EDIT:

OK, to keep this brief, the "kit" is now available for $328 each. This get's you the files, hardware, stickers, instructions, and "rights" to two copies of this puzzle (and / or the octahedral version, or any other external shape you may wish to create). I will sell a total of 25 "kits" here, and that's it. Ever.

For more details, refer to the below, or contact me directly at 24.cube at gmail.com (I've only set up this email account expecting it to now get spammed as I've posted it on a public forum; you'll recieve a reply from my "real" account.)

Cheers!

**********************************

OK, I've almost sorted out the details, but still have a few left to iron out so this isn't exactly set in stone, nor is it even ready to ship--I'll be done by the end of the weekend though.

Here's the background:

I printed every last piece of the puzzle I have (except the outermost caps, I cut those from polystyrene sheet), and it was expensive--both because of the high volume of plastic used, and also because things in New Zealand are not cheap. Except lamb, that's real good value. And mandarins around this time of year, too. But I'm getting off topic.

The point is that I am not willing to actually make any more myself for anyone, because it would involve either mouling and casting pieces (a skill I lack and have no interest in aquiring), or charging sums that I would not even be comfortable accepting, simply to cover the cost of printing.

At the same time though, I know more than a few people are interested in this puzzle, and with good reason--anyone who has attempted to design a mechanism for it in the past has come up against any number of surprising obstacles, all unresolved until now.

So, rather than either denying people access by way of my unwillingness to mould and cast, I've come up with a sceme...

The plan:

I will offer for sale a total of 20 (or 25) "kits" in the neighborhood of $350 each (maximum). The kit will include:

* A DVD with all the .stl files and a bill of materials
* A very detailed assembly manual in .pdf format (this thing only goes together one way)
* Stickers for two copies of the puzzle of your choice.
* All the screws, washers, springs needed
* A random extra

So why two, and what do I mean by "the puzzle of your choice"? First, by puzzle of your choice, I mean that since this puzzle is deep cut, it automatically gives all versions of the puzzle in one shot. Thus, I have designs for:

* A cube (The 24-cube)
* An octahedron
* A countoured tetrahedron
* A rhombic dodecahedron

Pick whatever shape you like, and make two *and only two* copies. The intent is that you can then sell either / both of the copies for whatever profit margin you like. Since the numbers are limited to only 20 (or 25) released copies, that means that the total number of puzzles in circulation will only ever be 40 - 50, assuming than everyone plays by the rules. Given the rather small nature of the community I think is a reasonable gamble. This way, anyone paying for this puzzle can be assured that it will hold it's value--I will not be mass producing this at any point in the forseeable future, and I doubt anyone else will be either!

So, that's the plan. There are now 6 people on the list as I write this, but maybe they will all back out. Who knows. Feel free to reserve a spot if you like, but I won't be accepting any funds until everything is ready to ship on my side.

Oh, as an added incentive, I will include a core (the most expensive item to print) for FREE to the first 8 buyers. Also, if you are planning to mould / cast the pieces, there are a total of 14 unique pieces on the cubic version of the puzzle, counting the outermost cap that I didn't actually print...

Cheers!

Interesting plan, you can definitely put me down for 1 set... and a pound of lamb. But no on a serious note, check your private messages

Ugh... man... I wish I could fund this... stupid paypal...

Cool plan.

I am in, as long as you promise to create more such ugly shape-changing puzzles in the future!




Pantazis

Please reserve a spot for me. Thank you.

Please reserve a spot for me as well!

1 for me too...

I'll take one.

Thanks!

Wow, I'm in the wrong business. Pulling in nearly $9k on a single puzzle design without having to physically make any of them ain't bad.
I'm not knocking ya, I think it's brilliant. Wish I had the cash to put out for one of these.

You're forgetting how much I've put out get the prototype up and running--you'll want to reduce your sum by well over 50%

Cheers!

Well, it shows. Amazing quality in all your designs.

