The Making of The Pentultimate
May 5, 2008 3:18 PM   Subscribe

The Making of The Pentultimate - a beautifully obsessive documentation of modeling, casting, and assembling a "deep cut vertex turning icosahedron", or put another way, the path from 820 fiddly little bits to an elegant 32-piece mechanical puzzle.
posted by Wolfdog (19 comments total) 10 users marked this as a favorite
 
I would call it face-turning, not vertex-turning, but I would also have lost most of the bits so he can call it whatever he wants.
posted by Wolfdog at 3:22 PM on May 5, 2008


The end result.

Which I just noticed is linked here on one of the last pages.

If penultimate is second from last, is pentultimate 5th from the end?
posted by quin at 3:26 PM on May 5, 2008


I like the way the guy in the video has already resorted to bandaging the ends of his fingers.
posted by rongorongo at 3:33 PM on May 5, 2008


So, this is neat and all, and very nice, but how the hell do you get all of that material, a computer-guided mill, and enough time to design that and build it? I guess the builder must be at some university somewhere, but damn.
posted by blacklite at 3:48 PM on May 5, 2008


enough time to design that and build it

I heard an interesting comment on the radio about this the other day. Some TV person asked some interweb pundit where people got the time to edit Wikipedia. His retort? You have no right to ask anyone where they get the time to do anything - if everyone stopped watching TV tomorrow and did something productive instead we could build pretty much anything. Americans collectively watch about 250 billion hours of TV a year. Which is about 2 hours per day per person. I bet he could do all that in 2 hours a day.
posted by GuyZero at 4:16 PM on May 5, 2008 [1 favorite]


This is amazing- there is a whole community of people who make custom, Rubiks-on-steroids type puzzles, and apparently this particular icosahedron model is something of a titanic achievement.

Here's the thread on twistypuzzles.com where the creator announces it (apparently many people have sketched and discussed such a thing, but he's the first to actually make it work, and quite beautifully too). Two things stood out to me:
1) First, several of the responders have in their sigs links to their own puzzle-making pages where they've crafted custom puzzles as well. I had no idea people where crafting these sorts of things, beautifully working hand-made puzzles.
2) Second, the original post was made at Thu Mar 20, 2008 1:31 am. Noah, who's apparently quite the math whiz/puzzle solver on those forums, then posted not 2 hours later at 3:27 am, said "Have you scrambled it yet? I wrote a solution for it already if you need one. It might not be optimal, but it works.". Wow!

I guess given that some other puzzle makers on that forum in their sites mention having a "Master's degree in Algebraic Graph Theory, and soon PhD in Control Systems Engineering", these people are the savants of the math and crafting of puzzles.
posted by hincandenza at 4:21 PM on May 5, 2008


I second that. I've done some prototyping like this and it isn't all that difficult. The digital based fab is pretty straight ahead. The RTV molding requires a much higher level of physical craft. What hasn't been mentioned is the cost. Those materials aren't cheap. Neither is the equipment. The website would be a lot more interesting to me if he accounted for the time and money associated with each step.
posted by Carmody'sPrize at 4:26 PM on May 5, 2008


I hear Pinhead has a whole box of those.
posted by Blazecock Pileon at 4:29 PM on May 5, 2008 [1 favorite]


posted not 2 hours later
Well, the puzzle was described years ago and has been available to play with in virtual form for a long time, so that solution probably wasn't cooked up in the 2 hours after the announcement of the working physical model.
posted by Wolfdog at 4:47 PM on May 5, 2008


You have no right to ask anyone where they get the time to do anything

That was Clay Shirky, talking about "cognitive surplus". (I don't know who Clay Shirky is, but he sure is pushing that book.)

I was talking about money and materials and opportunity, not just time.
posted by blacklite at 4:49 PM on May 5, 2008


1) Whoa.
2) I want...
3) Whoa.
4) ...one.
posted by DU at 5:08 PM on May 5, 2008


DU - Katsuhiko Okamoto's small run of 4-layer Master Pyraminxes sold for somewhere $600-$1000 a piece, as I recall, though I can't seem to document that at the moment. I think Jason (the pentultimate guy) might have similar goals in mind.
posted by Wolfdog at 5:18 PM on May 5, 2008


0.13mm! Hand me your calipers so I can pinch you with them!
posted by furtive at 6:09 PM on May 5, 2008


Not to nitpick, but that shape is actually a dodecahedron (pentagonal faces). An icosahedron would have 20 triangular faces. It sounds like he had the idea for an icosahedron puzzle but went with a dodecahedron instead.

Now, if it had been a pie-cosahedron that would have really been something.
posted by piers at 12:47 AM on May 6, 2008


Why does he need so many layers for the interlocking mechanism? It seems overly complicated, and I'm not surprised that he has trouble turning it
posted by jpdoane at 11:44 AM on May 6, 2008


I've honestly not entirely digested how the mechanism works - and more than the molding and all the physical processes it is the design that impresses me. But it looks like each of the twelve faces causes motion at a distinct level that it "owns", and the pattern of interlocking on that level makes the other related faces turn simultaneously at the same time. So each level's tabs encode the motions for one kind of face twist.
posted by Wolfdog at 1:45 PM on May 6, 2008


The model splits and rotates about its equator, thus it is not a dodecahedron. The "Pentultimate" is not a regular solid but looks to be a stellation of the icosidodecahedron.
posted by Tube at 1:15 AM on May 7, 2008


Yes, it's a dodecahedron, it just has cuts. Might as well say Rubik's cube isn't a cube because of the cuts.
posted by Wolfdog at 5:01 AM on May 7, 2008


Sectioning a cube creates new little squares on the faces. But sectioning a dodecahedron creates new little triangles, thus turning a dodecahedron into an icosidodecahedron, or in the case of the Pentultimate, a partially stellated icosidodecahedron.
posted by Tube at 8:23 PM on May 7, 2008


« Older The story of a post-bellum South family, through...   |   The Rise of the Rest Newer »


This thread has been archived and is closed to new comments