |
|---|---|---|
| 0 | \((3,3\epsilon,\pi h)\) | \((2,2\epsilon,2\pi c)\) |
| 5 | \((3,-\epsilon,\pi h)\) | \((2,0,2\pi c)\) |
| 1 | \((2,-\epsilon,2\pi c)\) | \((3,-\epsilon,\pi h)\) |
| 6 | \((2,0,2\pi c)\) | \((3,0,\pi h)\) |
Every other ideal row is unchanged. In particular, node 0's form rate is unchanged even though its gradient and phase rate change; event cost alone does not describe the local dynamical response.
The explicit family \(0<\epsilon^2<\mu/6000\) satisfies the common basin conditions: the energy bound is immediate and \(\|z\|^2<3/20000<9/800<r^2\). This proves nonemptiness of the strict open domain without executing a trajectory. It permits arbitrary positive held capacities and positive \(e,w,\beta\), without a uniform convergence-rate claim as those parameters approach excluded boundaries.
For \(\beta=1\), the fixed choice \(\epsilon=1/256\) gives
Two controls separate the obligations:
- Reverse exchange at the same state: replacing \((1,6)\) by \((0,5)\) costs \(+\epsilon^2/2\). Both endpoints still satisfy the same recovery bounds, but this reverse event violates the no-work passive premise. Recovery is not a substitute for an event budget.
- Zero contrast: at \(\epsilon=0\), both graphs are equilibria and either exchange has zero cost. No field response or energy preference selects an occurrence. This is the boundary of the strict example, not evidence of a strictly dissipative event.
Zero capacities remain admissible to the storage accounting but are outside this recovery theorem. At zero form dissipation the existing local theorem also excludes generic recovery; the event itself does not repair that limit.
Static implementation protocol and claim boundary¶
The static comparison is fixed before native-field evaluation: node order
\(0,\ldots,9\), the two ordered C5 cycles, old bridge \((0,5)\), new
bridge \((1,6)\), \(\epsilon=1/256\), \(e=w=1/2\), \(\beta=1\)
and unit capacities. Form is zero except at node 0. Ideal phases are
\(2\pi k/5\); the implementation preparation uses Python
2 * math.pi * k / 5 for both nodes \(k\) and \(k+5\).
Reuse fresh native fields and the shared support-reset owner on detached states. Compare actual storage changes using exact rational arithmetic on represented values. Use absolute tolerance \(10^{-15}\), with zero relative tolerance, for field agreement with the stated ideal formulas. No time integration, random input, fit or support event on a live graph belongs to this comparison.
The shared observe_relational_relocation and SDK relational_relocation
compare the old and new connected fields on detached state. The removed edge
must be a bridge between two nontrivial components, and the replacement joins
those same components while preserving every internal edge. Port updates,
field differences, support-reset accounting and signed supply assessment reuse
the attachment owners. The observer admits either sign of event cost; it does
not turn passivity into a graph mutation or a theorem verdict. See the
execution contract
and static controls.
The ideal phase preparation is not its binary64 materialization. Exact rational accounting of represented storage does not enclose trigonometric error or certify the ideal recovery inequalities for arbitrary floating inputs. Preserve rate/pressure defects and report this static comparison separately from the exact-real theorem. The two-bridge capture API does not certify either single-bridge graph.
Disposition: a supplied passive relocation can retain both prepared identities and enter an explicit continuous recovery domain on an open set. It reorganizes existing support; it does not create the substrate or select whether, when or which relocation occurs. The execution plan owns the constitutive decision that remains; no autonomous event law is introduced here.
11. Support-law closure: restrictions do not select occurrence¶
State, candidates and the missing row¶
Retain X=(x,theta,nu), the actual unit support E, held positive capacities,
and the complete relational field F_E. Fix a finite supplied family of
atomic bridge exchanges a=(removed edge, added edge). Each reset is
R_a(E,X)=(E_a,X): it changes support and retains every nodal coordinate.
Admit the old and new fields separately. If recovery is claimed, require a
valid common-basin argument such as section 10, not merely unchanged winding
or a favorable represented storage jump.
