lots of physical hyperpuzzle updates - new world record for physical 2x2x2x2, 2x3x3x3, 3x3x3x3, new twisty stacky puzzles

[Grant S]

Hey!

Its been a while since I've sent an update to the mailing list, all of the people on the discord have been hearing about these for a while. I was going to wait until I could make better videos about the new puzzles, but I have been busy with work, and then vacation, then was isolating with covid and couldn't use my normal filming setup, and now school is about to start again, so I will be busy with that, so in the meantime, here are some updates on physical 4d puzzles.

First and most recent news is that I have broken the physical 2x2x2x2 world record set by Rowan (sorry Rowan haha) It was a 1:23, (I average around 1:50 now)

There has been some interest on the discord recently about the method I use for orienting the 2 cells at the start of the solve because I do it without any algorithms. Others have been starting to switch to that method and Luna figured it out on the virtual puzzle also. So that was an interesting development, but here is the wr video:

https://www.youtube.com/watch?v=uEnk2yrJN7I

Next, 2x3x3x3

This puzzle was finished on July 6th.

The 2x3x3x3 is an interesting puzzle because all moves are accessible without gyroing the puzzle.

It is fairly easy to solve, and actually turns fairly well.

It also has a new piece type not seen in the image above, a 1C, which is just a solid color cube.

This also led to the opportunity to use half of this puzzle as a twisty stacky 3x3x3 puzzle. Generalizing from the 2x2x2 twisty stacky that you can take any layer and switch it to the opposite side of the puzzle.

twisty stacky 3x3x3 was a very interesting solve.

I also took half of the 2x2x2x3 and used it as a twisty stacky 2x2x3 and half of the 2x2x3x3 and used it as a twisty stacky 2x3x3. Twisty stacky cuboids are actually fairly difficult to solve. Still fun though.

Ok, so, physical 3x3x3x3, how does it work.

Back in February, I made this render and posted it to the discord.

Which just seems crazy, but hear me out.

A little later, Rowan made a video about a 2D representation of a 2x2x2 rubik's cube.

which looked something like this:

He used paper to make these pieces and then slid them around to do moves.

You can see how this is similar to the physical 2x2x2x2 design, just everything one dimension down.

So, we can go from this....

to this

to this

to this

Then, for a 2D version of a 3x3x3, we have to add pieces in new places

The rest of the white and yellow pieces can go in extra middle layers.

Then the green and blue 1Cs have nowhere to go except for chilling on the outside.

So this leads to the design for physical 3^4

This thing is definitely not practical to speed solve or anything, but at least it exists.

I only had time for a quick video with it, and I didn't have my regular setup, so it doesn't have any audio, but here is a quick video showing a few turns with mc4d doing the same moves.

https://youtu.be/EkH0XYjlvj4

The Green cell turns are very tough but that is mostly because of the camera angle and that I am using one hand to hold the puzzle up to film the video. It is more manageable if you can use both hands for it.

At least it exists!

Having the pieces in the new layer means you have like multiple types of gyros. Gyros to switch out the secondary axis (The name I have given to the pink and purple pieces in the picture above) or the primary axis (white and yellow in the picture above)

This playlist has all of the videos on hypercubing I have made so far:

https://www.youtube.com/watch?v=OyvMomKXXQ8&list=PLMfvDuBNaWYOv-KRKEkQHAe8hwVkxwIJG

I'm sure there will be plenty of questions, feel free to ask and I will do my best to answer!

 - Grant

[Marc Ringuette]

Wow, Grant, awesome, an actual physical 3^4, nicely done!   So much work and so many magnets!  

The flatland versions are very insightful, I love that kind of thing, and the 2x2x3x3 and 2x3x3x3 both look good. 

For the full 3^4, some of the moves are indeed awkward, about what I was expecting a few years ago when I was trying to work it out and only got part way there with the coloring and magnets -- those last two sets of mid slices, the ones along the shared left-right axis, do have to jump around a bit.   If people are trying to understand where they go and why, this old video of mine using wood cubies moved around on a workspace marked on a cardboard box lid, might help.   This is as far as I went in this direction 4 years ago.    https://youtu.be/mj4EteEQ3Kg?t=202

Grant, you might want to spray-paint a spare cubie black, to use as an 81st no-color piece, if you want to try my workspace method of working through the moves.   Having a no-color piece allows you to arrange the puzzle into three 27-piece sub-cubes, with the "mid" sub-cube 1 dimension reduced from the other two, which I find interesting.

Cheers and congrats on the builds, and particularly pushing through to the big 3^4 milestone!  Woot!
Marc

[Grant S]

Putting the extra pieces into a 3x3x3 cube of pieces is something that Luna tried last year when making the first drawings of the 2x2x2x3 design, I attached one sketch from then, however, if you try to color these it just becomes a mess and which pieces are adjacent to which other pieces becomes completely inaccurate to the virtual puzzle.

Having a 0C piece is an interesting idea for these puzzles, but considering that there is no 0C/ core piece in either the 2d 3x3x3 or in the virtual representation of the 3x3x3x3, it doesn't make sense for it to be present in a physical 3x3x3x3. It is possible that a design with it could be made though.

--
You received this message because you are subscribed to the Google Groups "hypercubing" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hypercubing...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hypercubing/6078bf33-cbfb-61a7-4fc0-cdd5952eebbf%40solarmirror.com.

[Robert DuBeau]

Just wow, this is insanely cool. Thank you for doing this and sharing your achievement here. From someone who so far can barely solve the 2x2x2x2 I just have to say it is nice to see that the puzzles you made could be done. Congratulations! They are really elegant.

--

You received this message because you are subscribed to the Google Groups "hypercubing" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hypercubing...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hypercubing/CADE7h9ruxk32GjRe7_nWbSSezYOBtCYpfqOs_-0XgjOA93g9tQ%40mail.gmail.com.

[Melinda Green]

On 8/8/2022 6:00 PM, Grant S wrote:
> considering that there is no 0C/ core piece in either the 2d 3x3x3 or in the virtual representation of the 3x3x3x3, it doesn't make sense for it to be present in a physical 3x3x3x3.

That depends upon how you look at these puzzles. Puzzle makers typically start with a mechanism as a guide to discovering potentially interesting puzzles. Personally I prefer to start with interesting puzzle groups/topologies and look for possible mechanisms with which to implement them. This is an instance of the former which seems to naturally suggest a 0C piece. That doesn't mean it should have one, but if it helps mechanically, then I would definitely use it.

-Melinda