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

TwistyPuzzles.com Forum

It is currently Fri Jul 25, 2014 11:34 am

All times are UTC - 5 hours



Post new topic Reply to topic  [ 3096 posts ]  Go to page Previous  1 ... 46, 47, 48, 49, 50, 51, 52 ... 62  Next
Author Message
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Tue Jan 11, 2011 2:10 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
Recently solved all the maze and arrow cubes. The maze 5x5 was hard, I had much to concentrate while reducing the centers. Now I'm glad that I did it, it was a big hill for me. If there is any interest, you have to know the pieces for solving it. Especially for the 5x5 maze cube. I prepared some pictures for that reason. And you need a picture of the solved cube. Or it is a really hard puzzle, wich is nearly impossible to solve.
Attachment:
3.1.3b.png
3.1.3b.png [ 19.03 KiB | Viewed 4136 times ]
Attachment:
3.1.4b.png
3.1.4b.png [ 21.26 KiB | Viewed 4136 times ]


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Wed Jan 12, 2011 12:56 am 
Offline
User avatar

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
Julian wrote:
bmenrigh wrote:
Yesterday I solved 3.2.13 thinking that my reduction to a Dino Cube would beat Julian but nope
I reduced 3.2.13 to a Dino Cube too! :lol: I made some intuitive moves then 5-move algos to swap the 2-sticker pieces, (3,1) algos to swap the squares, then solved the reduced Dino Cube, in 48 + 80 + 12 = 140 moves.
Okay so I know the Dino Skewb isn't the most interesting puzzle around but I have a record for you to beat 8-) .

Edges paired in 27, reduced in 97, solved in 109. I'm sure if I could conjure a little inner Elwyn I could solve this in slightly less than 100.

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 13, 2011 2:03 am 
Offline
User avatar

Joined: Sat Apr 22, 2000 2:51 am
Location: New Ulm, Minnesota, USA
Gelatin Brain could you please answer my question here?

viewtopic.php?p=242178#p242178

I also have another question. Do you have hidden puzzles on your website? If you do why do you do that and how do you get to play and see them.

_________________
Darren & Traiphum's Dual Helicoptrahedron
X-TownCuber wrote:
Are my eyes deceiving me, or is this the coolest puzzle ever?


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Fri Jan 14, 2011 12:28 pm 
Offline
User avatar

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
Firstly, I want to mention that I had my forum name changed due to various personal reasons. Most of you already know me as "Katja" from the rankings, but since this is mainly the only place I'm interested in posting these days, I thought it would suffice to announce it here :)

Recently I decided to have a go at 3.4.1 (2x2x2 + Skewb), and I was finally able to find a pure (5,1) commutator for the center triangles (I have no idea how to refer to these pieces as correctly as possible). After making this discovery I tested the algo on some of the other 3.4.x's, and it works for almost all of them. However, not always pure but can easily be commuted to cycle pieces pure.

At last, the mystical world of 3.4.x's is open to me :D


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Fri Jan 14, 2011 6:04 pm 
Offline

Joined: Sat Jun 13, 2009 7:02 am
I just started out on Gelatin Brain. It takes a little time to get used to solving virtual puzzles, it's quite different from solving physical puzzles.
I started with the easier puzzles, like 3.3.20 and 3.7.6 and I got a decent time on both.
I liked the extra challenge the curvy copter gave over the normal heli cube.

For some reason I'm having more difficulty with normal cubic puzzles than with the other, more exotic ones. I got a terrible time on the normal 3x3x3. :lol: The viewing angle takes a little getting used to.

To GelatinBrain: I accidentally misspelled my last name, it should be "van der Laan" with an "N" at the end, not an "M". So the name in the ranking list is slightly incorrect. Could you change it for me?


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Fri Jan 14, 2011 6:53 pm 
Offline
User avatar

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
Welcome! :D We don't get a lot of new "members" in this thread.

I also felt the same way starting to solve GB puzzles, but with practice and some getting used to I feel it's more natural to solve computer puzzles rather than physical.