The continuous law and the reset map leave distinct data unspecified:
| Item | Restriction inherited from the present premises | Remaining choice |
|---|---|---|
| Candidate family | Must transform with the full state and preserve the declared internal support | Which exchanges are eligible is itself a premise |
| Event budget | Additional zero-supply passivity requires Delta S_a<=0 |
A nonpositive cost does not require execution or rank two eligible events |
| Identity | The stated pre/post recovery conditions must both hold | Recovery does not choose an event or prove repeated-switching stability |
| Selection | An equivariant deterministic choice must be fixed by the state stabilizer | Several fixed choices can remain; ties need not be symmetries |
| Timing | Event rates have inverse-clock units and transform with the complete flow | Waiting law, guard, threshold, randomness or memory remains unspecified |
| History | Include every consumed mark, age, budget or previous event | Such memory is not present merely because a diagnostic report is retained |
Abstention, denoted bottom, is a valid outcome. A candidate set, a selected
candidate, an occurrence and a waiting time are not interchangeable. The
current read-only observers provide the candidate's field/budget comparison;
they implement none of these missing causal choices.
Exact symmetry acts on events, not their list positions¶
Let a supplied finite group act on nodes, the support and every consumed
state/history coordinate. It acts on each event by sending both its
removed and added edges to their images. Candidate eligibility must be
invariant under the stabilizer H of the complete declared state. For a
deterministic equivariant selector s,
Thus only stabilizer-fixed events or abstention are possible. If no eligible event is fixed, a unique deterministic non-abstaining choice is obstructed. A singleton candidate orbit removes this particular obstruction; it does not derive a selector, its continuity, or an event time. Sorting node names, taking the first candidate or hiding a mark in insertion order does not respect this premise.
The existing finite-action owner already proves this restriction. To apply it to edges without a parallel selector, lift the action to typed node and event slots. Retain the complete exact nodal labels, actual support relations and each event's added/removed endpoint incidences. Permute the event slots by the induced edge action, then pass those slots as candidates. This is an exact declaration, not a live graph certificate or an assertion that rounded phases have exact symmetry.
A strict passive witness exists without breaking node-label symmetry. Use
two C5 rings and bridge (0,5), synchronized phase zero, common positive
capacity, and x=epsilon*e_0, epsilon!=0. Restrict candidates structurally
to replacement bridges between neighbors of the two old ports:
{1,4} x {6,9}. Every candidate has
The independent reflections of the two rings fix the complete old state and act transitively on these four events. None is fixed. Hence strict storage preference for relocation does not supply a unique equivariant choice. This synchronized witness is distinct from the winding-one recovery preparation in section 10. With that latter primitive phase held fixed, equal costs need not be generated by any full-state symmetry.
If randomness is independently postulated, invariance forces equal probabilities or intensities within each stabilizer orbit. It does not fix the total event rate, the mass assigned to different orbits, or the probability of abstention. Symmetry therefore does not derive stochastic dynamics either.
The event clock must transform with both continuous rows¶
For a constant clock conversion tau=b*t, b>0, with form units held fixed,
the same mathematical paths are represented by
Here lambda_a is an independently supplied occurrence intensity, when such
a law is chosen. The integrated hazard integral(lambda_a dt) and event
choice probabilities are invariant. Storage and its event jump are unchanged;
continuous storage work and loss are divided by b. Increasing solver dt
without transforming capacity changes the evaluated evolution, not its units.
The engine normalizes its two pressure weights: dividing both by b leaves
the effective weights unchanged and cannot substitute for this conversion.
For d tau/dt=alpha(t)>0, absorbing the clock into capacity instead gives
A nonconstant alpha generally leaves the held-capacity family. A
state-dependent clock has dot(alpha)=L_F alpha; if it depends on support,
an event can also reset the transformed capacity. One cannot suppress those
terms to manufacture a native event clock. The
full-state clock owner
retains the regularity requirements for treating this as a state chart.
For example, dividing all capacities by their mean discards their common
scale unless it is retained separately; it is not an invertible full-state
chart. Held transformed capacity requires alpha to be constant along each
flow segment and preserved at its events, not merely positive.
At a complete equilibrium a time-independent deterministic guard sees the same state forever. It cannot both remain inactive initially and first become active after a positive finite delay without additional time/history input. This excludes that particular state-only waiting mechanism, not every event law. An event-rate postulate can introduce randomness, but the equilibrium does not derive it.
