Tiling almost-squares with smaller distinct almost-squares
An almost-square is a k × (k+1) rectangle. Erich Friedman asked (Math Magic, Problem of the Month, May 2012) for which n the n × (n+1) almost-square can be cut into almost-squares of distinct sizes, all smaller than n, and conjectured that every n ≥ 20 works. We prove that this is possible exactly when n ∈ {4, 10, 12, 14, 15, 18} or n ≥ 20. The construction scales a perfect squared square and moves each cut line by an integer so that every square becomes an almost-square; one counting identity proves it correct and gives a parity lemma explaining why the squared square of side 112 reaches only even n and the one of side 110 only odd n. Explicit tilings cover the remaining n from 20 to 248, and an exhaustive search proves that the other n ≤ 19 are impossible, which earlier searches had suggested but not established. The theorem is formally verified in Lean 4 (the cases n = 16, 17, 19 by native_decide, also settled by two independent searches), and the certificate data are rechecked by an independent program. This record is identical to the folder almost-square-tilings at tag almost-square-tilings-v4 of https://github.com/jvvk/mathematics. Version 4 of the preprint; versions 1 to 3 (8 to 23 October 2026) were released on GitHub only.
Authors
- Vamshi Jandhyala (ORCID: https://orcid.org/0009-0006-6708-5805)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-19
- DOI
- https://doi.org/10.5281/zenodo.24375062
- Primary Topic
- Mathematics and Applications
- Type
- preprint