3.3.20's a great puzzle and if you want to try another variation of a Helicopter Cube, you could give 1.4.1 a go. It's basically a 12-sided Helicopter Cube. For extensions of the normal 3x3x3 I've been enjoying 3.1.36 recently, or if you haven't solved it already, the standard circle 3x3x3 - 3.1.7 - is a good place to start if you want to get into circle puzzles.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sat Jan 15, 2011 7:56 am 
Offline

Joined: Sat Jun 13, 2009 7:02 am
Thanks Katja!

I just gave 3.1.36 a try. It was easier than I thought, but a lot of fun to solve. It took like 8 minutes on my first solve because I messed up twice, but I'm sure I can do much better. It took me a while to realise I could just apply void cube parity algorithms to fix the centers.

The helicopter dodecahedron looks like a lot of fun too. I'll give that one a go tonight.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sat Jan 15, 2011 1:37 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
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 8-) .

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.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sat Jan 15, 2011 1:53 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
@Stefan: Thanks for itemizing the pieces of 3.1.3b and 3.1.4b. It could be a long time before I have the courage to attempt them, but if/when I do, it's nice to know your diagrams are here. :)


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sat Jan 15, 2011 4:58 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
I solved 3.1.15 two times and I was surprised, to get the move-count record with 145 moves. Here is my outline, wich I don't suppose to be optimal:
Attachment:
3.1.15.png
3.1.15.png [ 8.86 KiB | Viewed 3927 times ]

I and II: 4 circle-edges of one color, then parity check (corners) and fix with one move, then the rest of the circle edges wich are 12 pieces all together. Use a (1,1).
III: circle-corners with a (1,1), also 12 pieces.
IV: corners with (3,1)'s
V: edges with (3,1)'s wich also use slice moves.

I and II: 27 moves
III: 19 moves
IV: 23 moves
V: 76 moves
= 145 moves.

I first thought ok it solves but it is hardly optimal. It seems not to be optimal to solve the edges with (3,1)'s. Maybe you can improve my method. Take the challange.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sun Jan 16, 2011 1:35 pm 
Offline
User avatar

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
3.1.15
Stefan Schwalbe wrote:
Here is my outline, wich I don't suppose to be optimal:
:idea: After I read this I had a revelation and immediately knew how to solve it, thanks Stefan! :D I used the exact same steps as you;

1 - Circle edges - which I realized can be solved the same way as the invisible edges on puzzles like 1.1.29 etc.
2 - Circle corners - same trick as the circle edges.
3 - Corners - using a standard 3x3x3 algo, such as: U', F', U, B', U', F, U, B,
4 - Edges - also using a standard 3x3x3 algo, such as: U&2, F', U', F, U'&2, F', U, F,

I included some hidden spoilers to each steps, which give away the puzzle and makes it surprisingly easy. Highlight to read.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sun Jan 16, 2011 7:43 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
3.1.31 - Improved solution

(Original solution here.)

1. Solve all pieces except the small edges like a Rubik's cube, but every move must be made as an "anti"-move, moving the concave part of the puzzle only, so for F do instead B with a clockwise rotation around the F face.

2. Solve small edges pure with a (6,1) commutator. Try doing the familiar Rubik's algo F U R U' R' F' with anti-moves, and you have isolated a small edge in the D layer.

Average solves of around 180 moves should be possible: step 1 with Fridrich in around 60 moves, and step 2 solving a pair of edges each time in 6 cycles, with an average of 3 setup moves.

Edit: Elwyn is right, I over-estimated the setup moves needed to solve the small edges. Perhaps half as many are actually needed, around 1.5 moves per cycle? Also Jessica Fridrich estimates a 56 move total for her method, which gives a better estimate of 56 + (6*17) = 158.


Last edited by Julian on Mon Jan 17, 2011 5:24 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Mon Jan 17, 2011 7:27 am 
Offline
User avatar

Joined: Tue Sep 08, 2009 8:41 am
Location: The Blue Mountains, Australia
Julian wrote:
3.1.31 - Improved solution
My method was going to be similar to your old one but with a (6,1) alg for the detached edges which was just F R U R' U' F'. I'd still use it if i wanted to be fast as i think it would save time as there are far less anti-moves which are confusing.

