Flat strips, counting conventions, and a width bound

Definitions and mathematical statement

The supplied verification identifies this record as commentary after Question 6.2.1 in Section 6.2, "Zero curvature geodesics and flat strips," rather than an independent question. The extracted "Question 6.1.1" prefix and page header are artifacts. The preceding question concerns a closed genus-at-least-two surface with a smooth, or at least $C^k$, $k\geq3$, nonpositively curved metric and a nonclosed geodesic along which Gaussian curvature vanishes identically. This is context supplied from the neighboring record, not a newly assigned problem here.

The commentary reports: with curvature only $C^0$, an example can be constructed whose two geodesic ends are asymptotic to closed geodesics; with curvature $C^2$, a zero-curvature geodesic cannot spiral into a closed geodesic [Ruggiero1998]. It attributes to Wu [Wu2013] a negative answer to the preceding question when $\{K<0\}$ has finitely many components, and finiteness of the free homotopy classes containing zero-curvature closed geodesics under that hypothesis. These are regularity statements about curvature where so worded; they are not silently changed to metric regularity.

A positive-width flat strip is a totally geodesic isometric immersion of $\mathbb R\times I$ with its product metric, where $I$ is an interval of positive length. The source views zero-curvature geodesics heuristically as infinitesimal-width strips. It attributes to the preprint [Cao-Xav] periodic closing of positive-width strips in the $\mathbb R$ direction and finiteness at each fixed lower width $\delta>0$. It mentions a shallow-crossing argument producing a long rectangle of width near $2\delta$. It further reports possible finiteness of all strips, known under Wu's finite-component condition, and an unsuccessful general attempt [Rodriguez Hertz2003]. The counting convention is essential and is not literal counting of every substrip.

The aid gives three quantifier repairs. First, an embedded flat cylinder of width $W>\delta$ contains uncountably many distinct embedded subcylinders of width $\delta$, by moving the transverse interval. Thus literal all-substrip finiteness is false whenever such a cylinder exists.

Second, on the closed nonpositively curved genus-at-least-two surface, choose a primitive unoriented free homotopy class, represented by a deck transformation $\gamma$ of the CAT(0) universal cover. With translation length $\ell(\gamma)$, its minset is \[ \operatorname{Min}(\gamma)= \{x:d(x,\gamma x)=\ell(\gamma)\}. \] The reported canonical maximal-strip convention uses its product decomposition $\mathbb R\times[0,W_\gamma]$, allowing $W_\gamma=0$, with finite maximal width. Conjugacy and inversion do not change this width. The quotient gives a maximal flat-cylinder object for that class when the width is positive; it need not be globally embedded in the surface.

Third, let $C_1,\ldots,C_N$ be embedded flat cylinders with pairwise disjoint interiors and widths $w_i\geq\delta$. Let $\operatorname{sys}(M)>0$ be the shortest length of a noncontractible closed curve. Their circumference lengths satisfy $\ell_i\geq\operatorname{sys}(M)$, so the reported area bound is \[ N\operatorname{sys}(M)\delta\leq \sum_i\ell_iw_i\leq\operatorname{Area}(M),\qquad N\leq\left\lfloor\frac{\operatorname{Area}(M)}{\operatorname{sys}(M)\delta}\right\rfloor. \] The disjointness and embeddedness are essential. This bound does not establish finiteness of all intersecting or immersed maximal strips, and the repairs are not a new solution of the neighboring zero-curvature-geodesic question.

Short English statement

This record is commentary, with an erroneous question prefix, about zero-curvature geodesics and flat strips. The supplied aid clarifies that literal counting of all substrips is infinite, proposes a canonical maximal-width convention by deck-transformation minsets, and bounds disjoint embedded cylinders by area divided by systole and minimum width. The original historical references retain their qualifications.

Sources

• Local source record: catalogue entry, catalogue entry; canonical aim-geometry-notes.json, notes[236], canonical source item 6.1. Title: "Flat strips, counting conventions, and a width bound." The supplied verification corrects its placement to commentary following Question 6.2.1 in Section 6.2.
• Keith Burns and Vladimir Matveev, Open Problems and Questions About Geodesics, AIM workshop "Geodesics," supplied reference https://aimath.org/pastworkshops/geodesicsproblems.pdf. The original remark field is "Remark." The literature field is empty. This draft uses the supplied text without accessing a PDF.
• Supplied source references [Ruggiero1998], Wu [Wu2013], Cao and Xavier's preprint [Cao-Xav], and [Rodriguez Hertz2003]. No full bibliographic entries are supplied. The extracted page number and author header are not mathematical content.
• Registry dated 14 August 2026: status "open," classification "OPEN-TRIAGE," statement status "exact." Attempt 1 is marked context only and not a problem, rigor "proved," novelty confidence low; its candidate contribution is the three-part counting repair. The machine-generated, non-peer-reviewed aid requires independent verification before citation.