Support run 78c499b4-bc7d-4e1a-9efa-47571547b826, 2026-08-24. Serving run 872d7dce (hunt r_c7f779), which asked one bounded literature question and is still alive to receive the answer.
The question this directory answers
Two parts, both about published work rather than about a computation.
- Has a Cohn–Elkies / Delsarte linear-programming certificate been applied to a one-dimensional pair sum
sum_{p != q} f(t_p - t_q)whosefis "positive part of a band-limited function minus a band-limited nonnegative square", and is it known whether the LP bound is tight against the periodic lattice optimum in dimension 1? The load-bearing sub-question: is there published technology, or a published obstruction, for LP certificates against a non-band-limited target of that positive-part shape? - Is there a published theorem that settles, in dimension 1, whether the LP value equals the lattice value for such an
f, the "LP is sharp iff a dual optimiser is supported on the lattice's distance set" line?
Scope
This directory only, plus case-log entry Hunt #114 in hunts/README.md. Nothing in zeta/, ontology/, harness/, lean/, meta/, hunts/frontier_math/, hunts/r_c7f779/, or any root markdown file.
The brief named hunts/r_c7f779/arm_e/PRIOR-ART.md as the delivery path and then forbade writing to any other hunt directory. The scope rule wins: the prior-art page is RESULTS.md here, and the parent can copy it.
This is a literature answer, not a computation. No mathematics is claimed, no number is measured, and nothing here bears on RH (docs/08). The reserved word is not used.