Anyway I solved it using your method instead :) 3x3x3 took me 58 (could be a little better) the rest took 98 so all up 156. There was nothing lucky about the second step either, not one edge was already in place and two cases where two pieces were in each other's position and one edge was in place but oriented wrong. It takes 2 three cycles to solve just three pieces for both of those situations. I'd say three set-up moves is too many, three was the most i used and a lot of the time 1 or 2 was enough... perhaps I was just lucky there rather than in other ways.

Oops, forgot to post this then did another similar solve of 152 (3x3x3 took 52 the rest took 100).

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Mon Jan 17, 2011 4:24 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
Gelatinbrain you often surprise me with new ideas. 3.1.17 behaves not as I had expected. The inner slice moves turn only the inner circles of one side. Thus this puzzle isn't fully symmetric! - that's strange.
Attachment:
3.1.17.png
3.1.17.png [ 22.29 KiB | Viewed 3654 times ]

Katja wrote:
3.1.15
Stefan Schwalbe wrote:
Here is my outline, wich I don't suppose to be optimal:
:idea: After I read this I had a revelation and immediately knew how to solve it, thanks Stefan! :D I used the exact same steps as you;
...
Your algs are very similar to mine.
I have decided to solve all the 3.1.x's. I'm now at 3.1.17. For 3.1.16 I also got the move-count record. I find the time's are a bit more important than the move-count's. I may share my 3.1.16 approach.
3.1.16:
Attachment:
3.1.16.png
3.1.16.png [ 8.58 KiB | Viewed 3887 times ]

I with (1,1) fix the parity of the inner 2x2 (II pieces in the picture) before you complete I
II with (3,1)'s. its the inner 2x2
III with (1,1)'s, again, fix the parity of the corners (V) before you complete
IV and V like a 4x4 with already reduced faces
If you end with two swapped edges, here is a fixer:
u, (L', F, U', L, F'), u', d', ( F, L', U, F', L), d
I have found the last algo somewhere in the forum, wich I don't remember exactly where.


Last edited by Stef-n on Thu Jan 20, 2011 2:48 pm, edited 2 times in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Mon Jan 17, 2011 4:52 pm 
Offline
User avatar

Joined: Thu May 31, 2007 7:13 pm
Location: Bruxelles, Belgium
Stefan Schwalbe wrote:
The inner slice moves turn only the inner circles of one side. Thus this puzzle isn't fully symmetric - strange.

This is a bug. I've just fixed it for both exe and java. :wink:

_________________
Virtual Magic Polyhedra
Applet(Online)
Executable Jar Installer
Win32 Executable(Download)
troubleshooting


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Mon Jan 17, 2011 5:04 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
gelatinbrain wrote:
Stefan Schwalbe wrote:
The inner slice moves turn only the inner circles of one side. Thus this puzzle isn't fully symmetric - strange.

This is a bug. I've just fixed it for both exe and java. :wink:
Ah, this is a bug :P


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Mon Jan 17, 2011 5:52 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
With the shorter algos found recently by Elwyn, schuma, Brandon & Katja, there aren't many puzzles left that need longer than [8,1] commutators to solve, so I think it's time for an updated list that expands downward to x > 6. The same [7,1] can be used when solving the puzzles marked in bold, and it has been proven optimal by a computer program written by Brandon and discussed here.

If one solves the Gelatinbrain puzzles* entirely with [x-or-less, 1] commutators and setup moves, where the 1 is a regular move or a slice move, the following puzzles seem to have a minimum possible x > 6:

x = 7
1.1.32, 1.1.33, 1.1.34, 1.2.9, 1.2.15x, 1.4.12, 1.5.1, 1.5.2, 2.1.5, 3.3.7, 3.3.10, 3.3.15, 3.3.17, 3.3.19, 3.10.3x, 4.3.3

x = 8
1.1.26, 1.1.27, 1.1.31, 1.1.35x, 1.1.50, 1.2.8, 1.3.2, 2.1.8, 2.4.1, 3.3.11, 3.4.23, 3.6.5, 3.7.2, 3.7.5, 4.3.2, 4.5.1, 4.7.2

x = 10
1.4.3x, 1.4.7x, 2.3.1, 3.2.14, 3.5.1

x = 12
3.3.6, 3.4.24, 3.5.3, 4.3.4

* Excluding the spheres


Last edited by Julian on Wed Feb 09, 2011 7:22 pm, edited 7 times in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Tue Jan 18, 2011 4:25 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
Hi Julian, seems you have an own classification system based on the length of the move-sequences. If someone finds something shorter, wich may happen, you would update it. It would be nice to have a print with the best move-sequences, wouldn't it?


