m5 yay

Prompt

RESEARCH QUESTION (plane topology; adversarial audit of Lemma TWO-SIDE (prime ends, dense channels)). Self-contained. Label every claim PROVED (you prove it), SOURCE (exact reference; unchecked?), LAB (from this prompt, status as given), CONJECTURE, OPEN, or RESTATEMENT (a gap reworded). If a definition is ambiguous, state your reading in one line and continue. Say where your argument stops. A. DEFINITIONS Setting (compressed from the lab notes D2595, D2596, D2599; DERIVED-U there, unaudited here; lines marked (INFERRED) are my bridging). - Zhat := the profinite integers; D: Zhat \ {0} -> E_amb is continuous and even with D(2N) = sigma D(N); E := the cluster set of D at 0 (a Cantor set, the fibre labels, E = X_theta). Fibre points p_m(f) (m in Z, f in E), fibres P_m := {p_m(f) : f in E}, K := the closure of the regular points g(z) = (z, (D(z-j))_j), X := K x {0,1}. The host Y = Y_D := ([0,1/2] x K)/folds is a planar continuum with q: [0,1/2] x K -> Y the quotient map; the fence F joins the two ends q({0} x {x}), q({1/2} x {x}) of the arc q([0,1/2] x {x}) (the chord of x). Dynamics H[t,x] = [2t,2x] (no periodic point), h := H^-1, m_k := min_Y d(y, h^k y) > 0. Ray l_e(T) = q(T, 0^e), with H l_e(T) = l_{sigma e}(2T); U := C \ Y is the complement. - 2PAGE := a fixed embedding P: X -> S^1 in which F and the fold are noncrossing matchings (D1535); a stable cut order <_* on E from depth M_s: the fibre orders of all far fibres are <_* or its reverse. For |n| >= M_0 (far), Hull(P_n^0) is the smallest closed arc of S^1 containing P(P_n x {0}) (length < 1/16), H^0(n) its preimage, K_n := {x in K : (x,0) in H^0(n)}, and diam K_n -> 0. The band n is B_n := q([0,1/2] x K_n); a slot of the band is an open region of U between two consecutive chords of B_n (INFERRED from (E1)-(E2) below). - Folded coordinates: for real u, eta(u) := frac u if frac u <= 1/2, else 1 - frac u; A(u) := floor u if frac u < 1/2, else -1 - floor u. A real t is generic if t is not in Z/2 (so eta(t) is in (0,1/2)); then the slice of Y at depth t is C(t) = q({eta(t)} x K_A(t)). - A jump J = (e', e) is a pair of consecutive labels of <_*; J_n is the open arc of S^1 between the walls (p_n(e'),0) and (p_n(e),0) inside Hull(P_n^0); the content of J at n is X cap J_n. **(EMB) Plane model (stated convention; DERIVED-U from the 2PAGE disc realisation of [T3], as used by rig_thick's line law).** X sits on a circle, fence chords are straight segments inside the disc, and fold chords lie outside and are pinched (Moore). For |n| >= M: - (E1) the region R_n bounded by the two extreme chords of the band and the arcs Hull(P_n^0), Hull(P_n^1) satisfies R_n cap Y = B_n := q([0,1/2] x K_n); - (E2) U cap int R_n is the union of the open slots between consecutive chords of B_n; - (E3) for points y_1, y_2 of B_n at one height eta in (0,1/2) there is a level arc gamma in R_n from y_1 to y_2 with gamma cap Y = {q(eta, x) : x in K_n between x_1 and x_2}, crossing each such chord once; - (E4) level arcs in R_n have diameter -> 0 as |n| -> infinity. A4 Crosscut: an open arc q in U whose closure is an arc with distinct ends in X. Sh(q) = bounded component of R^2 \ (X u q), Out(q) = the unbounded one. diam q = diam cl q. A5 (BFMOT 3.6) phi: Riemann map, outside of unit disk onto U u {inf}; ray R_t = phi({r e^(2 pi i t): r > 1}). A chain: disjoint crosscuts Q_i, diam -> 0, Q_j in Sh(Q_i) for j > i; equivalent chains form a prime end E_t. Pr(E_t) = {z: a chain defining E_t converges to z}. The (X,x)-end of R_t (x on R_t): the bounded component of R_t \ {x}; R_t crosses Q essentially iff some (X,x)-end lies in Sh(Q). x in X is accessible iff some arc in U u {x} ends at x. **Crosscut tree (RECALL; Pommerenke ch. 2, standard).