Explicit compatible rival laws¶
Nonuniqueness remains even when a supplied candidate has strict passive slack
and both supports have proved recovery. Work on a common open domain U
where the finitely many supplied candidates satisfy these hypotheses. Follow
the admitted deterministic old-support flow until its first event or first
exit from U. Stop this comparison at exit; after an event retain its new
support and the existing recovery law. This construction needs at most one
event and asserts no arbitrary repeated-switching result.
Define the dimensionless slack and a rate from quantities already present:
These do not select a law. For example, the following are independent logical countermodels, not inferred TNFR rules or proposed engine defaults:
The last two explicitly postulate competing stochastic first-event clocks. On every compact subdomain they have finite continuous rates and positive total rate. Their first-event survival probability along the unchanged pre-event path is
up to the stopping time. The conditional instantaneous event-type weights
are lambda_a/sum_b lambda_b; they are not generally the probabilities
integrated over the entire future path. All three models retain the same
continuous field, passive budget, recovery premise, node relabeling and
common form/phase origins. Two strictly positive models remain different
even if abstention is excluded. Multiplying every rate by a positive
dimensionless constant changes waiting while preserving instantaneous type
weights; holding F_E fixed makes this a different event law, not a common
clock conversion.
The rates also satisfy the selected model's
form-unit covariance.
Under x'=a*x, beta'=a^2*beta, tau=b*t, a,b>0, and normalized coefficients, put
k=e+a*w, e'=e/k, w'=a*w/k, nu'=k*nu/b. Then s'_a=s_a and
r'=r/b. Transform U and its recovery inequalities as well; keeping a
numerical radius in mixed form/phase coordinates unchanged is not a unit
conversion. This covariance restricts admissible formulas without choosing
the exponent or probability law.
A concrete discriminator uses the aligned winding-one phases and
x=epsilon*e_0+(epsilon/2)*e_1, with supplied candidates (1,6) and (2,7)
replacing (0,5). Their exact jumps are -3*epsilon^2/8 and
-epsilon^2/2. Their ideal common equilibrium has zero bridge gaps; the
section 10 basin proof applies to each support for sufficiently small
nonzero epsilon. The two positive rival models predict instantaneous
type ratios 3/4 and 9/16, hence first-type weights 3/7 and 9/25.
These are analytic conditional predictions of different assumed laws,
not measured event frequencies or a reason to install either law.
Integration and disposition¶
The selection controls exercise the existing exact-action owner and real relocation observer; the clock controls compare the complete shared fields and a bounded Euler unit-conversion control. Exact arithmetic on represented storage is kept separate from the ideal phase preparation and its recovery proof. No event selector, random generator, timer, automatic graph mutation or duplicate SDK report is introduced.
Result: full-state availability, event passivity, recovery, symmetry and clock covariance do not uniquely close support evolution. No independently justified additional occurrence premise was obtained in this audit. Retain support changes as supplied interventions. This closes the bounded nonselection question; it does not prove that a future independently justified principle could never determine an event law. Fixed-support pattern formation and interaction remain valid and need no primitive rewiring to exist.
12. Nodal reorganization and connection in one action¶
Revising the event premise, not discarding the nonselection result¶
An NFR's proposed ability to act through operators belongs to the same research question as connection formation. An operator specifies an action on state. Deriving its internal activation additionally requires a map from the acting pattern's state to the target, action and time. Neither an external controller nor a conscious agent is required by that question; the current implementation's invocation rules do not already supply its physical answer.
The earlier pure-addition restriction held every nodal coordinate fixed.
Actual UM changes phase and can change capacity while adding edges. RA, EN
and AL can change form. It is therefore necessary to examine a complete reset
(E,X)->(E_plus,R X), rather than transfer the frozen-triad restriction to
every operator-mediated contact.
The implementation audit reuses existing owners:
| Action | Existing mechanism | Premises still supplied |
|---|---|---|
| UM, Coupling | Snapshot phase proposals and optional capacity alignment; functional links to phase-compatible nonneighbors | Invocation, candidate inventory, phase limit, affinity mixture/threshold, sampling and merge policy |
| RA, Resonance | Form mixing and configured phase/capacity changes on existing compatible neighbors | Invocation, factors, target set and ordering; RA itself creates no edges |
| EN, Reception | Incoming-form blending on its declared execution path | Which incoming data are available and when the operator runs |
| AL, Emission | Form change on an existing node | Source/amplitude and invocation; it does not create the substrate |
EN's actual form input is the unweighted mean of existing incoming neighbors (predecessors on a directed graph), not a U3-filtered or phase-ranked mix. Its source-ranking telemetry does not select or weight that mean. AL and EN write no phase or support on these basic paths. Operator-class callbacks and complete words retain their separate execution contracts.