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Tue Jan 18, 2011 4:43 pm 
Offline
User avatar

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
Stefan Schwalbe wrote:
Hi Julian, seems you have an own classification system based on the length of the move-sequences. If someone finds something shorter, wich may happen, you would update it. It would be nice to have a print with the best move-sequences, wouldn't it?
Yeah he has been updating it. I'm sure many can be improved upon. Doing so is likely to be very hard though!

Julian, I'm really glad you are keeping a list. I think you might want to mention 3.3.7 (little Chop) has been proven by way of my program to require at least (7,1) (see this post).

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Tue Jan 18, 2011 5:55 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Stefan Schwalbe wrote:
Hi Julian, seems you have an own classification system based on the length of the move-sequences. If someone finds something shorter, wich may happen, you would update it. It would be nice to have a print with the best move-sequences, wouldn't it?
Yes, I'll either update it or link to a newer list. I have thought about putting actual move sequences beneath the list (probably in invisible ink), and color-coding related puzzles that use exactly the same longest algo, in a later list.

I like the idea of classifying puzzles this way by analogy with chess. If we agree that it is harder in chess to see a winning/drawing idea that pays off 5 moves ahead than one that pays off 3 moves ahead, then surely a similar difference exists with twisty puzzles. How many moves ahead do you need to explore to find a solving algorithm? And, similarly to chess, it doesn't take many moves before it is impossible to explore all of the possibilities without a computer program, so we must rely on good instincts and observation to explore the most promising avenues.

Most mass-marketed twisty puzzles have x <= 5 if we extend move-sequences beyond commutators to include also conjugates, mirroring, and repetition of 2-gen fragments. Tricky exceptions include the Square-1 and Bandage Cube. Square-1 has x = 7 (the maximum length to restore cubic shape, and also to resolve an odd edge perm), and the Bandage Cube has x = 11 (following Jaap's analysis, the longest sequence needed to build a 2x2x2 block around BLD and be able to rotate F, R, and U).

Possible uses of this classification of the Gelatinbrain puzzles:

1) When working on a solution to a difficult puzzle, see if your x matches the value in the list. If your x is higher, you can choose to look further to find an improvement; or if your x is lower, you can share your sequence and the rest of us learn something new, and the list improves its accuracy.

2) For fun and curiosity, see what the most complex puzzles have in common. Just looking at a puzzle's cuts, can we predict what kind of puzzle requires at least one long algorithm in its solution?

3.) Following on from 2), the list can help us estimate how efficient a solution is to a brand new puzzle. Does the new puzzle seem like an x = 7 or 8 specimen, or harder, or easier? Our instincts are often correct! :)

4) To avoid the most difficult puzzles for a while, especially when new to Gelatinbrain, or at least have an idea of what you're letting yourself in for if you do take the plunge!


Last edited by Julian on Tue Jan 18, 2011 6:23 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Tue Jan 18, 2011 6:18 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
bmenrigh wrote:
Julian, I'm really glad you are keeping a list. I think you might want to mention 3.3.7 (little Chop) has been proven by way of my program to require at least (7,1) (see this post).
I intended to do that, to mark 3.3.7 and related puzzles as having a proven optimal x and linking back to your post, and also of color-coding puzzles with related algos, etc. But fatigue took over and I ended up just pushing my scribbled list online! I'll do a basic edit to my post now.