** U = C \ Y is connected and U cup {inf} is simply connected (Y is a nonseparating continuum). Every crosscut c splits U into two components; D(c) is the bounded one (the pocket). - Ancestors: c' is an ancestor of c if c subset D(c'). The ancestors of c are linearly ordered by pocket inclusion (two disjoint crosscuts whose pockets share a point have nested pockets). - Spine crosscuts are those whose pocket contains a tail of the outchannel chain. Every ancestor of a spine crosscut is a spine crosscut. - Impression: Im(E_t) := the intersection over a defining chain of the closures of Sh(Q_i); Pr(E_t) is contained in Im(E_t) (standard; BFMOT 3.6). A prime end E_t with Pr(E_t) = Y = dY is a dense channel (BFMOT, quoted below). E(h^-k) := the outchannel of h^-k, i.e. the unique geometric outchannel which BFMOT Theorem 6.2.1 gives for X with the map h^-k (Lemma IB* of D2595 says BFMOT applies to every h^-k); BFMOT := the plane-continuum fixed point paper arXiv 1004.0214 (authors Blokh, Fokkink, Mayer, Oversteegen, Tymchatyn, INFERRED from the initials). **Definition.** Let J = (e', e) be a jump, w in {e', e} a wall label, and n a far fibre. - (J, w) is **wet at n** if (p_n(w), 0) is a limit of points of X cap J_n (content of J accumulates at the wall from inside J). - Otherwise (J, w) is **dry at n**. Then P(p_n(w), 0) is an endpoint of a gap of P(X) lying in J_n. A strip is dry at both walls. - **ALL-DRY** := every (J, w) is dry at every fibre n with |n| >= N_*, where N_* is the far threshold of Lemma CH (A(t) for t >= M_s, 2t >= 3M_*/2). B. LEMMAS (status; where used) **Lemma AC (ancestor diameter bound; DERIVED + RECALL).** For a crosscut x and D > 0 there is d*(x, D) > 0 such that every ancestor c of x with dist(c, x) >= D has diam c >= d*. **Lemma CH (i) (D2595; DERIVED + RECALL; used once, in step 6 of the target).** Assume S* (standing hypotheses of the 2PAGE host above), (EMB) and a stable cut order from depth M_s. Take k >= 1, a generic far t >= M_s with 2^k t >= 3M_*/2 (M_* a far threshold of the lab), and e in E with e and sigma^k e both different from min and max of <_*; put p := l_e(t). Then there is a chain (g_i) defining E(h^-k) such that each g_i is a slot segment at depth t (the part of the level arc at height eta(t) inside one slot) inside a jump J_i of the fibre order, and g_i -> p. C. THE TARGET **Lemma TWO-SIDE (DERIVED + SOURCE + RECALL; P yes / M none (any dense channel) / C side of approach).** Let E_t be a prime end of U with Pr(E_t) = Y (a dense channel). Let n be far, let x in K_n be a point whose chord in band n is accumulated, inside Hull(P_n^0), by chords of X-points on one side A of x, and let p = q(eta, x) with eta in (0, 1/2). Then E_t has a defining chain of crosscuts converging to p, each lying in a slot of band n between two chords of points on side A. In particular, **at a wall that is wet at n, every dense channel approaches p = l_w(t) from inside J, for every generic t with A(t) = n.** SOURCE (papers/lit/bfmot_1004.0214.txt): - l.1665: "Pr(Et) = {z in bdU | for some chain {Qi} defining Et, Qi -> z}" (symbols lost on disk, read from context); - l.1684-1687: "An external ray Rt is said to cross a crosscut Q essentially if and only if there exists x in Rt such that the (T(X), x)-end of Rt is contained in the bounded complementary domain of T(X) cup Q"; - l.1704-1706: "If moreover Pr(Et) = bdU = bdT(X), we say Et is a dense channel"; - l.3184-3186 (Theorem 6.2.1): "there exists a unique geometric outchannel Et for X, which is dense in X". *Proof.* 1. **Approximating points.