Very interesting design and original marketing plan.

I have couple of questions though:
1. Do you have pictures what the other shapes will look like (contoured octahedron, rhombic octahedron, etc.)

2. Is it possible to assemble 3D-printed pieces directly or do I need to go through molding (sorry for my ignorance in the puzzle-building area

I admire this idea.
I wouldn't know where to start with .stl files and my moulding/casting is rudimentary. So my best chance is to buy one of the 20-50 puzzles being made.

I'll have some pictures soon, in the next day or two.

In the meantime, since I've been asked a few times, it's not that I'm unwilling to sell a complete puzzle, it's that a complete puzzle costs *a lot* of money to produde via 3D printing HERE. In other words, I'd have to charge sums equal to what Drew's teraminx is currently going for just to cover the printing costs. So, if you really want, I guess I can get a copy printed for you, but I'll obviously need at least 50% up front.

As well, I've also recieved quite a few comments along the line of "I'd like to make one, but have no funds" and "I have the funds, but don't know anything about casting / moulding". Since each "kit" is intended to produce two puzzles (maximum), why not simply pool resources? Just a thought of course...

Anyway, as I say, nothing is set in stone yet, and I havn't even finished off all of the manual or pricing out all the hardware--two items left to deal with. So there is no rush or anything.

Cheers!

I have a possibility to make parts (full set of parts or masterparts for moulding).
In case there is somebody who would like to share the costs and/or would like to make parts out of masterparts which I could have made (all within the rules set by the.drizzle), than please let me know and send me a proposal.

Thanks,

Geert

OK, this has taken a little longer to get done than I had planned, owing to the large number of items on my plate at the moment; sorry about that.

First, I have completed the .stl files for the octahedral skinning of this puzzle, and made the sticker templates as well. As you can see, I have opted to pillow the faces. This was because the puzzle was just too big in a planar form, and this has the added bonus of now costing less to print as well. And it looks cool too The size is roughly 85mm measure face-center to face-center, so it "matches" the 24-cube (85mm cube).

As well, I have opted to *not* offer the pillowed tetrahedron version of this puzzle unless somebody actually wants it. The reason for this is partially because it's just plain huge, and the other reason is that unlike the cube and octahedron variants, the tetrahedron requires two different outermost pieces as opposed to just one, so the resulting puzzle would be shape-shifting + jumblable, which would seem to me to not be very fun. Like I say though, if you really want it I'll be happy to provide it.

Similarly, I am waffling on wether or not I will offer the rhombic dodecahedron variant, as it may be a bit too easy to solve. That is, the puzzle would consist of 12 faces, each with 2 pieces. The problem is that since the positions of the faces can be swapped around, it would be hard to define a (semi) unique solution to the puzzle, and might be a bit on the easy side. Again, I will be happy to offer this *if* somebody wants it, but otherwise I won't be struggling to finish the design.

So, I just need to put some finishing touches on the manual, do a few sums, and the package will finally have a fixed price and be available. Oh, and since the NZ dollar has gone down recently, so will the price of the "kit"!

In the meantime, check out the octaheral variant of the puzzle, the 24-octahedron. To view it, you'll need to first rename the file to 24_octo.rar (I had to change it to .txt to bypass the positng filter), and then decompress the 3d .pdf file, which was too large to post in it's original form.

Cheers!

We appreciate the effort you put into this. Pillowed looks good and V-style LOL
Indeed, reduced amount of plastic will reduce costs in an exponential way.

Can't wait for the final update!




Pantazis

I wish I could afford one, but I am broke.

Great selling plan by the way.

I've always wanted this puzzle so please put me on the waiting list.

Some questions...

(1) Aside from knowing what 3D printing is I have no experience with it... or casting for that matter but I do know the cost of 3D printing is something that is just going to get cheaper with time so I'm tempted to buy a kit and just wait on the costs to come down. Can you tell us what you spent getting the prototype printed? If New Zealand is unusually high... anyone else care to comment what this might cost to get printed in the USA and could they let me know of some good places to go to for 3D printing.