Speaking of your prog for 3.3.7, it may be very helpful in homing in on the value of what I call X, the longest sequence length needed to solve all of the Gelatinbrain non-spheres. Because the small edges of 3.3.6 and 4.3.4 move exactly the same as the pieces of the Little Chop, if we can prove that no possible [7-11,1] Little Chop pure cycle makes a pure cycle when pasted into 4.3.4, we have proved that those small pieces require [12,1] or greater to solve. Then if we can prove that the other piece types of 3.3.6/4.3.4 cannot be solved pure with shorter than [12,1] either, we have established that X >= 12. The alternative is that we do find a shorter pure algo than [12,1] to finish 4.3.4, in which case only two puzzles stand between us and reducing X from 12.

Please can you PM me all the [7-11,1] algos your program found? I'm really curious.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Tue Jan 18, 2011 6:51 pm 
Offline
User avatar

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
Julian wrote:
[...]But fatigue took over and I ended up just pushing my scribbled list online! I'll do a basic edit to my post now.
Oh I totally understand. Fatigue has been keeping me from finishing classifying every Pentultimate pattern my program (with GuiltyBystander's optimizations!) has found When I saw you post I thought "oh man that must have taken a long time". Do you write routines down? I don't and I would never be able to remember the exact move sequence length and re-finding some of them has proven quite difficult in the past..

Julian wrote:
Speaking of your prog [...] Because the small edges of 3.3.6 and 4.3.4 move exactly the same as the pieces of the Little Chop, if we can prove that no possible [7-11,1] Little Chop pure cycle makes a pure cycle when pasted into 4.3.4, [...] Then if we can prove that the other piece types of 3.3.6/4.3.4 cannot be solved pure with shorter than [12,1] either, we have established that X >= 12. [...]
Yeah I had the same thought. A sort of proof-by-equivalence.

Julian wrote:
Please can you PM me all the [7-11,1] algos your program found? I'm really curious.
You may be surprised by how many routines the program found for the longer sequences! I'm not trying to keep the routine secret but I also don't want to flood the forum with a gigantic list that is questionably useful.

So instead, I have zipped out the raw text output of my program. The zip is 18 megs but it expands to 221 megs: http://noh.ucsd.edu/~bmenrigh/little_chop_routines.zip. It should be obvious which routines are which based on the filenames.

As an aside, your list of sequence lengths often achieves short routines by toying with the solve order so that the harder pieces can be solved non-pure. For newer solvers this can be quite difficult. I'd hate for a new solver to look up a puzzle, see that it isn't on the list and then be unable to find a (5,1) sequence, only to learn that your expected solve order is quite complicated to avoid long sequences.

It would be great to have a pair of sequence lengths, the one you have listed above and also the length of the shortest known pure sequence. Obviously this would be ton of extra work so I'm not asking for you to actually work on the list or anything like that. One day though :D .

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Tue Jan 18, 2011 9:09 pm 
Offline
User avatar

Joined: Tue Sep 08, 2009 8:41 am
Location: The Blue Mountains, Australia
bmenrigh wrote:
Do you write routines down? I don't and I would never be able to remember the exact move sequence length and re-finding some of them has proven quite difficult in the past..
I was wondering the same thing yesterday and i'm pretty sure he must have a rather interesting long list of algorithms.

A couple of months ago i spent far too long looking for an alg that i had found before but forgotten, this caused me to get a little angry at myself and so i started writing all the more difficult to find algs in a word document. I'm thinking Julian does the same or has one of the most incredible memories ever :lol:

On a different note, Julian, there's something i've been meaning to ask you. You talk about adding set-up moves to algs, and change say a (3,1) to a (5,1) to help isolate a piece in a turning layer. I've always found it much easier to instead change the latter part of the commutator and change a (3,1) to a (3,3) which is the same amount of moves as a (5,1) but you never seem to do this and i was wondering why?

As an example, for the new 1.3.13 I can cycle the small inner triangle pieces pure in 12 moves only if it's a (3,3) and the shortest (x,1) i can find is (6,1) because i never really got my head around adding the set-up moves to algs...... Wait, i just found a (5,1) with your set-up move system... so it seems i can find them but i don't find them easily, so i'll just go back to my original question of why you never use algs of the form (x,y) only ever (x,1).