The UM kernel searches
all graph nodes or the supplied _node_sample for nonneighbors. That inventory
is a potential-contact relation, not a relation created by the search itself.
UM first needs an existing compatible neighbor at its target; it can join
nontrivial components or attach an isolate, but does not bootstrap two
isolates. RA uses the shared neighbor stage.
The existing child/coupling feedback and
THOL transport already retain supplied
targets and schedules; they are useful mechanisms, not forgotten autonomous
selection theorems.
In particular, UM's functional-link score mixes phase affinity, normalized absolute-form similarity and Si similarity. Its form term changes under a common EPI offset, whereas the selected continuous relational law is offset invariant. Limited proximity sampling also uses snapshot rank for ties. These configured policies cannot be imported as a derived relational law without revising and testing their premises. They remain separate from the read-only storage accounting below.
Full reset accounting¶
For symmetric nonnegative conductance W and simple undirected support U,
define the same storage functional, with their distinct roles retained:
At unit weights this is the selected relational storage. Reading it on other weights is valid endpoint accounting, not admission of a weighted relational evolution law. In particular a zero-weight support edge still contributes phase storage, consistently with the native unweighted phase neighborhood.
Writing X_minus and X_plus for the actual endpoint states gives the exact
decomposition
The first term is nodal reorganization on the old support; the second is
support work at the new nodal state. This intermediate evaluation is an
algebraic counterfactual, not an asserted execution order or an extra event.
Capacity does not enter this storage explicitly, but any capacity change
must be retained because it changes subsequent rates. A negative first term
can offset a positive second term in this same action. No reservoir of
past dissipation is inferred. Event passivity remains an additional premise,
tested against the full Delta S, and no timing law follows.
A strict UM attachment funded by its own phase reset¶
Take two existing unit edges (0,1) and (2,3), constant form x_i=m>0,
equal positive capacities and equal Si. Set
Invoke one bidirectional UM stage at node 1 with phase factor
0<eta<=1, and explicitly supply node 2 as its only candidate nonneighbor.
The compatible old pair has circular mean h, so the ideal phase proposal is
The proposed bridge (1,2) has positive phase cost
beta*(1-cos(eta*h)). Uniform form makes its transport cost zero for any
nonnegative functional-link weight. The full change is nevertheless
Indeed f(0)=0, f(1)=cos(2h)-cos(h)<0, and
Convexity gives f(eta)<=eta*f(1)<0. This is a conditional finite-action
theorem for the declared one-candidate UM reset. It proves neither universal
UM passivity nor its autonomous invocation, and these two-node components
are a minimal mechanism witness, not a demonstrated maintained NFR identity.
The concrete production control uses h=pi/8, eta=1/4, m=1/2, capacities
one and Si 0.8. All phase gates and the compatibility threshold have strict
slack. The ideal new weight is 63/64, and the new gap is pi/32.
The strictly negative budget therefore also persists under sufficiently small
admitted perturbations of the inputs with this candidate policy fixed.
The theorem concerns exact-real phases; the test separately reads the actual
binary64 stage endpoints and their represented storage.
The generated weight is not one. The current unit-support relational executor consequently rejects that output. Do not silently replace its weight or reuse the unit-bridge recovery theorem as if this operator event were the same model. The result establishes feasible contemporaneous reorganization and attachment, not subsequent pattern maintenance.
RA form redistribution can also offset an attachment cost¶
On the same two edges, take forms (m+d,m-d,m-d,m+d), m>d>0, uniform
phase and equal positive capacities. One all-target RA stage with unclipped
form-mix factor rho changes each pair's contrast from d to
c*d, where c=1-2*rho. A subsequent admitted UM stage at node 0, with only
candidate 2, can add their positive-conductance bridge. Calling its actual
conductance omega, uniform phase gives
For rho=1/4 and 0<omega<=1, this is strictly negative, while the new
edge cost 2*omega*c^2*d^2 is positive. The production control uses m=1/2,
d=1/8; RA produces forms (9/16,7/16,7/16,9/16). It retains the actual
capacity amplification and UM conductance. The combined two-event budget is
telescoping endpoint accounting: RA's reduction and UM's positive increment
must also be reported separately. This is not zero-supply passivity of each
individual event or a claim that prior dissipation is a stored work reserve.