(2) I probably know even less about moulding but considering the number of parts and their complexity I'm thinking this option won't produce as good a puzzle as 3D printing. Would someone that knows more please comment? Since I don't know how to do it 3D printing is really my only option unless I team up with someone... which if that will save me some money I'm certainly willing to do.

(3) Has the final cost been pinned down yet? I see $350 maximum above and the mention that it might be less but that is it.

(4) Does the finally cost need to be paid soon? Or can it be split up over 2 or 3 months? Money is tight at this end and Christmas is just around the corner which will soon make it even tighter, but this puzzle has been in my mind since I first saw a Skewb and a 2x2x2 side-by-side many many years ago and I'd really love to have one someday.

Thanks for all the great work you've put into your designs,
Carl

(5) Just thought of another question... since we can build two puzzles can I pick two shapes? The cube is my first pick but I'd probably want to build the pillowed octahedron as my second if that is doable?

Some answers...

1. I don't want to divulge the exact price as it's quite a bit higher than what I thought I would have spent. Let's just say that myself and whomever purchased Drew's teraminx are now on par for spending obscene amounts on a puzzle. Keep in mind though that I had the work done locally, printed every last piece, and EVERYTHING is expensive* here, especially by American standards.

I tried to get a ballpark figure from 3DPartz.com, but he flatly refused to deal with a non-US customer. As such, another forum member is helping me get a US-based quote, both for a set of masters and a complete build. I have a hunch a set of masters should cost under $200, but the core cannot be cast. I have no idea what a complete puzzle would cost, though. Give it some time, the data CD still needs to make it via post (upload 500 Mb from NZ? Yeah, right).

2. I've already suggested this point above

3. Not quite; I need to check one thing yet. Sorry for the delay, the last month has been rather hectic.

4. Deferred payments are fine, within reason. Contact me and we can chat.

5. Getting both is fine; I should clarify that I'm not requiring anyone to make two copies--feel free to make only one, or even none! The make-two-sell-one approach is a means of (potential) cost recovery for builders, while still keeping the overall numbers low enough to ensure that the puzzle holds its value as a collectors item--which would be the primary market this puzzle will ever appeal to.

Cheers!

(I'm trying to have the lot done by the end of the day, so feel free to be excited too! Oh, and the only-one-in-the-world copy is currently on its way to The Netherlands for DCD '08. Enjoy!)



*For example, in Auckland a Big Mac combo costs about US$5.00, fuel costs US$1.40 per litre (thats US$5.60 per gallon), a CD costs US$24.30, and a luxury city apartment costs US$4,200 per month and up. Meanwhile, the average income is only US$24,107. Woot!

Can someone post a jpg/png of the octahedron version? I can't seem to open the pdf.

1: the file size is 500mb? that's way too large. an STL file is no larger than 3megs for a few hundred thousand polygons. which is more tha n enough resolution considering FDM. I think you can and should optimize your STL files.

How's this:

Could you please share those quotes when they come in?


I'm pretty sure I'm going to buy the kit myself as I'd eventually like to have both the cube and the pillowed octahedron. My question was more aimed at if molding from a set of masters could make as good a puzzle as working from a full set that was printed. Just the complexity of the puzzle and the number of pieces would make me worry about the molding option. The one puzzle that actually exists at the moment was printed so we know that works... I'm just less sure about the molding option as its not something I have experience with. As I don't plan on getting mine printed right away I guess I have time to wait and see how successful others are with the molding method.


Will do. I'll wait and see what the finally price is and I'll propose a payment plan.


Thanks!!! As I want both for myself and I'm not in a rush to get them printed recovery of costs isn't my big concern.


PLEASE... anyone that gets a chance to play with this puzzle please offer up a review.

And I thought of another question...