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


Last edited by Elwyn on Tue Jan 18, 2011 9:54 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Tue Jan 18, 2011 9:39 pm 
Offline
User avatar

Joined: Thu Jul 23, 2009 5:06 pm
Location: Berkeley, CA, USA
Elwyn wrote:
bmenrigh wrote:
Do you write routines down? I don't and I would never be able to remember the exact move sequence length and re-finding some of them has proven quite difficult in the past..
I was wondering the same thing yesterday and i'm pretty sure he must have a rather interesting long list of algorithms.

A couple of months ago i spent far too long looking for an alg that i had found before but forgotten, this caused me to get a little angry at myself and so i started writing all the more difficult to find algs in a word document. I'm thinking Julian does the same or has one of the most incredible memories ever :lol:


I wonder the same thing. I'm maintaining a long google document for myself. I downloaded the file today as a TXT and it's ~1MB of pure text. It contains the major steps and necessary algorithms for all twisty puzzles that I can solve including all GB puzzles. Many algorithms are much longer than the ones in this thread, but I'm content because they are my hard work anyway.

Unfortunately most of the comments are in Chinese. So sharing the whole file with you guys requires a lot of translation.

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Tue Jan 18, 2011 10:02 pm 
Offline
User avatar

Joined: Tue Sep 08, 2009 8:41 am
Location: The Blue Mountains, Australia
schuma wrote:
It contains the major steps and necessary algorithms for all twisty puzzles that I can solve including all GB puzzles.
:lol: I thought you might say something like that... but it still surprised me. That must be one amazing document.

Back to 1.3.13 I thought i'd point out i wouldn't need that (5,1) anyway as i'd solve it with reduction to 1.1.48... not saying that would be easier but i think it would be more efficient. Even without reduction i'd solve the centres last pure (3,1) so i could solve the triangles (3,1).

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Wed Jan 19, 2011 4:24 am 
Offline
User avatar

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
schuma wrote:
It contains the major steps and necessary algorithms for all twisty puzzles that I can solve including all GB puzzles. Many algorithms are much longer than the ones in this thread, but I'm content because they are my hard work anyway.
I do a similar thing. I have a folder named Gelatinbrain's Applet containing 52 text documents with algorithms that are just my own. Sometimes I add others, but then I write that in. For some reason I don't write mine in Norwegian, I've found that my own language lacks certain words when it comes to solving puzzles etc. There is for example no good word for solve, or even twisty puzzle. So I do mine in English :lol: I took a screen shot:
Attachment:
Skjermbilde 2011-01-19 kl. 10.15.29.png
Skjermbilde 2011-01-19 kl. 10.15.29.png [ 129.92 KiB | Viewed 3702 times ]
A lot of the text files are on puzzles I've solved and then there are some for puzzles I've simply played with and found an algorithm I'm satisfied with. I should mention that even though I write all of them down, I still memorize them before going in on a solve.


Last edited by Katja on Wed Jan 19, 2011 9:52 am, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Wed Jan 19, 2011 9:32 am 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
I collect my solution sheets (one sheet per puzzle) in a physical folder. For the effects of move-sequences I use small drawings instead of letters. ImageFor the move-sequences I often use GB-notation or if GB-notation is too long, my own notation.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Wed Jan 19, 2011 1:21 pm 
Offline
User avatar

Joined: Mon Sep 29, 2008 9:41 am
Location: Spijkenisse, the Netherlands
Haha! I too have a word-document full of GB-algorithms (that can be put in the algorithm-bar) with a screenshot of what is exactly does. It's not very orderly, more like in a chronological order of solving.

_________________
Olivér Nagy wrote:
43,252,003,274,489,856,000. Or the full number in Hungarian is:
Negyvenháromtrillió-kétszázötvenkétbilliárd-hárombillió-kétszázhetvennégymiliárd-négyszáznyolcvankilencmillió-nyolcszázötvenhatezer :wink: )


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Wed Jan 19, 2011 3:57 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
bmenrigh wrote:
When I saw you post I thought "oh man that must have taken a long time". Do you write routines down? I don't and I would never be able to remember the exact move sequence length and re-finding some of them has proven quite difficult in the past..
Yes, I write my routines down. I had the intention of organizing them in a folder but instead I've ended up with several spiralbound notebooks with no organization at all! I have a pretty good memory for the overall shape of solutions and the nature of the algos and their length, but I have a poor memory for the details and if I didn't write things down I'd have to rediscover most of them. I usually include diagrams, similar to the example Stefan posted. If it is tricky to see how the q move varies the position of the 3rd piece of a cycle, I put a ring around the default position and draw in hollow pieces for the other positions. While figuring out a solution I have shorthand symbols for "cycles this piece type", "also moves these other piece types", "promising but needs tweaking", and "probably the best I'll find for this stage".

