Online since 2002. Over 3300 puzzles, 2600 worldwide members, and 270,000 messages.

TwistyPuzzles.com Forum
 It is currently Thu Jul 24, 2014 1:42 am

 All times are UTC - 5 hours

 Page 50 of 62 [ 3096 posts ] Go to page Previous  1 ... 47, 48, 49, 50, 51, 52, 53 ... 62  Next
 Print view Previous topic | Next topic
Author Message
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 24, 2011 5:45 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Stefan Schwalbe wrote:
3.1.30

I: edge-wings and corners like a 4x4 without face-pieces
II: faces (kites) with [[1:1],1]'s. Cycle also blocks of two kites.
This part is expensive! I had all these pieces solved after 231 moves, including 22 moves to fix two swapped 3x3x3 edges. Here is the start of my method:

1. Solve all the kites intuitively, similarly to 4x4x4 centers.
2. Pair 8 edges using typical 4x4x4 edge-pairing techniques, which leave the kites intact.
3. Pair the last 4 edges using [3,1] commutators.
4. Solve the reduced 3x3x3. If necessary fix two swapped edges with two more [3,1] cycles and short setups. Fixing a single flipped edge is trickier; I'll try to find an efficient fix and I'll add it to this post.

The next part of your method is better than the way I solved it (which was [1,1] for the first orbital of small triangles, and [3,1] for the second orbital). Pairing didn't occur to me, but here we definitely save moves by pairing the triangles, not caring where they go, then solving the pairs.

My total was 930 moves. If you decide to solve this puzzle again, and solve the 4x4x4 in 208 moves, and take the same number of moves as last time to solve the rest, you will set a record of 800 moves!

Flip FU edge at the end of 4x4x4 reduction (37 moves):

[R&2,F'4,]x4, R&2, /* 4-cycle */

D',R2, /* setup moves */
U, R'&2,D,R&2,U',R'&2,D',R&2, /* 3-cycle */
R'2,D, /* undo moves */

B',R',L2&2,D, /* setup moves */
R'&2,D,R&2,U,R'&2,D',R&2,U', /* 3-cycle */
D',L2&2,R,B /* undo moves */

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Jan 25, 2011 3:28 pm

Joined: Sun Aug 29, 2010 1:56 pm
regarding 3.1.30
Julian wrote:
1. Solve all the kites intuitively, similarly to 4x4x4 centers.
2. Pair 8 edges using typical 4x4x4 edge-pairing techniques, which leave the kites intact.
3. Pair the last 4 edges using [3,1] commutators.
4. Solve the reduced 3x3x3. If necessary fix two swapped edges with two more [3,1] cycles and short setups. Fixing a single flipped edge is trickier; I'll try to find an efficient fix and I'll add it to this post.

I solved it again and I beat to flyers with one clap. 899 moves and a time of 58:44.
I solved the kites first as you suggested (around 80 moves), then pairing the wings (73 moves). I found that the last 4 wings could be paired too with my standard 4x4 method. Then the reduced 3x3 (56 moves). Then I had to fix both edge paritys wich cost me 62 moves. I hope you find a shorter way. Then pairing the triangles (207 moves), solving the triangle-pairs (154 moves), then the wedges (267 moves).

Edit: played a bit with your 37 moves alg, found some shorter moves:
2nd Edit: shortened two alg's from 29 to 25 moves
twisting the FU edge (25 moves):
[l2:D2],
(r,F4)x4, r,
[l2:D'2],
r2, [F2,[U:b]], r'2
twisting the FU edge and swapping two edges UF<->UB (25 moves):
[l2:D2],
(r,F4)x4, r,
[l2:D'2],
r2, [B2,[U':f']], r'2
swapping two edges UF<->UB (20 moves):
[(l2,D2,F4):r2],
[(U2,r',U4):[[f:d'],U2]]

The (r,F4)x4, r, is a good find by Julian. I searched for this a long time, and was absolutely amazed when I saw this for the first time. My addition is, that I bring the UF and the BU wing on the r-slice before I do the single r-move. I found this some time before and used it for the bigger cubes. So, after the single r move, we have only one 3-cycle of wings to do. My second edit shortened the setup-moves for that 3-cycle to one move.

Last edited by Stef-n on Wed Jan 26, 2011 2:22 pm, edited 5 times in total.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Jan 25, 2011 4:48 pm

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
3.1.31

Just solved it today and got a pretty high move count but OK time. After reading both help links I found my algorithm for the large edges, (that at first seemed to need a cycle to swap 3 single triangles but they form pairs to make up 3x3x3 edges) which is a [6,1] pure. I suppose mine is just a variation of the "F U R U' R' F' with anti-moves" that Julian mentioned in his improved solution, except that mine doesn't cycle the "small edges". I suppose it would be a better strategy to solve it small edges last with a pure cycle, but I'd rather not have to deal with all those confusing anti-moves. Also, I wasn't able to find a pure cycle for the small edges that didn't involve over 50 moves, that is

Although hard and confusing, dealing with these anti-moves was a new experience. I'll definitely be sure to try out a few of the others!

Last edited by Katja on Wed Jan 26, 2011 10:05 am, edited 2 times in total.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Jan 25, 2011 5:37 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
3.1.30

Excellent, Stefan! Your UF<->UB algo is 20 moves due to an overlap near the end on the U face. So we have an average edge fix at the end of the 4x4x4 stage of: (29 + 29 + 20 + 0)/4 = 19.5 moves.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Wed Jan 26, 2011 2:36 pm

Joined: Sun Aug 29, 2010 1:56 pm
Julian wrote:
3.1.30

Excellent, Stefan! Your UF<->UB algo is 20 moves due to an overlap near the end on the U face. So we have an average edge fix at the end of the 4x4x4 stage of: (29 + 29 + 20 + 0)/4 = 19.5 moves.

I was able to shorten the 29's to 25's. I have changed this in my post.
Katja wrote:
3.1.31
...
Although hard and confusing, dealing with these anti-moves was a new experience. I'll definitely be sure to try out a few of the others!

I solved 3.1.31 for the first time too a few days before. I also used Julian's last solution, wich is excellent. Instead of the usual letter notation I used color notation.
white,
yellow,
green,
blue,
red,
orange
That I transformed back into letter notation, and got things like: B R F D' L' D', L2, D L D F' R' B', D2
I will too go on and try to solve 3.1.32 and the following in the next days.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Fri Jan 28, 2011 2:36 am

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
Julian wrote:
bmenrigh wrote:
Okay so I know the Dino Skewb [3.2.13] isn't the most interesting puzzle around but I have a record for you to beat .

Edges paired in 27, reduced in 97, solved in 109.
I really like 3.2.13. I've tightened up my pairing of the edges since my first solve. My method is: match squares to edges for the first 1 or 2 edge pairs, then continue to pair edges without caring about the squares, storing paired edges by the "north pole" (UBL corner) or "south pole" (DFR corner) and using upper or lower equator slice moves to do the pairing. If a 7th pair can be made with a slice move after the poles are fully stocked with 6 pairs, this means no more than 2 cycles are needed with 5-move sequences to finish pairing the edges.

I think the biggest factor in a low move count is sheer luck going into the 2nd stage: how many squares are already correct? If the edges are paired in <=27 moves and 6 or fewer cycles are needed for the squares, or if the edges are paired in <=18 moves and 7 or fewer cycles are needed for the squares, then there is realistic hope for a sub-100 solve.