An internally justified composite action would still need its own definition.
The continuous law can create phase admission¶
At held unit support, phase and capacity, an AL/EN form jump d changes the
next freshly evaluated relational phase row by an exact linear identity:
For one unclipped EN target i with mix 0<=rho<=1, its existing neighbor
mean gives d_i=-rho*q_i/degree_i and all other entries zero. Consequently
This redistributes existing contrast; uniform form remains uniform. A
targeted AL increment a from uniform form instead costs
degree_i*a^2/2 on nonisolated support. Such an increment can provide a phase
response, but its supplied action and budget must remain explicit. The
identities do not transfer unchanged to simultaneous multi-target EN,
clipping, a capacity/phase reset or arbitrary operator words.
For a supplied candidate pair i,j in independently admitted components,
retain a common phase reference and a declared native U3 limit
0<gamma<=pi/2. On a
regular lift, set delta=theta_j-theta_i and
M=cos(delta)-cos(gamma). The existing phase row gives
Positive M is phase compatibility for this limit. It is not an edge or an
instruction to invoke UM. Consider the two separate edges (0,1) and (2,3)
with phases (0,0,gamma,gamma), unit capacities and forms (m+a,m,m,m).
For a>0, H_i=pi, q=(a,-a,0,0), so
Smoothness supplies a transverse crossing and nearby preparations with the
second component rotated to gamma+epsilon that start incompatible and enter
compatibility in finite positive time for sufficiently small epsilon>0.
The implicit-function argument also gives
t_cross(epsilon)=beta*pi*epsilon/(w*a)+O(epsilon^2).
At a=0 each component is an equilibrium and remains incompatible after that
rotation. Reversing a reverses the initial crossing direction. Thus this
admission timing is inherited from the complete continuous law; no new
phase-speed equation is imposed.
Both components can rotate independently without changing their internal
dynamics. Their cross-component phase difference therefore requires the
supplied common reference; it is not reconstructible from independently
phase-quotiented observations. Moreover the unit bridge at the displayed
boundary would cost a^2/2+beta*(1-cos(gamma))>0. Phase admission alone still
does not make a frozen-state attachment passive or cause its occurrence.
Integration and the all-operator audit¶
The shared reset observer
and Network.relational_reset(after, storage_scale=...) retain both supplied
endpoints, actual conductances and phases, plus the shared transport snapshot's
capacity and stored pressure (zero defaults when those attributes are absent).
Those defaults are not evidence of measured zero capacity or pressure. They
reuse the transport reset and represented half-sine phase cost to separate
form/phase changes into state and support contributions. Exact rational
accounting of represented values is not an enclosure of ideal trigonometric
error, authentication of an operator event, or admission of future flow.
The API contract
owns the wider snapshot domain and supplied-work assessment.
Actual UM/RA controls, AL/EN and phase-admission controls and SDK/export controls reuse these owners. Native AL/EN serialized uniform-real BEPI passes shared signed-scalar admission directly; no test-only state conversion substitutes for integration. Rich, complex and unrepresentable scalar inputs still reject.
The all-thirteen mechanism map adds IL phase relaxation, capacity-only versus edge-aware VAL/NUL resets, stored-pressure lifetime, THOL's isolated birth and the three REMESH paths. It distinguishes implemented writes, eligibility, configured dispatch and autonomous occurrence. In particular UM can move phase at uniform form where the relational phase velocity is zero; its reset is not automatically a continuous step of the selected law.
Sustained rhythm synchronization is a candidate activation premise, distinct
from instantaneous U3 compatibility. The fields already give
delta_dot=theta_dot_j-theta_dot_i. Equal rates once do not prove persistence;
equal sustained rates can retain a noncompatible offset. A locking hypothesis
needs a declared time interval or an invariant dynamical condition, a candidate
relation/common reference and a separate reason why locking causes attachment.
Internal precontact evolution above avoids using the future bridge to explain
its own admission; it does not establish the remaining occurrence principle.
What this opens, and what remains to derive¶
The two results are complementary: existing form/phase dynamics can create admission, and an actual nodal action can make a positive-cost connection compatible with total nonincrease. They are not yet one autonomous trajectory: their preparations, candidate access and execution contracts differ.