Recording algos is very advantageous when tackling new puzzles. If I see a new puzzle with a similar shape and depth to one I have already solved, one of the first things I'll do after improvising for a few minutes is to look up algos that worked with the previous puzzle, and see what they do with the new one. It is rare that at least one of them isn't useful, either the same or with a bit of tweaking.

bmenrigh wrote:
You may be surprised by how many routines the program found for the longer sequences!
I was! Thanks for the files.

bmenrigh wrote:
As an aside, your list of sequence lengths often achieves short routines by toying with the solve order so that the harder pieces can be solved non-pure. For newer solvers this can be quite difficult. I'd hate for a new solver to look up a puzzle, see that it isn't on the list and then be unable to find a (5,1) sequence, only to learn that your expected solve order is quite complicated to avoid long sequences.
Without wanting to sound like a gruff puzzle bootcamp guy :lol: I don't see that as a problem. When I was new here I was cycling pieces one at a time, struggling to find efficient algos, and in awe of least move solves that have since been surpassed by miles. When I played around with 3.7.2 I was happy just to find a solution at all, which initially finished with pure [30,1] cycles. I then knocked that down to [14,1] and made a solve, and since then I've found [8,1] but not re-solved yet. I think it's unrealistic for newer solvers to expect to be competing for least move solves straight away. It can take time and experience of solving a good variety of puzzles before one develops a personal style of discovering efficient algos, a least moves "knack".

One aspect I find fascinating about the Gelatinbrain puzzles is that GB has effectively created and presented a wide variety of interactive logic-art. Solving is a form of appreciation, as is discussing solution methods; the rankings table is like a guest register; and the shortest time and movecount are invitations as much as challenges. "This is possible, care to try too?" Whenever you see a movecount for a puzzle that is waaaay under what you think you can do, it's a wordless hint. The number is saying: "You are missing something. Something is possible with this puzzle, and you have not spotted it yet. It could be sneakily hidden, or it could be right in front of your nose. What could it be?"


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Wed Jan 19, 2011 4:32 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Elwyn wrote:
I'm thinking Julian does the same or has one of the most incredible memories ever :lol:
I have one of the worst memories ever, especially short-term; with the exception of a few puzzles, if I use more than 2-3 setup moves I have to write them down or I mess up! Sad but true.

Elwyn wrote:
On a different note, Julian, there's something i've been meaning to ask you. You talk about adding set-up moves to algs, and change say a (3,1) to a (5,1) to help isolate a piece in a turning layer. I've always found it much easier to instead change the latter part of the commutator and change a (3,1) to a (3,3) which is the same amount of moves as a (5,1) but you never seem to do this and i was wondering why?
I got in the habit of building everything around a single move for the q part of [p,q]. I like having a clear visual picture of [this region has stayed the same] but [this has been swapped] at the midpoint. Doing sequences differently would confuse me. (By the way, it has belatedly occurred to me that by using rounded brackets all this time, I may have been committing and encouraging mathematical sin, pinching notation that more rightly belongs to sets, combinatorics, graphs, and coordinates. So from now on I will use [p,q] for commutators p q p' q' and [p:q] for conjugates p q p', as recommended by the nice people at speedsolving.com.)


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Wed Jan 19, 2011 5:46 pm 
Offline

Joined: Sat Jun 13, 2009 7:02 am
I just finished solving 3.6.4 (helicopter cube + skewb). Very nice puzzle, fun to solve. It was the first time I ever solved a puzzle of this type. It took me a while on my first solve though, 36 minutes. But I'll try a faster solve this weekend.