Is the solved state of the cube unique? It looks like the arangement of the colored faces could be moved around. For example... could the top color end up next to the bottom color in an apparently solved cube? What about the mirror image of the original state?

Thanks,
Carl

I'm not sure about the pillowed-octrahedral version, but the cube version could have many "solved" states, since all the pieces are interchangeable... A "harder" version of this already difficult puzzle could be an equivalent of a "super" cube, where orientation (neighbor coloring)
matters...

Cheers,
D

Could or does? Sounds like you are saying it does have many "solved" states? I just wasn't sure. Which makes me think of a related question. This puzzle (in cube form) looks alot like the Dino Cube. However there its easy to see the bottom color could never be next to the top color in a "solved" state (as there is no piece with both the top and bottom color on it) however could a 6 color dino cube be "solved" into the mirror image of its original starting position?

Carl

since all the pieces can swap around, this puzzle DOES have multiple solved states, assuming there is no way to identify a specific "solved" position

Frankly, I wish you allowed multiple puzzles made as this puzzle is extremely moddable. I myself would want to build both offered shapes, and there are other mods I'd been thinking of. Maybe allowing for max two copies of each different shape variant would suit better? That would not limit the possible mods.

Also, are you planning on giving your customers an upgrade? The supposedly cleaner mechanism with fewer parts? I'd love an upgrade for a nominal fee!

Just something to think about.

I have no problem with that at all--I'm simply trying an approach that will keep the numbers of this puzzle low, thus rewarding everyone who gets in on it. If, for example, you want to simply use just the mechanism to build a totally unique facade, feel free .



Huh? What re-design is this?



Uh, I think it can be done, actually, although I think you'd need to bring in the "jumbling" moves to get there. Each piece of the (outer) puzzle is a triangle with only one colour on it. That is, 6 faces with 4 pieces each, or 24 pieces. The 24-Cube. Or, better yet, 24^3.

Cheers!



Oh, one more thing. Looks like $328 is the number, 25 sets total, contact me via pm or my email (if you know it) to get things rolling. I'll be shipping things once per week on Thursdays until all are sold, and I'll post a few more details in a few hours.

not my quote, but I think it is possible to swap adjacent faces without jumble-moves...

Actually the question you copied above was regarding the Dino Cube. It only has 12 pieces and each piece has 2 triangles on it. It looks just like your 24-Cube but it has turns affecting only 3 pieces at each corner. And there are no jumbling moves on the Dino Cube. I was just curious... if like the 24-cube it would have more then one solved state. I think it might have 2. The original state and its mirror image. The 24-cube could have 5 colors opposite the top face, that leaves 3 colors that could be opposite the front face, and the last 2 colors could swap sides making a mirror image so is that 5x3x2 or 30 possible solutions? I'm probably missing something in that math but if correct I'm saying the Dino Cube has 2 potential solved states and the 24-Cube has 30 potential solved states. Actual solved states is another mater for both puzzles, some of these may be able to be ruled out for parity or other reasons.



Off to send you a PM now. Thanks,
Carl

Compare this solved picture:
download/file.php?id=6649

With the color scheme of the applet:
http://users.skynet.be/gelatinbrain/App ... exa_e3.htm

I think that the 24-Cube has quite a few solved states. Possibly one for each color scheme...

I've seen the means of "counting" 30 solved states a few different ways, but what I'm unsure of is if all of the 30 states are achievable? That is, would some of them require a parity-like piece swap--which I'm told is not possible, at least not without "jumbling"...

As I've mentioned more than once on these forums, solving these puzzles is not my strong point

The dino cube does have 2 solutions (as I believe was pointed out a couple posts back, but my iPhone doesn't like the links).

The 24 cube seems to have 5! = 120 color schemes, no? One face is pre-defined, 5 possibilities for the next, then 4, 3, 2, and 1 for the following faces. I suppose the reason it's 30 has to do with the fact that orientation of the cube doesn't matter. But even so, I don't exactly see why the number to divide by is 4. Would somone be so kind as to explain it?