Last night I got a lucky 26 + 61 + 14 = 101, where 9 squares were correct and I only needed 6 cycles.
I tried to beat your 101 count a few times and came to the same conclusion that the second stage requires a bit a luck. In fact, based on another ~5 solves, I'd say my 105 move solve was a bit lucky too -- just not as lucky as yours.

But it occurred to me a few mornings ago that it should be possible to replace the pure square 3-cycle commutator with a non-pure conjugate. After some playing tonight I came up with a [3:1] conjugate (7 moves) and all of the variations on it to solve the squares more efficiently.

I just gave the puzzle a non-lucky 36 + 54 + 10 = 100 solve. I was a bit to aggressive with the square-matching during the edge pairing phase. After paired all of the edges I had 9 squares solved. Solving the squares did not go smoothly and I only got one case where I solved three squares at the same time. I still was able to solve the phase in only 54 moves because armed with the pure [3,1] commutator and flexible [3:1] conjugate I hardly had to use any setup moves.

Using this technique a luckier solve should get 90 moves or even a bit less.

In other news, today (January 28th) is my 1-year Gelatinbrain applet solving anniversary. It has been really nice to get to know you guys and share solving ideas. Even though our only interaction is this thread and Gelatinbrain's scoreboard, in a strange way I can't help but think of you guys as friends. You guys have made it a GREAT year and I'm looking forward to many more .

_________________
Prior to using my real name I posted under the account named bmenrigh.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Fri Jan 28, 2011 7:25 am

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
3.1.33
Attachment:

Skjermbilde 2011-01-28 kl. 13.05.55.png [ 22.64 KiB | Viewed 4032 times ]

1 - Solved just like the circle edges in 3.1.15, which I wrote about here.
2 - Solved just like the circle corners in 3.1.15, link above.
3 - Using a standard, pure [3,1] for 3x3x3 corners.
4 - Using a pure [6,1] to 3-cycle the edges.

Step 4 in this outline is the only step you have to do differently from a normal 3.1.15 solve (in 3.1.15 you'd use a [3,1] for this as well, but since there are no slice moves on 3.1.33 this is not possible).

I got a pretty good time; 13 minutes. In my opinion, this puzzle is easier than 3.1.31.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sat Jan 29, 2011 9:44 am

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
1.1.1/Megaminx
Elwyn wrote:
I am usually around 100 moves when i get to the last layer. I use a very basic orient edges, permute edges, orient corners, permute corners all with simple commutators no complex algorithms.
I recently found and used this nice site to help me solve the last layer more efficiently. (Previously I was doing OE-OC-PC-PE using my own algos, averaging 7 + 10 + 9.5 + 19.5 = 46 moves.) My new personal best for the Megaminx is 144 moves: 41 for F2L, 82 to reach the last layer, 9 for OLL, and 12 with a move overlap for PLL. Using the names from the site, I got a PLL case of "Mushroom", subcase "Jefferson Airplane"! I might be wasting precious moves by making a complete star on the first face, then putting up 5 separate corner-edge pairs. I think I can blockbuild the F2L more efficiently than that...

Edit: On a later solve, I managed to get to the last layer in 117 moves, my best ever attempt, but the last layer was tough so I only improved my PB by one move, to 143.

Last edited by Julian on Sat Jan 29, 2011 2:53 pm, edited 1 time in total.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sat Jan 29, 2011 10:00 am

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
3.2.13
bmenrigh wrote:
But it occurred to me a few mornings ago that it should be possible to replace the pure square 3-cycle commutator with a non-pure conjugate. After some playing tonight I came up with a [3:1] conjugate (7 moves) and all of the variations on it to solve the squares more efficiently.
Thanks, I shall try this myself and see how I do.

bmenrigh wrote:
In other news, today (January 28th) is my 1-year Gelatinbrain applet solving anniversary. It has been really nice to get to know you guys and share solving ideas. Even though our only interaction is this thread and Gelatinbrain's scoreboard, in a strange way I can't help but think of you guys as friends. You guys have made it a GREAT year and I'm looking forward to many more .
Happy anniversary! I feel the same way.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sat Jan 29, 2011 10:11 am

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
3.1.33
Katja wrote:
I got a pretty good time; 13 minutes. In my opinion, this puzzle is easier than 3.1.31.
I solve 3.1.33 the same way as your outline. 13 minutes is a good time! I reached the last stage after about 20 minutes and then I really struggled. 3.1.31 has harder recognition for the last stage pieces because they are pairs, but it has regular setups, while 3.1.33 has easy recognition for the last stage pieces because they are single, but it has anti-move setups. And the anti-move setups melted my brain. I just wanted to finish and go to bed, but every time I tried to set the edges up I lost track or had one flipped by mistake, and as I got more tired, I got worse!

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sun Jan 30, 2011 2:22 am

Joined: Tue Sep 08, 2009 8:41 am
Location: The Blue Mountains, Australia
Julian wrote:
I might be wasting precious moves by making a complete star on the first face, then putting up 5 separate corner-edge pairs. I think I can blockbuild the F2L more efficiently than that...
You most probably can. On my 120 move solve (yeah that's right, i got the record back ) I had done F2L at 34 moves and i used block building rather than getting the whole star. I was up to the last layer at 97 moves OLL took 9 PLL took 14.

To get to the last layer in under 100 takes some rather clever block building.

I only did 2 solves, i wonder how many i would have to do to get to the last layer in under 100 then get either a PLL or OLL skip, that would get the record down to around 110.

bmenrigh wrote:
In other news, today (January 28th) is my 1-year Gelatinbrain applet solving anniversary. It has been really nice to get to know you guys and share solving ideas. Even though our only interaction is this thread and Gelatinbrain's scoreboard, in a strange way I can't help but think of you guys as friends. You guys have made it a GREAT year and I'm looking forward to many more.
So around 1 puzzle every two days, nice. I don't know when i started but i think i've been here a little over 18 months. It's a little strange having such a long conversation/thread with people you know so little about

_________________
Some PBs
3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sun Jan 30, 2011 3:56 am

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
bmenrigh wrote:
In other news, today (January 28th) is my 1-year Gelatinbrain applet solving anniversary. It has been really nice to get to know you guys and share solving ideas. Even though our only interaction is this thread and Gelatinbrain's scoreboard, in a strange way I can't help but think of you guys as friends. You guys have made it a GREAT year and I'm looking forward to many more .
Happy anniversary! More than 180 puzzles in one year is pretty impressive, I reckon. Like Elwyn pointed out it is kinda strange having such long conversations with people you don't really know, but still I also feel that we're all friends on some level.
Julian wrote:
3.1.31 has harder recognition for the last stage pieces because they are pairs, but it has regular setups, while 3.1.33 has easy recognition for the last stage pieces because they are single, but it has anti-move setups. And the anti-move setups melted my brain.
For some reason, I felt the exact opposite; I had much more trouble with the setups for the last stage pieces in 3.1.31. I found it harder to find the proper setup for when you get two edge-pairs in the correct spots, but oriented incorrectly. And on 3.1.33 I found the fix for this problem to be rather easy.

