Square Packing Atlas
where everything on this site comes from

Sources

The packings are their finders' (most from Ellsworth's catalogue) and the classical proofs are their authors'. Original here: the interactive tools, the contact and rigidity analysis, and the 2026 computer-checked results in §5 marked "this project". Preprints are unrefereed; this page says which is which.

1 · The catalogue this site is built on

The record packings, every n ≤ 324, are drawn from our own certified copies (search/regularize/lists): for each n the best known packing, solved exactly at its side and written as a rational certificate that two separately written exact checkers accept. Each is regularized for display: squares whose angle the packing does not fix are turned to join an existing rotation group, loose squares are placed tidily, symmetric where the packing allows it, with faces made flush and the orientation of the source drawing kept where possible. It stays the same packing as the record, reachable from the source drawing without enlarging the box. This copy is the one shown everywhere the packing appears, including in the history after it has been beaten (Compare, the leaderboard); the drawing it was made from (the catalogue's SVG, or the coordinates as posted) stays in the repository and is linked from Explore's source line.

Everything else (older records, the catalogue's alternatives, n above 324) is parsed from David Ellsworth's SVG files, at the precision they give: exact constants for most, decimals for the ~60 packings that were optimised numerically (Explore marks those as having numeric contacts). Dates and "found by" attributions for those, and for the records still in his catalogue, are quoted from his wording, not independently re-derived.

Not every record is in his catalogue. Since mid-September 2026 smaller packings have been posted for 76 counts up to n = 308: Francisco Couzo's (25 records now; franciscouzo/square-packing from 27 Sept, and registration requests jlevy/squares#451, #476, #488, #583, #607), Nate Chaoweeraprasit's (itsnaka's) SQUISH packings (18; #401, #481, #500), Ryan Xu's (ry-xu's; 9; #432), Daniil Ostashev's (8; #528, #562, #575), Mishapolk's (3; #470), Joost de Winter's (3; #668), PingYou Ltd's (2; #503, #504), and ours (8: 132, 146, 154, 155, 236, 266, 271, 308; §5). Each record is credited to whoever posted that packing first, with the date of the post; Explore lists later posts of the same packing too. A packing found by starting from someone else's is credited to its poster, "based on" the earlier one, unless the two are the same local minimum (Leaderboard, "How credit is counted"). 70 of the 76 are accepted on jlevy/squares' register; the other 6 are open registration requests, and Explore and Bounds say so. State as of 17 Oct 2026.

Some n have more than one packing at the record side. Explore's packing menu lists them, and Compare morphs between them. The catalogue's other drawings at the record side are compared with the record over the eight symmetries of the box: the same packing up to symmetry and free squares (not marked), the same tilts with squares placed differently (perhaps the same packing moved along its free motions; not tested), or a different tilt structure, an alternate. Ours, at 259, 266, 267 and 302, come from merging rotation groups of the record. Each is a valid packing at the record side, certified like the records, but our path test (turning the changed squares in small steps without enlarging the box) did not connect it to the record, so we list it as an alternate rather than as the same packing.

2 · Origins

The rest of the early field reached Gardner by reader mail and has no separate web presence: Pertti Hämäläinen (1980, improved n = 17–18), Robert Wainwright (1979, n = 19), Charles F. Cottingham (1979, diagonal-strip packings up to n ≈ 49), Evert Stenlund (1980, records for nearly all n ≤ 100), and Mats Gustafsson (1981, an alternative n = 18). Their attributions here follow Friedman's survey and Ellsworth's page.

3 · Proofs: exact values and lower bounds

Asymptotics: how much area large packings must waste (7 items)

4 · Packers and their tools

Below are the people whose packings the record table stands on and how they found them, in the order the field usually cites them.

Current records by finder

FinderRecords
David Ellsworth46
Thomas Schadt9
Frits Göbel7
M. Z. Arslanov, S. A. Mustafin & Z. K. Shangitbayev (joint)7
Károly Hajba6
David W. Cantrell5
Evert Stenlund4
Erich Friedman3
David Ellsworth & David W. Cantrell (joint)2
Michael J. Kearney & Peter Shiu (joint)1
Walter Trump, John Bidwell, Pertti Hämäläinen, Robert Wainwright, Sigvart Brendberg, Maurizio Morandi, Joe DeVincentis1 each

Counted from the "Found by" (or "Found first by") line on each row of Ellsworth's main table (not the sub-pages), one credit per current record; a joint find is one credit for the pair, not one each. This is a tally of attribution text, not a claim about who actually deserves credit for disputed or independently-rediscovered packings.

5 · Computer-checked lower bounds (2026, unrefereed)

All the items below are preprints, repositories or blog posts, none peer-reviewed, all with published code or a verifier so the claims can be checked independently. We link others' certificates as their authors published them and replay only our own; the recent s(11) and s(17) certificates use strict-core and subset-charge formats that our checkers do not read. Several of these repositories changed daily in September 2026, so the values below are dated and the repositories have the latest.

6 · Explainers and popular coverage

7 · This site

Built in 2026 from Ellsworth's 30-digit constants (§1); the parsing, contact and rigidity analysis are ours, written independently of his. Gaps and free squares are computed in double precision; "exact contact" flags are checked at 50 digits, so roundoff is never mistaken for touching. Code: github.com/evand/square-packing.

Checking our rigidity analysis against Ellsworth's own "Rigid" marks…