** Choose distinct X-points y_j on side A with y_j -> (x, 0), and put y^_j := q(eta, y_j). Then y^_j -> p, and y^_j != p. Choose radii r_j > 0 with r_j < (1/3) dist(y^_j, {p} cup {y^_i : i != j}); this is positive because the y^_i accumulate only at p. Take r_j -> 0 and the balls B_j := B(y^_j, r_j) so small that B_j cap Y consists of band-n chords of points on side A ((E1), (E2)). The B_j are pairwise disjoint and miss p. 2. **Chain elements near each y^_j.** y^_j is in Pr(E_t), so some chain defining E_t converges to y^_j. - Take one element c_j of it inside B_j. It lies in one slot of band n. - It joins the slot's two chords. Otherwise its pocket would be a Y-free disc, and the impression would not be Y, against Im supseteq Pr = Y. - R_t crosses c_j essentially. Elements of a defining chain are crossed essentially (RECALL; BFMOT Prop 3.6.5 and the definition of a defining chain). So the end of R_t lies in the pocket D(c_j) for every j. 3. **Nested.** The c_j are pairwise disjoint (disjoint balls), and their pockets pairwise meet (each contains a tail of R_t). Disjoint crosscuts with meeting pockets have nested pockets (RECALL, crosscut tree of D2596 sec.2). So {D(c_j)} is totally ordered by inclusion. 4. **No increasing chain.** Suppose D(c_{a_1}) psubset D(c_{a_2}) psubset ... . Then every c_{a_l} is an ancestor of c_{a_1}. Since c_{a_l} -> p while c_{a_1} subset B_{a_1}, we have dist(c_{a_l}, c_{a_1}) >= r_{a_1} for large l, and diam c_{a_l} <= 2 r_{a_l} -> 0. This contradicts D2596 Lemma AC (ancestors at distance >= D from x have diameter >= d*(x, D) > 0). 5. **The chain at p.** An infinite totally ordered set contains an infinite strictly monotone sequence (RECALL). By step 4 the sequence is decreasing: c_{b_{l+1}} subset D(c_{b_l}). - Closures are disjoint, diam -> 0, and c_{b_l} -> p, so this is a null chain. - R_t crosses every element essentially, so the chain defines E_t (RECALL: a null chain is equivalent to every chain whose prime end's ray crosses all its elements essentially; Pommerenke ch. 2). - Each element lies in a slot on side A. 6. **Slot segments.** Replacing each element by the slot segment of its slot at height eta gives an equivalent chain (D2595 Lemma CH (i), same step). QED *Remark (scope).* The proof uses only Pr(E_t) = Y, the band structure (E1)-(E2), and Lemma AC. It uses no h, no CP1, no variation and no SH+-. It holds for the outchannel of every h^-k and of every h^k, and for every other dense channel (BFMOT l.3180: "X could have other dense channels"). D. ASK Check every numbered step of the TWO-SIDE proof above, including the SOURCE quotes (BFMOT) and the two RECALL claims. For each step: VALID, VALID WITH REPAIR (write it) or BROKEN (give the configuration). E. CONTROLS (all must hold) (C1) The proof claims (Remark at its end) to use only Pr(E_t) = Y, the band structure (E1)-(E2) and Lemma AC; it also cites Lemma CH (i) once (step 6) and the RECALL facts. Name any further input it secretly uses. (C2) A dense channel that is not an outchannel must be covered. (C3) Step 4 (no increasing chain) must be checked when pockets are nested but not strictly. F. DEAD ENDS (not progress) General prime-end theory without pointing to a step. G. FORMAT Line 1 exactly: VERDICT: TWO-SIDE CORRECT | CORRECT WITH REPAIRS (list) | BROKEN (failing step + configuration) | OPEN. Then SUMMARY (<= 100 words), the argument (<= 2,000 words), STILL OPEN EXACTLY. Short of effort: line 1 + summary; an honest OPEN is valid. No finite computation as proof; no theorem beyond its hypotheses; UNKNOWN if unsure a cited result exists. RED FLAGS: trusting the RECALL steps without checking them; a counterexample with Pr(E_t) != Y