But - the puzzle I struggled most with was 3.1.32. I solved it small edges and corners on one face, then I placed the hidden edges on the first face while I at the same time places the second-layer small edges. Then I placed the same piece-types on the last layer, which left all of the circle edges. I cycled two at a time using my [6,1] from both my 3.1.31 and 3.1.33 solution. This melted my brain

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sun Jan 30, 2011 7:07 am

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Katja wrote:
the puzzle I struggled most with was 3.1.32. I solved it small edges and corners on one face, then I placed the hidden edges on the first face while I at the same time places the second-layer small edges. Then I placed the same piece-types on the last layer, which left all of the circle edges.
A pair of circle-corner pieces is tied to each small edge, so I just ignored the circle-corners. I solved the small edges and corners like a Void Cube, which automatically solved the circle-corners. The main thing to avoid was the odd edge perm, then I cycled the centers into position in 8 moves (4 pairs of opposite moves).

Katja wrote:
I cycled two at a time using my [6,1] from both my 3.1.31 and 3.1.33 solution. This melted my brain
Same here, and it was tricky, but I did okay because it was earlier in the evening! A couple of times I deliberately solved just one circle-edge pair with a cycle, either to avoid getting one flipped, or to move a flipped one into another position ready to solve with a later cycle.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sun Jan 30, 2011 8:08 am

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
Julian wrote:
A pair of circle-corner pieces is tied to each small edge
Yeah, I noticed that because I was looking for any difference between 3.1.32 and 3.1.34. Visually they're the same, but solving the small edges won't automatically solve the circle corners on 3.1.34 making it harder.
Julian wrote:
Same here, and it was tricky, but I did okay because it was earlier in the evening!
I'm glad to know I wasn't the only one solving it this way. Now I know I'll be able to improve my move count by being more careful on the final step. A bit off topic, but I never do good solves in the evening/at night. The optimal time of the day for solving for me is in the mornings, when I get to the point where I don't feel tired anymore.

On a different note:
Attachment:

Skjermbilde 2011-01-30 kl. 13.01.25.png [ 67.7 KiB | Viewed 3807 times ]
I just beat Michael's 1.1.4 time record Barely by a minute and a half, but I've wanted this time record for a long time.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sun Jan 30, 2011 3:35 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
3.2.14 = 3.2.6 (Rex Cube) + Slice-Only 3.2.4 (Dino Cube)

This is the first Gelatinbrain puzzle I've seen that can be solved by reducing from one puzzle to another twice: first to 3.2.6 (Rex Cube), then to an easy Pyraminx (no edge orientation) or Cubominx.

1. Quick Dino Cube solve to be sure of reducing into an even center permutation.

2. Do a few quick Pyraminx-style [1,1] cycles of centers to position as many center squares correctly as possible.

3. Reduce the Rex Cube centers. The center squares are stuck in 2 groups/orbitals which can never intermix. Squares from one orbital can be cycled with side effects on the other orbital [7,1], or with no side effects [10,1]. The key to forming both algos is [slice move slice' : different move].

4. Reduce the Rex Cube leaf/petal pieces. Move the left-hand pieces to match the right-hand pieces or vice-versa. 3 pieces can be cycled pure with a [10,1] commutator, where the 10 moves are [2 setup moves : [2,1]] and of the [2,1], the 2 is a simulated Rex Cube slice and the 1 is a Dino slice (shift-click). This commutator can be shortened to 13 moves by stripping all plain (no shift-click) moves from both ends, because this is a reduction method and we don't care where the Rex Cube pieces move for now.

5. Solve the Rex Cube, which is the same puzzle as a Face Turning Octahedron (4.1.2) but in the cubic color scheme. The most efficient option for a low movecount is to build 4 big non-adjacent corners, then match the petals to each center, then solve the Cubominx.

-------------------------------------------------------------------------------------------

This is a tricky puzzle; I am adding it to the x = 10 list for now.

Last edited by Julian on Sun Jan 30, 2011 5:32 pm, edited 1 time in total.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sun Jan 30, 2011 4:10 pm

Joined: Sun Aug 29, 2010 1:56 pm
bmenrigh wrote:
In other news, today (January 28th) is my 1-year Gelatinbrain applet solving anniversary. It has been really nice to get to know you guys and share solving ideas. Even though our only interaction is this thread and Gelatinbrain's scoreboard, in a strange way I can't help but think of you guys as friends. You guys have made it a GREAT year and I'm looking forward to many more .
Congratulations to your 1-year applet solving anniversary. We meet here because of the gelatinbrain applet. I find it's a kind of friendship. Thank you for your contributions to this forum. I'm always happy to hear from you.
Julian wrote:
Katja wrote:
the puzzle I struggled most with was 3.1.32. I solved it small edges and corners on one face, then I placed the hidden edges on the first face while I at the same time places the second-layer small edges. Then I placed the same piece-types on the last layer, which left all of the circle edges.
A pair of circle-corner pieces is tied to each small edge, so I just ignored the circle-corners. I solved the small edges and corners like a Void Cube, which automatically solved the circle-corners. The main thing to avoid was the odd edge perm, then I cycled the centers into position in 8 moves (4 pairs of opposite moves).

Katja wrote:
I cycled two at a time using my [6,1] from both my 3.1.31 and 3.1.33 solution. This melted my brain
Same here, and it was tricky, but I did okay because it was earlier in the evening! A couple of times I deliberately solved just one circle-edge pair with a cycle, either to avoid getting one flipped, or to move a flipped one into another position ready to solve with a later cycle.

For me this (3.1.32) was also the hardest of 3.1.32 - 3.1.34. I just did it after I read this outline. It helped me a lot. Thank you.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sun Jan 30, 2011 6:35 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
3.2.13
bmenrigh wrote:
Using this technique a luckier solve should get 90 moves or even a bit less.
Or even 68 moves? Nice job!

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Sun Jan 30, 2011 8:31 pm

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
Julian wrote:
3.2.13
bmenrigh wrote:
Using this technique a luckier solve should get 90 moves or even a bit less.
Or even 68 moves? Nice job!
Thanks! Actually my 7-move conjugate technique is still limited to 85+ move solves. The 68 move solve came from a whole new idea that should work on a couple other puzzles too . I don't want to spoil your fun in finding it and then realizing it's pretty obvious. I haven't found any other puzzles where it is an obvious big win yet (except 3.2.13) but I'm sure there are a couple out there.

As for your 3.2.14 outline, it's embarrassing to say but 3.2.5 is my kryptonite. I see the edge and center pairs and I can cycle the pairs around pure. The issue is in solving the edges. I can't figure out how an edge can get reversed relative to the edge it needs to be paired with. At first I thought it was an orbital matching problem (hidden fixed orientation) where the edges could only come together correctly in certain orientations but that hasn't panned out either. I successfully solved both white-green, both white-orange, and both white-blue pairs and then been unable to bring both of the pieces with white-red edges together with the correct orientation.

3.2.5 is the only puzzle on Gelatinbrain's site with >= 10 solvers that I can't solve. I'm missing something obvious.

_________________
Prior to using my real name I posted under the account named bmenrigh.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 2:30 am

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
Hey Gelatinbrain, I love all of the exotic curvy-cut puzzles you are making! If you're looking for more 1.5.x puzzles, Sharon's Copterminx seems like it matches your aesthetics. Sharon said the puzzle makes use of a bit of fudging so I'm not sure if spherical / conical cuts are going to work out exactly like his puzzle.

You are adding puzzles faster than I can solve them. I think I'm going switching to macro solving for all of the puzzles that will take too long to solve by hand. I'm envious of all the cool puzzles Schuma has solved...

