Hey everyone,

Today is a good day. I've finished my GCSEs and high school, it's the first day of summer, but best of all, I get to show you lovely folks my latest puzzle, the NovaMinx.



VIDEO





The NovaMinx is a vertex turning dodecahedron, with 20 axes altogether. It is the third deepest vertex turning dodecahedron to date, behind the Chopasaurus and the Starminx III. The puzzle uses a three shell mechanism, but also having an additional stability lip near the surface of the puzzle, which I suspect has helped keep the puzzle so sturdy. This is similar to what Eric used on the Master Starminx.

As this puzzle is based on the icosahedral geometry, it jumbles. There are two different ways of jumbling it; one separates a small section from the main body of the puzzle, and the other separates a large section from the main body of the puzzle. These two different ways of jumbling are shown in the first two pictures below. By combining the two kinds of jumbling, the puzzle becomes really challenging. I couldn't solve the jumbling before I applied stickers; the small jumbling is so small that many sections can be broken off of the main body, and it causes some serious pain in the neck bandaging.



VIDEO

Hope you enjoy the puzzle!

Luke

WOW. Speechless. I saw the video on youtube and came here just to comment

Really gorgeous puzzle. Is this the first time two different types of jumbling moves have been available on a puzzle like this? (Not counting puzzles that always jumble or puzzles with combined geometries like the Helicopter Skewb.)

Very Very Nice Puzzle!!! Seeing the first picture I was thinking it had face turns and shallow corner turns. I'm glad I was wrong. I like the deep corner turns much more.

Carl

Congratulations, I know how hard you worked on this!

The puzzle is beautiful, it turns exceptionally well, jumbles like crazy, has a ton of pieces... How long have you been designing again???

And he does it again! Luke proves that he is the new breed of puzzle designers!
What a fantastic puzzle! Amazing job!!!

This is fantastic! Amazing! I wish my second puzzle could be so spectacular. I am definitely going to save up for this one.

sweeeet! Puzzle of the year nominee I'd say. Looks like an absolute beast to solve. Can't wait to see more from you!

The Starminx III I believe has the same type of jumbling moves as seen here so I'd say no. And looking at this more... I think this is basically a Starminx III where some of the interior pieces have been given surface area so they can be stickered. You could almost think of this as a step toward a Multi-Starminx III. I'd love to compare the designs of this with Greg's Starminx III or Drewseph's Starminx III. I suspect all the pieces seen on the surface here are present inside these 2 Starminx III puzzles but I don't know if design details of either of those have ever been released.

I just keep liking this one more and more,
Carl

Also this really shallow jumbling appears on any edge-turning cube deeper cut than the edge-midpoints. Gelatinbrain's 3.3.2 is an example of this. So is Drew's Quad-X and Greg's DeCETH. On a Helicopter cube only adjacent edges jumble. When you cross the edge midpoint non-adjacent edges will jumble too.

See https://www.youtube.com/watch?v=q0ojiaU6G5E&feature=player_detailpage#t=54s.

From what I know of Greg's design, he skipped some parts, such as the DeFTI tips. I kept these in my design, and a size comparison of those pieces with the core is here: http://prntscr.com/b0l8p.

Also, puzzles like the Quad-X and the DeCETH have this "two kinds of jumbling" property.

Thanks for the compliments guys! Much appreciated.

Yup, had to skip them because on a SMIII they are just too small.

This puzzle is awesome!
I still can't get my head around it!
Well done.

Great job!

Wow, great puzzle! Congratulations!

Amazing Luke,

The jumbling on this puzzle is very serious. What a fantastic piece of work. How long until 3 months is up???
The cuts look like they face turn too.. imagine that.

Cheers,
Burgo.

Just when I think I know which SW puzzle I will have to buy next, along comes another one I want. This is a great puzzle, well done. Do you know approximately how much this will cost so I can work out how much I need to save?

