this post was submitted on 24 Oct 2025
39 points (100.0% liked)

Science

17990 readers
62 users here now

Subscribe to see new publications and popular science coverage of current research on your homepage


founded 6 years ago
MODERATORS
 

In a breakthrough announced in October 2025, mathematicians Jakob Steininger and Sergey Yurkevich discovered the first shape proven to lack the "Rupert property" - meaning it cannot have a straight tunnel bored through it large enough for an identical copy to pass through[^1].

Named the "Noperthedron," this 90-vertex, 152-face shape disproved a centuries-old conjecture that all convex polyhedra would have this pass-through property, first demonstrated by Prince Rupert with a cube in the 1600s[^1].

The proof combined theoretical advances with massive computer calculations, examining approximately 18 million possible orientations. "It's a miracle that it works," said Steininger, who developed the proof with Yurkevich while both worked in Austria[^1].

This resolved a geometry problem dating back to Prince Rupert's royal bet that one cube couldn't pass through another. While Rupert won that bet, and mathematicians later proved many complex shapes could have pass-through tunnels, the Noperthedron finally provided the first counterexample[^1].

[^1]: First Shape Found That Can't Pass Through Itself | Quanta Magazine

top 8 comments
sorted by: hot top controversial new old
[–] Zerush@lemmy.ml 10 points 2 days ago (4 children)

I'm not sure about it, but I think that a sphere is such a shape by default and the shape found is an aproximate sphere. So I don't really understand where are the problem.

[–] ada@piefed.blahaj.zone 13 points 2 days ago

The property is (was) specific to convex polyhedrons.

[–] SarahVanDistel@hexbear.net 4 points 2 days ago (1 children)

The problem applies specifically to polyhedra. The shape found ("Noperthedron") is fundamentally different from a sphere. And, anyways, the sphere is not such a shape... i.e. it's essentially impossible to pierce a tunnel in a sphere which would accomodate for another sphere with the same or higher diameter.

[–] frongt@lemmy.zip 3 points 2 days ago

"Essentially" in the sense of "fundamentally", not the colloquial "nearly", yes. By definition, a sphere is identical no matter how you rotate it, so its projection into 2D space does not change.

[–] woodenghost@hexbear.net 3 points 2 days ago* (last edited 2 days ago)

It works for shapes that are also pretty close to a sphere. It works for a soccer ball and for a D120. This is the first counter example found.

[–] Shadow@lemmy.ca 3 points 2 days ago

Thats my confusion too, there's even a "sphere" (polygon with enough sides to be sphere like) image in the article.

[–] SkaveRat@discuss.tchncs.de 5 points 2 days ago

I knew this topic sounded familiar. The Tom Murphy from this article is @tom7@mastodon.social and he made a video about it last month:

https://www.youtube.com/watch?v=QH4MviUE0_s

[–] farbel@mas.to 2 points 2 days ago

@Zerush That's pretty cool. I'm wondering how it will be made useful. I'm sure it will.