_________________
Prior to using my real name I posted under the account named bmenrigh.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 3:04 am

Joined: Thu Jul 23, 2009 5:06 pm
Location: Berkeley, CA, USA
bmenrigh wrote:
Hey Gelatinbrain, I love all of the exotic curvy-cut puzzles you are making! If you're looking for more 1.5.x puzzles, Sharon's Copterminx seems like it matches your aesthetics. Sharon said the puzzle makes use of a bit of fudging so I'm not sure if spherical / conical cuts are going to work out exactly like his puzzle.

Do you know why it requires fudging to work? I don't see it. If you don't jumble, do you still need fudging?

bmenrigh wrote:
You are adding puzzles faster than I can solve them. I think I'm going switching to macro solving for all of the puzzles that will take too long to solve by hand. I'm envious of all the cool puzzles Schuma has solved...

I also feel that GB is adding puzzles pretty fast. For example, this afternoon I found four new puzzles. I worked for about 4~5 hours and solved only two of them. It's actually a lot of pressure to say "I've solved all GB puzzles". I wonder how hard for GB to create a puzzle, on average.

By the way, have you used macro to solve MC4D?

edit: GB just added another puzzle (2.1.15). It looks pretty cool......but too much to do.....

_________________
Check out some virtual puzzles I created at http://nan.ma

Last edited by schuma on Mon Jan 31, 2011 3:30 am, edited 2 times in total.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 3:26 am

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
schuma wrote:
bmenrigh wrote:
Hey Gelatinbrain, I love all of the exotic curvy-cut puzzles you are making! If you're looking for more 1.5.x puzzles, Sharon's Copterminx seems like it matches your aesthetics. Sharon said the puzzle makes use of a bit of fudging so I'm not sure if spherical / conical cuts are going to work out exactly like his puzzle.
Do you know why it requires fudging to work? I don't see it. If you don't jumble, do you still need fudging?
No, I don't see it either. I think a tiny center edge between the two main edge pairs may have been eliminated. If Gelatinbrain adds it we'll see how those curves work out.

schuma wrote:
bmenrigh wrote:
You are adding puzzles faster than I can solve them. I think I'm going switching to macro solving for all of the puzzles that will take too long to solve by hand. I'm envious of all the cool puzzles Schuma has solved...
I also feel that GB is adding puzzles pretty fast. For example, this afternoon I found four new puzzles. I worked for about 4~5 hours and solved only two of them. It's actually a lot of pressure to say "I've solved all GB puzzles". I wonder how hard for GB to create a puzzle, on average.
Yeah I saw a couple new ones today too. The new slice-only-edge-turning dodecahedron (1.4.12) looks brutal. I have mostly been ignoring new puzzles because I don't want to get distracted with the ones I've already been planning on solving. The trouble is, now there are many puzzles more interesting then the ones I was going to solve. Having too many cool puzzles is a terrible situation to be in .

schuma wrote:
By the way, have you used macro to solve MC4D?
I did and I feel a bit guilty about it. I have been planning on going back and re-solving it by hand. When I first solved it I had only solved NxNxN cubes using algorithms I didn't come up with on my own and didn't understand. I muddled my way through Roice's suggested solution over the course of a few days. I don't feel like I earned the solve and I want to go back and do a macro-free, sub-1000 move solve.

_________________
Prior to using my real name I posted under the account named bmenrigh.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 8:11 am

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
bmenrigh wrote:
3.2.13

The 68 move solve came from a whole new idea that should work on a couple other puzzles too . I don't want to spoil your fun in finding it and then realizing it's pretty obvious.
Intriguing! I shall look this evening.

bmenrigh wrote:
As for your 3.2.14 outline, it's embarrassing to say but 3.2.5 is my kryptonite. I see the edge and center pairs and I can cycle the pairs around pure. The issue is in solving the edges. I can't figure out how an edge can get reversed relative to the edge it needs to be paired with. At first I thought it was an orbital matching problem (hidden fixed orientation) where the edges could only come together correctly in certain orientations but that hasn't panned out either. I successfully solved both white-green, both white-orange, and both white-blue pairs and then been unable to bring both of the pieces with white-red edges together with the correct orientation.

3.2.5 is the only puzzle on Gelatinbrain's site with >= 10 solvers that I can't solve. I'm missing something obvious.
I've fixed my headline for 3.2.14 to say 3.2.4 instead of 3.2.5.

3.2.5 is a crazy puzzle and I can't claim to understand its edges properly. My impression is that an even number of edge-pairs can be flipped, but it's very difficult to see which moves are needed to fix them without flipping others. I made several attempts and after I did some experimental moves and solved the 5th edge-pair I looked around and noticed that they were all solved. I finished and submitted, then I heaved a sigh of relief and felt less guilty when Michael came along and took the record. I was just lucky. I shall revisit 3.2.5 sometime soon, and try to understand it properly. (Does the act of observing 3.2.5 edges change their orientation? The quantum entanglement of the pieces makes me suspicious... )

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 11:28 am

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
Julian wrote:
3.2.5 is a crazy puzzle and I can't claim to understand its edges properly. My impression is that an even number of edge-pairs can be flipped, but it's very difficult to see which moves are needed to fix them without flipping others. I made several attempts and after I did some experimental moves and solved the 5th edge-pair I looked around and noticed that they were all solved. I finished and submitted, then I heaved a sigh of relief and felt less guilty when Michael came along and took the record. I was just lucky. I shall revisit 3.2.5 sometime soon, and try to understand it properly.
Okay so I gave it another hour+ of twists last night. It seems like the puzzle has a hidden orientation that has to be correct in order for all of the orbitals to come together. I paired up 3 adjacent edge groups (red-white, red-green, green-white) and then moved them around the puzzle until I found a spot where the all came together nicely. Then a bit of twisting for the remaining edges solved everything except I had two edges in their correct position but the wrong orientation (they were flipped/reversed).

I found a 4-move sequence to move a flipped edge out of place and put it back in place reversed. Unfortunately this had enough side-effects (damn entanglement!) that I couldn't figure out how to use this in a sequence to flip both. I tried a bunch of different setup move strategies. At 2AM I quit. I am going to figure out how to flip a pair on a solved puzzle and then tackle 3.2.5 again.

Julian wrote:
(Does the act of observing 3.2.5 edges change their orientation? The quantum entanglement of the pieces makes me suspicious... )
I'm going to measure the state of 3.2.5 with a hammer .

_________________
Prior to using my real name I posted under the account named bmenrigh.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 2:06 pm

Joined: Thu Jul 23, 2009 5:06 pm
Location: Berkeley, CA, USA
bmenrigh wrote:
I'm going to measure the state of 3.2.5 with a hammer .

I also remember that I had a hard time solving 3.2.5. I revisited it today and it wasn't that scary. It looks like the edge pieces are as same as the cross-corner pieces in 4.1.11. And you do have solved 4.1.11, right? The setup moves are tricky. Sometimes I have to sacrifice some solved edge-pieces. But overall I don't think there is anything fundamentally hard.

_________________
Check out some virtual puzzles I created at http://nan.ma

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 3:17 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
3.2.5 -- solving the edges

The edges are in 2 orbitals. In the diagram below, the gray pieces are irrelevant. Provided we use only LFU, RUB, BLD, and DFR, the other edges will not be affected, and we can solve the centers at the end.