In this thread? I don't see it. If this question has been asked before in other threads I've missed it. The applet Puzzlemaster42 pointed us to was regarding the 24-cube puzzle.


I'm looking at it this way. Replace the six colors with the numbers 1 through 6. If all color schemes are possible in the solved state place the cube down with the number 1 on top. This is always possible. You can now have 5 possibilities for the color on the bottom. Now rotate the cube so that the lowest numbered face on one of the sides is facing forward. At this point the cube is in a unique orientation and you have 3 possibilities for the color on the back side. You are now down to the last 2 colors and there are two positions open so those are your last 2 possibilities. 5x3x2=30.

Looking at your math I think this can also be approached your way. With the number 1 on top you have 5 possibilities for the bottom color, 4 possibilites for the front color, 3 possibilities for the right color, 2 possibilities for the left color, and the back color is fixed. So that gives you your 5! = 120 but you need to divide by 4 because you can then take this cube and rotate it by 90, 180, and 270 degrees while keeping the number 1 color on top and get 4 schemes of your 120 that are really the same and differ only by a rotation of the entire puzzle.

I'm pretty sure the above is correct...
Carl

1. Look at puzzlemaster42's picture. He created a three-cycle with green, orange, yellow.
2. A swap of two opposite colors can be easily achieved by turning two parallel and adjacent edges successively three times.
3. By conjugation (see Jaaps page for theory) these two permutations can be combined two swap every two adjacend colors.
=> All 30 solved states are achievable.

The best "super" version I can think of is the truncated octahedron.

Would it be possible to post a picture of a typical kit? Since I'll be paying in a few installments I just wanted to see something to hold me over and to know exactly want to expect.

Thanks,
Carl

Interesting... is that a design you would consider making the.drizzle?

If so I think I'd prefer that over the pillowed octahedron as my second puzzle. If it was truncated would it still need to be pillowed?

Carl

By truncated, you mean an octahedron with the vertices cut down to create a face? If so then yes, I can definitely provide that, and some stickers for it too

Yes, that is what I mean by truncated. I believe contrabass is calling this a "super" version as each piece then has 3 colors on it I believe. And with 14 (6+8) total colors each piece should have a unique position in the solved state giving the puzzle 1 solution instead of 30.

Here is a puzzle I haven't worked out yet... but how many solutions does the pillowed octahedron have? I think at least two... the original state and its mirror image. The geometry isn't as easy to visualize as it is with the cube.

Carl

P.S. And thinking a bit more wouldn't a 14 color truncated cube also be a "super" version?

truncated cube


truncated octahedron

How about a 24-Dodecahedron?

And thinking a bit more... I think the perfect shape for the super version is the cuboctahedron



It could be considered both a truncated octahedron and a truncated cube and it would give the puzzle a nice symetry. The location where 4 pieces intersect would be in the center of the squares (and each square face would look like a 2x2x2 face) and the location where 6 pieces intersect would be in the center of the triangles. Each piece would have 3 roughly equal colored areas and I think look more balanced then either the pure truncated octahedron or truncated cube shapes shown above.

Ok, I want my second puzzle to be the cuboctahedron and I hope I don't appear too fickle. Technically as the cuboctahedron is a specific truncated octahedron I've only changed my mind once. Right?

Carl

Am I seeing this correctly--only two unique exterior pieces? If so, then sure, why not!

I initially sketched up a (regular) dodecahedron, but obviously embedded the mechanism incorrectly, as the pattern I ended up with was just a mess. The above is pretty attractive though...

Yes, the cuboctohedron and the truncated cube will work too. They will work any time (I think) when one of the others works. The truncated octohedron is the dual of the tetrakis hexahedron (which is the basic shape of the puzzle) and so will make a "super" shape mod. Incidentally, the dodecahedron may look cooler.

