Back

Putting Rigid Bodies to Rest

100 points11 hourstwitter.com
Cieric8 hours ago

I don't have enough time to read the paper in full right now. But I'm curious if using this they could possibly find the solution to the 3 sided coin problem. I haven't heard anything about it since I watched the matt parker video about it.

https://youtu.be/-qqPKKOU-yY

Or I guess if anyone else knows the answer, that would also satisfy my curiosity.

Scaevolus7 hours ago

They should be able to simulate it! Here's another answer: https://news.ycombinator.com/item?id=33776796

hnuser1234565 hours ago

Looks like that post author forgot to loop back to the original question once they found a model that fit their own simulations.

Just visually going off the chart, the answer is a "coin" has a 1/3 chance of landing on its edge when its height is 1.7x its radius, or 0.85x its diameter. (the blog author used half-height and the paper he found uses full height)

90s_dev9 hours ago

Does this mean 2d physics simulators are about to get N times faster? Because that'd be cool if N is big enough.

almostgotcaught8 hours ago

> our key observation is that we can identify dynamically stable configurations of a rigid body, and calculate their associated probabilities

> this model is purely geometric, and does not directly account for momentum

answer: no

spencerflem9 hours ago

I love this !!

Only kinda related but I love having the opportunity to share this website, cataloging every possible fair die: http://www.aleakybos.ch/Shapes.htm

(ie: not the sort of die in the post, they must have identical faces. this thread gave me a new appreciation for the non equal faced dice tho)

iwontberude9 hours ago

I don’t think it gets much better than this. How exceedingly clever.

pvg10 hours ago

Paper https://hbaktash.github.io/files/rolling_dragons_paper.pdf and more related stuff on the page of one of the authors https://hbaktash.github.io/

spwa47 hours ago

[flagged]

qoez7 hours ago

There's still way more signal there then really any other social network (even youtube is getting slop-y)

ninkendo4 hours ago

I’d never know, every time I follow a twitter link there’s just a full page login form that blocks me from viewing the link. Twitter should be treated like a paywalled site here, IMO. I have no ability to read links like these unless I make an account, which I will never, ever, ever do.