Attachment:

GB 3-2-5 edges.png [ 23.55 KiB | Viewed 3777 times ]

First we solve the 3 adjacent edges at FUR intuitively, which automatically solves the other pieces marked with a *, because these edges are attached in pairs. Now we need to solve the A-a, B-b, and C-c edge pairs. If none need flipping, they are either solved or a single move away from solved. Here is a [1,1] commutator with a single setup either side to cycle the pairs clockwise while flipping B-b and C-c:

(A,B,c) = DFR, BLD, LUF, BLD', LUF', DFR'

If we just want to flip B-b and C-c, we do the above followed by DBL'.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 4:31 pm

Joined: Mon Sep 29, 2008 9:41 am
Location: Spijkenisse, the Netherlands
schuma wrote:
edit: GB just added another puzzle (2.1.15). It looks pretty cool......but too much to do.....

I believe 2.1.15 is the dual of 1.2.7. BOTH of those look like monsters...

Yesterday I solved 3 new puzzles. I've been studying for an exam 2 weeks and finally got around to it again.

First 1.4.10
The corners and face triangles are the same as 1.4.8, but the thin pieces are of what seems to be a new kind, only found in deeper cut edge turnig dodecahedra.

Then 2.1.14.
After solving 2.1.10 and 2.1.13 this one was pretty straightforward. Unfortunately I remembered my past order of solving wrong, and solved all the bitten eye-pieces in vain.

It wasn't uploaded though, so here are 3 certificates (I don't remember which belongs to which puzzle )

b614a1358918a832d5c32ba09d623bc4
97772fd69b6437c89e953dc37689ec13
5eaabc436e8fdc347aa1f428d920b34d
7986f20db946738ce916d32c6799ce01

e393083fab970d2fa280003dc89e613d
c25e89e8135ea1bc439a91dd237a85f4
0bd226b34c6786e40db746458c2b16d1
2c6798ce31e51acb34e619cd32a7584f

8f0e10849b0bbc26c18e1db6f20db946
cd39a7584faf9d613bfe97742f069b66
37c89e613dc27689ec135ea1bc656ec4
(thanks!)

And finally, I solved 1.1.40
I should have started out with this one... I started collecting algorithms at 23:30 pm, that took an hour, and then solving, which took 3,5 hours...
I made a word-document with all of the GB-style-algorithms, and pictures of what it does. I also took screenshots of every solving stage, like Julian did with the 1.4.6.
I'll post the entire document as an attachment tonight, unless someone REALLY doesn't want any spoilers. It's not effective though (ahum 11000-something moves ahum)

Since I am up to 198 puzzles solved now, I will solve 2.1.1 next, and then for my 200th puzzle I will attempt to solve 1.4.6, since it has only been solved by the two masters, schuma and Julian, and Julians post about it a while back made me reall motivated to tr it once. Since it seems to be one of the hardest puzzles on the site, I think it will be fitting to make it my 200th.

_________________
43,252,003,274,489,856,000. Or the full number in Hungarian is:

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 5:15 pm

Joined: Mon Sep 29, 2008 9:41 am
Location: Spijkenisse, the Netherlands
Detailed 1.1.40 Guide with a terrible move countSpoiler alert!
Attachment:
PDF 1.1.40 guide.pdf [607.56 KiB]

It's a PDF file. I also wrote down some comments on every algorithm that may or may not be interesting.

But as I said, this will give you a move-count of over 10.000...

_________________
43,252,003,274,489,856,000. Or the full number in Hungarian is:

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 6:02 pm

Joined: Thu Jul 23, 2009 5:06 pm
Location: Berkeley, CA, USA
Sjoerd wrote:
It's a PDF file. I also wrote down some comments on every algorithm that may or may not be interesting.

But as I said, this will give you a move-count of over 10.000...

You have such a document for every GB puzzle that you've solved? This is really a very nice illustration of the solving process. Here is how I record the algorithms:

1.1.40: circle version of 1.1.39. The first few steps are as same as 1.1.39. Types of pieces: small trapezoids, large trapezoids, pentagons, edges, edge-triangles, corners, circle edge-triangles, circle corner-triangles.

1. Solve corners as in kilominx

2. Solve edges as in pyraminx crystal

3. Solve edge-triangles, 3-cycle, affecting pentagons and circle edge-triangles.
Algo: [H2,C'2],[A,K',A'],B'&2,[A,K,A'],B&2,[C2,H'2],
Mirror Algo: [I'2,F2],[A',K,A],B&2,[A',K',A],B'&2,[F'2,I2],

4. Solve pentagons, 3-cycle, affecting circle edge-triangles.
[H2,C'],
[[C,F',C'],G'&2,[C,F,C'],G&2],D&2,
[G'&2,[C,F',C'],G&2,[C,F,C']],D'&2,
[C,H'2],

5. Solve small trapezoids, 3-cycle, affecting circle corner-triangles.
[I',B,F2],
B'&2,[F'2&2,C2&2,F2&2,C'2&2],
B&2,[C2&2,F'2&2,C'2&2,F2&2],
[F'2,B',I],

6. Solve large trapezoids, clean 3-cycles.
[L',B,F2],
[I'2&2,A2&2,I2&2,A'2&2],[B'&3,K&3],
[C2&2,F'2&2,C'2&2,F2&2],[K'&3,B&3],
[F'2,B',L],

7. Solve circle edge-triangles, clean 3-cycles:
K,[E2,B'2,E'2,B2],A'&2,[B'2,E2,B2,E'2],A&2,K',

8. Solve circle corner-triangles, clean 3-cycles:
[L'2&2,G'],
[K2,C'2&2,E2&2,C2&2,E'2&2,K'2],A'&2,
[K2,E2&2,C'2&2,E'2&2,C2&2,K'2],A&2,
[G,L2&2],

For the puzzles I solved recently, I described the three locations of a 3-cycle by words, not figures. Since I combine some setup moves into the algorithms so that these three locations are always in the white and yellow faces (the purpose is to make further setup moves easy), it's not hard to describe by words.

_________________
Check out some virtual puzzles I created at http://nan.ma

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 6:06 pm

Joined: Thu May 31, 2007 7:13 pm
Location: Bruxelles, Belgium
bmenrigh wrote:
You are adding puzzles faster than I can solve them. I think I'm going switching to macro solving for all of the puzzles that will take too long to solve by hand. I'm envious of all the cool puzzles Schuma has solved...

schuma wrote:
I also feel that GB is adding puzzles pretty fast. For example, this afternoon I found four new puzzles. I worked for about 4~5 hours and solved only two of them. It's actually a lot of pressure to say "I've solved all GB puzzles". I wonder how hard for GB to create a puzzle, on average.

It's you who's too fast to solve. You are not obliged to solve all of them. Don't worry, this site will not disappear without warning. You can play them in your pace...

Adding new puzzles is quite easy for me unless it is not completely of new type. Normally, it's just a question of a few minutes. This is a great advantage of programming compared with mechanical puzzles. The only limit is that of your imagination.

But maybe I should stop to add more puzzles in this random manner. If I don't reorganize and reclassify these puzzles at some point,later it will become quite messy.

bmenrigh wrote:
schuma wrote:
By the way, have you used macro to solve MC4D?
I did and I feel a bit guilty about it. I have been planning on going back and re-solving it by hand. When I first solved it I had only solved NxNxN cubes using algorithms I didn't come up with on my own and didn't understand. I muddled my way through Roice's suggested solution

I've never seriously tried these puzzles but I'm regularly checking this site. If you are friend of these guys, I have a proposition , a 4D version of my 3.9.x
Each of 8 cubies of hypercube(tesseract) represents only 3D surface of the 4D object, isn't it?
Supposing 4 axes of 4D space w,x,y,z, the coordinates of 8 cubies of a unit hyper cube are respectively
w = -1, w = 1, x = -1, x = 1, y= -1, y = 1, z = -1, z = 1 .
Where is the inside of a hyper cube? It exists but there's no easy way to simulate it in 3D.