Yes, I agree all 3 shapes (truncated cube, truncated octohedron, and the cuboctohedron) would all make very nice puzzles. Since I can only make 2 and I want the basic 24-Cube too my personal favorite of these is the cuboctohedron. The tetrakis hexahedron is another interesting shape. Since the puzzle is jumblable I wonder how hard it would be just to restore the shape regardless of color. I thinking it would be rather easy but I'm not sure.



Agreed... this looks very cool. I believe this is also a "super" variant with a unique solution and it does so with only 12 colors, not the 14 of the 3 shapes discussed above. However it takes this reduction in colors at a price. It now needs 2 types of exterior pieces where as the ones above just need one and I must confess the thought of seeing that dodecahedron scrambled almost scares me. Shouldn't it be possible to put all the one color faces on the same side and the 3 color faces on the opposite side? Just what sort of weird shapes will that thing take on. I hope the.drizzle offers this mod and I hope someone builds it but just thinking about trying to solve it gets me confused. Now if I could pick 3 puzzles this might be my 3rd pick but I'll stick with the 24-Cube and the cuboctohedron. It will cost me more then enough just to get those 2 printed and even that may be some time off yet.

By the way, can someone define the technical difference between shape-shifting and jumblable? I think I understand in the context of this puzzle if all the exterior pieces are the same we are talking about jumblable only and if there are multiple exterior pieces it is also shape-shifting but I'm having a hard time pinning down the fundamental difference.

Thanks,
Carl

[edited to correct typos]

Shape-shifting is just that--the puzzle changes shape, like a mental block (as a familiar example). As the pieces change location and / or orientation the puzzle loses it's external shape.

"Jumbling" on the other hand, is the name that seems to be given to puzzles in this group (Helicopter / Bevel cube, Rua, Toru, 24-puzzle) that allow pieces to actually swap positions to achieve the changed shape. To think in terms of a 333 analogy, one would need to be able to swap a corner and edge piece to "jumble" it, which of course cannot be done.

The feature is more clear on the shallower cut versions of this puzzle (Helicopter / Bevel cube, Rua, Toru), since there are "corner" and "edge pieces to see, where as with this puzzle the feature is more difficult to visualize without the thing in your hand...

I'm ok with this definition.



This one bothers me... let me explain. Using this definition the 3x3x3 isn't jumblable. But what about the Mastermorphix which can be made from a 3x3x3? In that puzzle corners can become face centers and vice versa. If you stick with 180 degree rotation it maintains it shape just as the 24-cube does. So is the Mastermorphix jumblable? If not... what property does the 24-Cube have that makes it jumblable and the Mastermorphix doesn't have? If the Mastermorphix is jumblable... then I guess the next question might me is there a shape the 24-cube could assume that would make it a non-jumblable puzzle? I think a 24-cube in the shape of a sphere might qualify but that seems too trivial.

Carl

I think the key to jumbling is that functionally different pieces can change places and allow the puzzle to be further scrambled.

The corners and centers of Mastermorphix are both 3x3x3 corners and, though they appear different, they are functionally identical. (Just look at the parts of them that actually interact with, and make up, the mechanism.) If you turn the Mastermorphix an eighth of a turn then edges are where the corners started, and corners are where the edges started. These are functionally different. If that were a legitimate place to stop a turn then the Mastermorphix would be jumbleable, it isn't so it isn't.

Well... the 24-Cube has 24 identical pieces yet we say its jumblable.



So a knowledge of the mechanism is necessary to tell rather a puzzle is jumblable or not? And even if so this definition involving "functionally different pieces" seems to fall short for the 24-Cube. Don't get me wrong... I tend to agree the 24-Cube has a property that the Mastermorphix doesn't have. I'm just not sure how to generally specify that property so it can be applied to other puzzles. I don't think the mechanism or even the original shape of the puzzle need to apply. In other words I can see a spherical 24-Cube being called a jumbleable but non shape changing puzzle.

Carl