Continuous conductance is another possible model revision, but the native phase channel uses support independently of weight. A missing edge is not the limit of a present edge with weight approaching zero for that channel. Weighting phase too would change the constitutive model. The current operator reset route can be studied before introducing that separate extension.
An open mechanism is a justified internal activation/contact rule for a complete action: candidate access and common reference, state reset, occurrence and clock, and the post-action evolution domain. Emission or Reception may supply the form contrast used by the phase mechanism, but their input and work must then be included in the same account. A named operator does not by itself justify its source or timing. The sole plan defers that primitive-event question while prioritizing collective geometry below; no functional-link policy is relabeled as emergent.
13. Precontact rhythm, phase agreement and sustained locking¶
The hypothesis that a connection is caused by synchronized rhythms first
requires an identified rhythm of the actual nodal law. The
native pulse admission
reuses the existing phase/form modes and storage balance. A capacity is not
an angular velocity; a maintained phase pattern need not be a periodic orbit.
The auxiliary Network.rhythm() spectrum and arithmetic pulse studies have
separate laws. No additional primitive pulse variable follows from their names.
Matching phase and speed can still hide different futures¶
Fix a candidate port in each independently admitted P2 component. Write the
relative port phase as c=theta_b0-theta_a0, retaining a common reference.
The complete fields give c_dot=theta_dot_b0-theta_dot_a0. Instantaneous U3
compatibility bounds abs(wrap(c)); equality of phase speeds gives c_dot=0
at that instant. Neither assertion says that this equality is invariant.
An explicit counterexample needs no forcing or new parameter. Prepare both
pairs with uniform form m and common positive capacity nu. Their phases
are (0,a) and (0,-a), with 0<a<pi/2. Both candidate ports therefore have
the same form, phase, capacity and instantaneous phase velocity (zero).
Each pair has q=0, metric H=pi*sinc(a) and nonzero opposite form rates.
Differentiating the existing full field gives
Indeed each port has theta_dot=w*nu*u/(beta*H), where u=x_0-x_1.
Initially u=0, so the differentiated metric term vanishes, while
u_dot=2*w*nu*delta/pi. The acceleration comes from retained internal
phase geometry through form evolution, not a hidden external frequency.
The damping term also vanishes at this instant; the result holds for e>=0.
The two pairs even have the same consensus tangent spectrum. Matching that
spectrum or a port's current phase/speed cannot certify sustained locking.
Exact prepared locking does not select attachment¶
Conversely, let two supplied components be isomorphic with corresponding
capacities and model coefficients. Prepare their full forms/phases related
by the same node correspondence and constant offsets b,c:
Relabeling and common-offset equivariance, followed by uniqueness on a shared regular domain, preserve these relations for as long as both solutions exist there. Their corresponding phase velocities agree at every time. The components have no cross-edge, however; the existing law keeps their supports unchanged. This is inherited matching from a supplied preparation, not mutual entrainment across a missing connection or evidence that locking causes birth.
Any offset c is allowed by this independent-component symmetry. Thus sustained
equal rates can coexist with a port separation outside the selected U3 limit.
Even c=0 supplies no event law. The relative offset is a neutral preparation
freedom: replacing component B by a further constant phase rotation produces
another exact solution. Independent evolution cannot select a unique relative
phase for all those rotated preparations. Adding a contact rule which compares
the components introduces a potential-contact relation and a common reference
that must be declared and justified.
Existing fine support aligns ports while deforming regional geometry¶
An NFR understood as a coherent region need not have the same relations as an individual fine node. Distinguish the birth of a primitive graph edge from an effective interaction between regions on supplied fine support. The existing directed-triangle derived-form phase reduction already illustrates the latter under a different law and observation. Its contrast angle cannot be silently substituted for primitive relational phase.
There is also a direct causal control in the current joint law, without an
invoked UM event. Reuse the two unit C5 rings, common capacity nu>0, uniform
form, winding-one internal twist kappa=2*pi/5, and the supplied fine bridge
(0,5). Rotate the second ring by a small positive offset c. At that initial
state only the two ports have an ideal nonzero phase source. Define
Since q=0, all primitive phase velocities initially vanish. Nevertheless
x_dot_0=w*nu*g_0=-x_dot_5, so the full-support Laplacian gives
q_dot_0=4*w*nu*g_0=-q_dot_5. Differentiating the admitted phase row yields
Here c initially equals the port phase difference; once other nodes respond,
one must retain their full state rather than assume a closed rigid-ring angle.