Even if it is difficult to represent visually, it can be an extremely challenging puzzle with symmetrical beauty.
Indeed, geometrical or visualizable puzzles are only a tip of iceberg of all possible twisty puzzles.
Twisty puzzles as a subject of mathematical study can be far more diverse.

_________________
Virtual Magic Polyhedra
Applet(Online)
Executable Jar Installer
troubleshooting

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 6:32 pm

Joined: Mon Sep 29, 2008 9:41 am
Location: Spijkenisse, the Netherlands
schuma wrote:
You have such a document for every GB puzzle that you've solved? This is really a very nice illustration of the solving process.

No, I have one big document with pictures and the algorithm underneath, not in any particular order. However for this one I decided to make a nice documentation of my process, and make it look a bit better. I also include the reversed moves for every algorithm, but most of the time I just aplly a 3-cycle algorithm twice, that way I don't have to copy-paste as much.(this also results in waaay more moves of course...)

I just posted this one because I put a lot of work into it, and I'd like to see what Julian, Stefan, Brandon and Katja might say about it.

_________________
43,252,003,274,489,856,000. Or the full number in Hungarian is:

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 7:38 pm

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
On the topic of MC4D / hyper-twist-puzzles:
gelatinbrain wrote:
I've never seriously tried these puzzles but I'm regularly checking this site. If you are friend of these guys, I have a proposition , a 4D version of my 3.9.x
Each of 8 cubies of hypercube(tesseract) represents only 3D surface of the 4D object, isn't it?
Supposing 4 axes of 4D space w,x,y,z, the coordinates of 8 cubies of a unit hyper cube are respectively
w = -1, w = 1, x = -1, x = 1, y= -1, y = 1, z = -1, z = 1 .
Where is the inside of a hyper cube? It exists but there's no easy way to simulate it in 3D.

Even if it is difficult to represent visually, it can be an extremely challenging puzzle with symmetrical beauty.
Indeed, geometrical or visualizable puzzles are only a tip of iceberg of all possible twisty puzzles.
Twisty puzzles as a subject of mathematical study can be far more diverse.
I have created a mock-up of what I think a twist of a 4D 3.9.1 would look like:
Attachment:

hyper_3.9.x_mockup.png [ 35.31 KiB | Viewed 3780 times ]
Is this what you had in mind? One issue I can see in making an interface for this is that in the "Rubik's Tesseract" case, a "face" is a 3D cube and you only have 24 possible orientations for that cube (24 possible twists). In the case of 3.9.x, what you are twisting is the same dimension as the overall puzzle (twisting a cube inside of a cube). The 4D analogue should allow 4D twists inside of the 4D tesseract. 4D twists are very unintuitive and the result of a twist can feel like the shape in turning inside out on itself:

Just a reminder, the code for MC4D is open-source: http://code.google.com/p/magiccube4d/. The puzzle engine reminds me a bit of yours but it is less flexible. A year ago I dug through the code for about 8 hours and based on my fuzzy memory of how the underlying puzzle engine (written by Don Hatch) does things, I don't think a 4D 3.9.x is possible.

For me the primary hurdle in contributing to MC4D (besides time) is building the code. Roice and Melinda use Eclipse to program and to build. I couldn't figure out how to import the code into Eclipse in a way that it would compile. I spent many hours trying to build the jar on the command line but I'm not a Java developer and I didn't succeed.

As for 3.9.1, I did finally get around to solving it about a month ago. I literally brute-forced it, orienting one cube at a time with no thought to how it needed to be twisted. The puzzle is strikingly similar to 3.11.1 in that all of the pieces are identical and can go in any spot but the available orientations of the pieces separates them into two groups.

Edit: My mock-up isn't correct. I think the projection MC4D uses isn't 100% right for what a 4D 3.9.1 would be. In MC4D some pieces are different than other pieces (you have 1C centers, 2C face pieces, 3C middle-edges, and 4C corners). When you do a 4D 3.9.1 twist you move a 1C piece into a 2C position and a 2C piece into a 3C position. I simulated a pseudo 3D / 4D twist but a true 4D twist would effect the bottom face which isn't visible in the screenshot. More thought is needed...

_________________
Prior to using my real name I posted under the account named bmenrigh.

Last edited by Brandon Enright on Mon Jan 31, 2011 9:00 pm, edited 1 time in total.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Mon Jan 31, 2011 8:44 pm

Joined: Thu Jul 23, 2009 5:06 pm
Location: Berkeley, CA, USA
bmenrigh wrote:
As for 3.9.1, I did finally get around to solving it about a month ago. I literally brute-forced it, orienting one cube at a time with no thought to how it needed to be twisted. The puzzle is strikingly similar to 3.11.1 in that all of the pieces are identical and can go in any spot but the available orientations of the pieces separates them into two groups.

This distinction between two groups is not that crucial for 3.9.1, because the face centers are not required to be solved perfectly -- only the outside faces need to be correctly solved. For example, at the white face center, as long as the white color is seen from outside, it's fine. You can achieve this using pieces from either group. If you use a piece from from a wrong group, the internal surfaces must be wrong. But nobody cares.

However, in 3.9.1b, the distinction is crucial. At a face center, if you bring a piece from a wrong group, the arrow is always perpendicular to the correct direction so that you can notice it from outside. The worst thing is, one group has 14 pieces and the other has 13 pieces. The first time I tried to solve it, I ended up in the situation that there are two group-A pieces and one group-B piece, but there are one group-A slot and two group-B slots. And I had to discard the whole thing. It's a nice twist though.

_________________
Check out some virtual puzzles I created at http://nan.ma

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 3:21 am

Joined: Thu May 31, 2007 7:13 pm
Location: Bruxelles, Belgium
bmenrigh wrote:
On the topic of MC4D / hyper-twist-puzzles

I have created a mock-up of what I think a twist of a 4D 3.9.1 would look like:
Attachment:
hyper_3.9.x_mockup.png
Is this what you had in mind?

No, this is not what I meant.
The "interior" of the hypercube is nowhere in your picture.
The 7 visible 3x3x3 cubes in your picture and 1 hidden one are the "surface" of hypercube.
Even the core of each cube is not the interior. They correspond to 6 face centers of a 3D cube.

As evryone knows, the surface of cube consists of 6 planes.
If the edge length is 2 and the center of cube is (0,0,0),
formulas of each face are respectively
z == 1, z == -1, y == 1, y == -1, x == 1, x == -1
And the interior of cube is:
(-1 < x && x < 1) && (-1 < y && y < 1) && (-1 < z && z < 1)

Analogically, the "surface" of 4D hypercube consists of 8 cubes,
respectively z == 1, z == -1, y == 1, y == -1, x == 1, x == -1, && w == 1, w == -1
A point in the interior of the hypercube is
(-1<w && w< 1) && (-1<x && x< 1) && (-1<y && y< 1) && (-1<z && z< 1) .

The "inside out hypercube" I meant is a puzzle in which those surface and interior cells permutate.
Mathematically it's possible. The problem is how to represent visually. I have no idea...