[edit] I just thought, I suppose if you tell us how much it might cost on SW it might effect the auction price. [/edit]

When I received the template from you I though it will be Face and Corner turning puzzle. but this pentultimate like turn, with the jumble...
I can only say, that this is incredible!

Even deeper cut than Penultimate!? Must be very challenging for solving! Great Design~

This really looks like the puzzle of the year 2012!
Well done Luke!

If I buy this one it will probably be the last puzzle in my life, because I follow the rule that I'll buy new puzzles only, if I can solve all I own

I'm sure I wouldn't be able to solve it by hand. I can not reliably and consistently undo more than about 3 setup moves by hand which prevents me from solving my Super-MD. I have never gotten past the Starminx points stage.

The NovaMinx (GB's 1.2.5) has a lot of pieces and sometimes they need a lot of setups. I'm hesitant to call it "hard" since the longest routine you need is a [5,1] but it takes a long time even on a computer with macros.

I'm currently on a quest to play with awesome jumbling and there is no doubt, this puzzle has that. I can't way to jumble and try to unjumble it.

I have an idea on the price, but it may change, and as you said it may affect the auction price. However, if the winner of the auction is openly wanting the puzzle to go on Shapeways early, then I will be selling it sooner than three months.



They're two very different geometries. The Pentultimate is a face turning puzzle, and this is a vertex turning puzzle. I think Oliver was meaning that some of the cuts somewhat resembled cuts on a Pentultimate.



DKwan has explained to me briefly how he would go around solving the puzzle, and it doesn't sound too difficult (yet I'm still reluctant to try solving it ). I think with more break-in time and with enough commutators it should be a tough but not insanely difficult puzzle to solve.

Thanks for the compliments guys!

Well done, great design -
Seems very hard, but incredibely beautiful

What it is edge length?
What is its weight?

BTW: Thank you for editing the images.

Edge length: 40mm
Weight: Approx. 310g
Part count: 425 pieces; 92 pieces move in a single turn.

Congrats Luke! It's so beautiful, excellent job with the design!



It should be noted that the Starminx III has a piece type which this puzzle does not (the +centers).

EDIT: To clarify, by "+centers" I mean off-center pieces oriented towards the edges. In the SM3 you might prefer to call them the "split edges".

Ok I had to go back and check. And yes, you are correct. I was being sloppy and hadn't noticed that.

Here is the animation I made a few years ago of the Corner-Turning Multidodecahedron.



Then the Starminx III was know simply as Gelatinbrain 1.2.6 and the NovaMinx was known as Gelatinbrain 1.2.5. Here is how Andreas grouped/named these 18 piece types.

And its piece 10 which is on the Starminx III and NOT on the NovaMinx. Dare I ask how hard it would be to make the puzzle I call the Hybrid of 1.2.5 & 1.2.6 in this animation? My guess is it wouldn't be that easy unless the puzzle was a fair bit larger and therefore a fair bit more expensive but I can hope I'm wrong.

Carl

I tried to go deeper (the initial puzzle was a shallow Pentagram/Master SkewbMinx, but then kept going deeper until pieces started to lose their mechanism and/or my computer crashed). When I tried to design a deeper cut NovaMinx, I couldn't get the pieces to have any mechanism, however I haven't really gone back and looked over it again. Although for ease of stickering it would probably need to be bigger. If anyone is interested in printing a version where the Starminx III split edges are present I could have a shot at it again, although it is unlikely to work.

I think you're confusing this with 1.2.4, since solving in GB doesn't require anything longer than a [3,1] for 1.2.5. Since one of the [3,1]'s uses a slice move though, I would be inclined to call it a [4,1] on the physical puzzle. Since the algorithm lengths are on the exact same order as 1.1.5, I would say that solving a non-jumbled scramble shouldn't be hard, just very very tedious (easily a few hours). Luke is a speedsolver though, so maybe he can do it in under an hour... that would be impressive xD

With the jumbling of course, who knows how long it could take to solve it! =X

Hmm I thought it was [3,1] too but when I checked my notes I had a [5,1] written down. I need to re-visit the puzzle.