For 0<c<pi/2, g_0>0 and all initial edges are acute. Removing the fine
bridge while keeping both prepared rings unchanged makes each a twist
equilibrium, so this acceleration vanishes. The dependency is thus
existing relation -> phase pressure -> form contrast -> phase response.
It supplies an interaction-induced port response, not a connection caused
by an already assumed alignment. Crucially, it is not instantaneous alignment
of the entire regions. Each internal neighbor of port 0 has initial phase
acceleration -w^2*nu^2*g_0/(beta*H_n), with H_n=2*pi*cos(kappa), and
the corresponding neighbors of port 5 have the opposite sign. Hence the
difference of the two regional mean phases satisfies
for sufficiently small positive c on C5. At zero offset,
H_p=pi*(1+2*cos(kappa))>2*H_n because cos(2*pi/5)<1/2.
Thus the ports initially approach while the regional means initially separate.
This is an explicit geometry-dependent deformation, not a contradiction or
evidence that one scalar phase per region closes the future response.
For positive dissipation and sufficiently small offset, the existing local recovery theorem also gives convergence to the common aligned-twist equilibrium modulo common offsets. This reuses its local theorem, not a new quantified basin or a global synchronization result. The bridge already belongs to the fine model; interpreting a resulting regional relation as an emergent effective connection must state that distinction. It does not derive the underlying graph's birth.
A candidate composite identity, with full internal geometry retained¶
This suggests a more precise ontological target than primitive edge creation: two coherent regions may acquire a joint geometric identity on supplied fine support. The existing interaction theorem already establishes the relevant local distinction. Disconnected acute C5 twists have four independent neutral offsets (form and phase per component) and sixteen stable tangent directions. The joined graph has only two global neutral offsets and eighteen stable tangent directions under the positive- capacity/dissipation hypotheses. Relative offsets are restored by the joint dynamics rather than remaining arbitrary independent choices.
At aligned ports the bridge adds quadratic stiffness
[(u_0-u_5)^2+beta*(v_0-v_5)^2]/2 to perturbations of form and phase.
Degrees and phase metrics change too. The two lost neutral freedoms therefore
do not define an isolated two-dimensional oscillator; they mix with internal
deformation. The existing tangent and nonlinear closure results
already prevent replacing this state by just two rigid regional clocks.
The full joined equilibrium geometry, modulo common form/phase origins, is a conditional candidate composite identity: it has a restoring response to sufficiently small relative perturbations under the same law. Calling it a closed autonomous coarse NFR additionally requires a sufficient retained state and justified constitutive reduction; the known hidden-state memory cannot be discarded. A sustained composite pulse is also a separate claim, subject to the dissipative/reversible distinction above. Supplied fine support and prepared component patterns remain explicit premises.
The smallest useful discriminator observes internal shape. For the receiver
ring define chi_x=x_5-mean(x_6,...,x_9). The phase-offset preparation gives
chi_x_dot(0)=-w*nu*g_0<0. Independent components or a rigid-region
approximation give zero. This is already a static analytic distinction; a
new run is justified only to test a separately frozen quantitative finite-time
prediction, not to rediscover transmission or claim physical constituents.
Execution evidence and research consequence¶
The contact controls
differentiate the actual native field for the acceleration counterexample and
advance the shared Euler owner for short prepared matching controls. The former
is a static directional-derivative check; the latter verifies finite execution,
not exact sustained locking or continuous error bounds. The ideal result is
the equivariance/uniqueness argument above. A separate static two-C5 control
checks both signs of the bridge-induced response and its removed-bridge null
case. It does not rerun the completed transmission or recovery campaigns.
No new oscillator, selector,
observation window or phase law is installed. Full field.phase_rate, internal
state and the existing port/work observations supply the needed information.
The collective-identity question retains independent-component and rigid-region controls. Primitive support birth remains unresolved and has its own connection-mechanism admission; collective organization on finer support does not settle that question. A threshold, dwell time or phase-slip count can define a configured observation, but is not derived merely by naming synchronization. The sole queue owns the next quantitative admission without reopening completed modal, transmission or memory calculations.