_________________
Virtual Magic Polyhedra
Applet(Online)
Executable Jar Installer
troubleshooting

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 3:42 am

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
gelatinbrain wrote:
The "inside out hypercube" I meant is a puzzle in which those surface and interior cells permutate.
Mathematically it's possible. The problem is how to represent visually. I have no idea...
Gelatinbrain, I came to the same conclusion on my drive home from work. Per my edit, I realized something wasn't right. On further thought I understood what you were trying to say about the 8 cubes being the surface of the tesseract.

If a 3x3x3 cube is sliced into 27 cubes then wouldn't a 3x3x3x3 be sliced into 81 hypecubes? I'm not sure what sort of projection could be used so that they could be seen but if MC7D can pull off a projection of 7D into 4D -> 3D -> 2D computer screen I'm sure it is possible.

For what it's worth, forum member Mrrl (Andrey Astrelin) is the author of MC7D and he is a geometry expert, especially with hyperbolic and hyperspace geometries.

You should john the 4D_cubing group (it's basically a mailing list). Andrey, Melinda, Roice, and probably Matt Galla and Schuma would love the discussion.

_________________
Prior to using my real name I posted under the account named bmenrigh.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 3:50 am

Joined: Thu Jul 23, 2009 5:06 pm
Location: Berkeley, CA, USA
gelatinbrain wrote:
bmenrigh wrote:
On the topic of MC4D / hyper-twist-puzzles

I have created a mock-up of what I think a twist of a 4D 3.9.1 would look like:
Attachment:
hyper_3.9.x_mockup.png
Is this what you had in mind?

No, this is not what I meant.
The "interior" of the hypercube is nowhere in your picture.
The 7 visible 3x3x3 cubes in your picture and 1 hidden one are the "surface" of hypercube.
Even the core of each cube is not the interior. They correspond to 6 face centers of a 3D cube.

As evryone knows, the surface of cube consists of 6 planes.
If the edge length is 2 and the center of cube is (0,0,0),
formulas of each face are respectively
z == 1, z == -1, y == 1, y == -1, x == 1, x == -1
And the interior of cube is:
(-1 < x && x < 1) && (-1 < y && y < 1) && (-1 < z && z < 1)

Analogically, the "surface" of 4D hypercube consists of 8 cubes,
respectively z == 1, z == -1, y == 1, y == -1, x == 1, x == -1, && w == 1, w == -1
A point in the interior of the hypercube is
(-1<w && w< 1) && (-1<x && x< 1) && (-1<y && y< 1) && (-1<z && z< 1) .

The "inside out hypercube" I meant is a puzzle in which those surface and interior cells permutate.
Mathematically it's possible. The problem is how to represent visually. I have no idea...

I agree that the interior of the hypercube is not represented in Brandon's illustration. But the program doesn't have to show it, right? For example, in 3.9.1, if I turn off the animation, I can never see the interior of the cube. However I can still play it. If I need to learn the interior, I can flip it out, see it and put it back in. I think the same is true for the 4D analog. One can use imagination and memory to keep track of the interior of the hypercube.

_________________
Check out some virtual puzzles I created at http://nan.ma

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 5:25 am

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
Sjoerd wrote:
Detailed 1.1.40 Guide
My solution is kinda different from yours, as I used reduction to 1.1.3/Pyraminx Crystal and also ignoring the circle pieces until the end. If you scroll a bit down after my outline, Elwyn gives a suggestion to swap some of my steps to get better move counts for the 1.1.39 part. I really liked your outline with illustrative photos and algorithms below. Here's the ones I used:

Step 1 - Trapezoid, non-pure: (F&2, B'&4, F'&2, B'2, F&2, B&4, F'&2, B2,). I also have a pure [4,1] for these pieces, but it's not necessary when you do this step first.

Step 2 - Large triangles, non-pure: (C&2, I'&2, C'&2, I&2, H&2, C&2, I'&2, C'&2, I&2, H'&2, C&2, I'&2, C'&2, I&2,). However, this algo is pure for 1.1.39.

Step 3 - Diamond-shaped pieces, non-pure: (C&2, H, F', H', F, C'&2, F', H, F, H',). This algo is also non-pure for 1.1.39.

Step 4 - Thin triangles, pure: In my original 1.1.39 outline and when I solved 1.1.40, I used a 32-move algorithm for these pieces. But later I was able to find a pure [3,1]: (E2, C&2, E'2, C&4, E2, C'&2, E'2, C'&4,). Also you'll need the mirror image, but I won't bother including it.

Step 5 - Solve like Pyraminx Crystal.

Step 6 - Circle edges, pure: (B&2, B', F', B, H, B', F, B, H', B'&2, B, C, B', I', B, C', B', I,).

Step 7 - Circle corners, pure: I would like not to include my algorithm here, as I used a very non-efficient two-two swap at about 40 something moves. I really like your routine, Sjoerd!

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 10:34 am

Joined: Mon Sep 29, 2008 9:41 am
Location: Spijkenisse, the Netherlands
Finally got around to 2.1.1. Next up, 1.4.6!

_________________
43,252,003,274,489,856,000. Or the full number in Hungarian is:

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 11:57 am

Joined: Sun Aug 29, 2010 1:56 pm
Sjoerd wrote:
I just posted this one because I put a lot of work into it, and I'd like to see what Julian, Stefan, Brandon and Katja might say about it.
It's well done, especially the pictures of the progress. They may make someone understand, that wouldn't understand it. The puzzle has so many pieces I'm a bit tired of the many pieces - puzzles. The length of the algo's seem not optimal, but I find it is ok, to solve with it, until you have no shorter ones, and just don't worry.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 1:23 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
3.2.5
schuma wrote:
It looks like the edge pieces are as same as the cross-corner pieces in 4.1.11. And you do have solved 4.1.11, right? The setup moves are tricky. Sometimes I have to sacrifice some solved edge-pieces. But overall I don't think there is anything fundamentally hard.
Of course, thanks!

@Brandon: Remember how it bothered you that you didn't know how the split corners of 4.1.11 relate to other puzzles? It has finally clicked. They, and the edges of 3.2.5, are equivalent to the edges of two separate Pyraminx puzzles, where each of the 12 stickers of the 6 edges of each Pyraminx is represented by a piece, and each Pyraminx is tied to an orbital of 4 non-adjacent turning points. Of course, with 3.2.5 this connection is difficult to see because the stickers of the Pyraminx edges are so strangely spread out and kind of interlocking.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 1:44 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Sjoerd wrote:
I just posted this one because I put a lot of work into it, and I'd like to see what Julian, Stefan, Brandon and Katja might say about it.
I agree with Stefan, it's a nice-looking outline, easy to follow and understand. Step 4 has pretty long algos. To solve the edge-aligned triangles I would use the following [5,1] commutator (pure):

E2,E2&2,D2&2,E'2&2,E'2,
F,
E2,E2&2,D'2&2,E'2&2,E'2,
F'

It's one of those funny algos where it looks like the first 5 moves from the solved position haven't done anything!

------------------------------------------------------------------------------------------------------

Good luck with 1.4.6!!!

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 2:22 pm

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
Julian wrote:
To solve the edge-aligned triangles I would use the following [5,1] commutator (pure)

E2, C&2, E'2, C&4, E2, C'&2, E'2, C'&4,

Which is a [3,1] pure

Also, I had another look at 1.4.6 today. Now I have all the necessary algorithms to solve it. My longest one is 15 moves, which I reckon is pretty OK considering what a beast this puzzle is. I'm moving to Australia on Friday, so I don't have much time for such solves but this and 1.2.7 is definitely on my to-solve list! (I feel like at 135 puzzles solved it's time to take on the harder puzzles on the site.)

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Tue Feb 01, 2011 2:42 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Katja wrote:
Julian wrote:
To solve the edge-aligned triangles I would use the following [5,1] commutator (pure)

E2, C&2, E'2, C&4, E2, C'&2, E'2, C'&4,

Which is a [3,1] pure
Yes, better! Now that I look more carefully at my messy notes to 1.1.39, I have a mirror image of that algo with a nice star by it to indicate "optimal, I think". But I was distracted by my earlier [5,1] with the pretty diagram next to it. I never got around to organizing my scribblings or deciding on a solve order for 1.1.39 or 1.1.40, and I probably won't attempt them for a long time. Good luck with the move.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Wed Feb 02, 2011 7:06 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Sjoerd wrote:
I'm not sure if what Julian said about 1.4.6 a while back is still true, but if it is still the puzzle with the most pieces on the entire site, I might save it for my 200th puzzle solved
I wasn't sure either, so this evening I clicked on puzzles with lots of pieces and counted/calculated. Apologies in advance for any omissions and/or miscounting, but I believe these are currently all the puzzles with 300 or more pieces:

1.1.41 (Teraminx) = 530

1.1.40 = 470
1.4.6 = 470

2.1.16 = 440 410

1.1.34 = 392

1.1.39 = 350

1.4.14 = 342

1.1.33 = 332
1.1.36 = 332
2.1.8 = 332

1.4.13 = 330

3.3.16 = 314
4.3.9 = 314

1.1.31 = 302
1.4.5 = 302

1.1.43 = 300

------------------------------------------------------------------------------------------------------------

1.4.6 may not quite have the most pieces (any more), but I think it is a bigger and trickier solve than 1.1.41 or 1.1.40 and definitely a special choice for your 200th solve.

Last edited by Julian on Thu Feb 03, 2011 5:30 pm, edited 1 time in total.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Thu Feb 03, 2011 2:06 am

Joined: Thu Jul 23, 2009 5:06 pm
Location: Berkeley, CA, USA
Julian, thank you for counting the pieces. One day I also had this thought but I didn't do it. Although 1.1.41 and 1.1.40 have more (and equal number) pieces than 1.4.6, I notice that Sjoerd has solved these two puzzles. So no doubt 1.4.6 is the best choice.

Actually I think, maybe I will solve two puzzles that Sjoerd hasn't solved, and register using the name "Sjoerd"... He must be crazy. I'm kidding. I'm wondering if this kind of things happened or not.

_________________
Check out some virtual puzzles I created at http://nan.ma

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Thu Feb 03, 2011 2:23 am

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
On puzzle solving impostors / impersonators:
schuma wrote:
I'm wondering if this kind of things happened or not.
Doubtful. I feel like there is a strong trust, honesty and integrity among the core solvers. If it happens among those not on the leader board it isn't really noticeable enough to matter.

Lets not start. I really don't want Gelatinbrain to have to take drastic measures to insure honesty and integrity...

Several years ago there was a site at www.therandomgame.com that I cheated into the ground. No matter what control they added, I cheated it. I so thoroughly compromised the statistics on the site that the admins eventually gave up and shut it down. At the time I thought it was funny to cheat a game based 100% on a PRNG running in your browser. Unfortunately the site had a small but close-nit following and I completely ruined the site and the fun for the ~20 or so regular users.

I don't want to see that happen to Gelatinbrain.

_________________
Prior to using my real name I posted under the account named bmenrigh.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Thu Feb 03, 2011 5:24 pm

Joined: Sun Aug 29, 2010 1:56 pm
Hi Katja. Congrats on your 8:24 on 3.1.33! I had not the pleasure to see my 9:24 on the 1st place, for you had already replaced it. Maybe I will try again.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Thu Feb 03, 2011 5:51 pm

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Stefan Schwalbe wrote:
Hi Katja. Congrats on your 8:24 on 3.1.33! I had not the pleasure to see my 9:24 on the 1st place, for you had already replaced it. Maybe I will try again.
You are both very fast!

I have added 1.4.12 to my list of puzzles needing longer algos (x = 8).

Because 1.2.7, 1.2.8, 2.1.15 & 2.1.16 are so closely related, this weekend I will make a solution outline for all of them together.

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Thu Feb 03, 2011 6:01 pm

Joined: Thu May 31, 2007 7:13 pm
Location: Bruxelles, Belgium
bmenrigh wrote:
Several years ago there was a site at http://www.therandomgame.com that I cheated into the ground. No matter what control they added, I cheated it. I so thoroughly compromised the statistics on the site that the admins eventually gave up and shut it down.

Adding password is not enough to prevent impersonators?
Without cracking into the server, you have no way to know others' password. Or maybe that's what you did?
For my scoreboard, I don't think it's necessary. No one has interest to post by others'name.

BTW, the new installer still doesn't work on your Linux?

_________________
Virtual Magic Polyhedra
Applet(Online)
Executable Jar Installer
troubleshooting

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Thu Feb 03, 2011 6:06 pm

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
Stefan Schwalbe wrote:
Hi Katja. Congrats on your 8:24 on 3.1.33! I had not the pleasure to see my 9:24 on the 1st place, for you had already replaced it. Maybe I will try again.
Thanks! I really like this puzzle so I'm feeling a little competitive about the time record. There's nothing wrong with some friendly competition, right? Also I've realized that time records are the only records I stand a chance at; I'm not efficient enough to compete with i.e you for FM. I think your 163 moves are really impressive!
Julian wrote:
Because 1.2.7, 1.2.8, 2.1.15 & 2.1.16 are so closely related, this weekend I will make a solution outline for all of them together.
I just came up with a 1.2.7 solution myself a few weeks ago so I'm looking forward to this; it will be fun to compare notes!

Top

 Post subject: Re: Gelatin Brain's Applet Solutions Discussion ThreadPosted: Thu Feb 03, 2011 6:13 pm

Joined: Sat Jun 13, 2009 7:02 am
I just finished solving 3.1.28.
It was much easier than I thought, though I did use macros. I kept loosing track of my setup moves while solving the rounded center pieces intuitively, so I decided to use the input field on this one. I was getting sleepy half way through the solve.
I got a pretty decent time, a little under 19 minutes. My move count was quite high though, 1105 moves.

My method was nothing special:

1. solve 3x3x3 edges and corners.
2. Solve center triangle pieces intuitively
3. Solve rounded center pieces with algorithm.

I would like to try to solve the whole puzzle intuitively some time.
And I'm going to give 3.1.29 a go as well.

Top

 Display posts from previous: All posts1 day7 days2 weeks1 month3 months6 months1 year Sort by AuthorPost timeSubject AscendingDescending
 Page 50 of 62 [ 3096 posts ] Go to page Previous  1 ... 47, 48, 49, 50, 51, 52, 53 ... 62  Next

 All times are UTC - 5 hours

#### Who is online

Users browsing this forum: No registered users and 8 guests

 You cannot post new topics in this forumYou cannot reply to topics in this forumYou cannot edit your posts in this forumYou cannot delete your posts in this forumYou cannot post attachments in this forum

Search for:
 Jump to:  Select a forum ------------------ Announcements General Puzzle Topics New Puzzles Puzzle Building and Modding Puzzle Collecting Solving Puzzles Marketplace Non-Twisty Puzzles Site Comments, Suggestions & Questions Content Moderators Off Topic

Forum powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group