3.6.4

My solution:
1. Solve white edges.
2. Put in corners on white side paired up with 2 edges each. Some edges are going to be in incorrect orbits and need to be corrected. It did this with a skewb move as setup move and a helicopter move.
3. Orient corners on yellow side.
4. Permute corners with simple algo.
5. Permute edges that are already in correct orbits with commutators.
6. Using skewb moves as setup moves, permute last edges with same commutators as in step 5.
7. Pat yourself on the back.

All in all it wasn't a very difficult puzzle, but it was a new challenge for me.


Last edited by luke1984 on Thu Jan 20, 2011 6:09 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 2:38 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
3.1.27
Attachment:
3.1.27.png
3.1.27.png [ 5.44 KiB | Viewed 3705 times ]
I first solve the corners like a 2x2. Then I solve quarter-faces with a (3,1) until they go out. Then I solve the rest of the (1/8) face pieces with a (7,1). Julian, do you have a shorter alg. for the last step, it's not on your list.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 3:46 pm 
Offline
User avatar

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
Gelatinbrain, would it be possible to add pirsquared's/Eitan's new Celticube? It's got to be one of the strangest looking face-turning puzzles I've seen! At first it seemed to be vertex-turning, but then I realized it was rotating around that tiny little kite shaped piece, acting like the center of a face-turning puzzle :shock: Edit: never mind, Brandon is right :lol:
3.6.4
luke1984 wrote:
All in all it wasn't a very difficult puzzle, but it was a new challenge for me.
Well done. I don't really have a lot of steps for this puzzle. I solve the corners using intuition and a mix of Skewb and Helicopter Cube moves. Then I use a Helicopter Cube (3,1) cycle for the x-centers, which is easier to set up thanks to the addition of Skewb moves!


Last edited by Katja on Thu Jan 20, 2011 4:41 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 4:34 pm 
Offline
User avatar

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
Katja wrote:
Gelatinbrain, would it be possible to add pirsquared's/Eitan's new Celticube? It's got to be one of the strangest looking face-turning puzzles I've seen! At first it seemed to be vertex-turning, but then I realized it was rotating around that tiny little kite shaped piece, acting like the center of a face-turning puzzle :shock:
It sure is a cool looking puzzle. It doesn't maintain shape with each twist though and it has 24 axis of rotation which would keep it from fitting in with the puzzle types Gelatinbrain already supports.

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 4:35 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
Julian wrote:
So from now on I will use [p,q] for commutators p q p' q' and [p:q] for conjugates p q p', as recommended by the nice people at speedsolving.com.)

:shock: This is exactly what I have searched for a long time. I'm sure it's better to use short notation for conjugates. I take this and will use this from now on. [p,q] and [p:q]. The brackets make it clear. Thank you Julian.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 4:45 pm 
Offline
User avatar

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
Stefan Schwalbe wrote:
Julian wrote:
So from now on I will use [p,q] for commutators p q p' q' and [p:q] for conjugates p q p', as recommended by the nice people at speedsolving.com.)

:shock: This is exactly what I have searched for a long time. I'm sure it's better to use short notation for conjugates. I take this and will use this from now on. [p,q] and [p:q]. The brackets make it clear. Thank you Julian.
Yeah I'm glad to see this change too. It has bugged me that we've been using () rather than [] but that started long before I joined so I haven't said anything.

But then, we have been describing the |x| and |y| as (n,m) so we haven't been terribly strict with our notation.

I welcome the [n,m] and [n:m] change :D .

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 4:51 pm 
Offline
User avatar

Joined: Tue Feb 16, 2010 12:15 pm
Location: Sandnes, Norway
bmenrigh wrote:
I welcome the [n,m] and [n:m] change :D .
The only reason I don't use those characters is because I don't have them on my keyboard :lol:


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 4:58 pm 
Offline
User avatar

Joined: Wed May 13, 2009 4:58 pm
Location: Vancouver, Washington
Katja wrote:
Gelatinbrain, would it be possible to add pirsquared's/Eitan's new Celticube? It's got to be one of the strangest looking face-turning puzzles I've seen! At first it seemed to be vertex-turning, but then I realized it was rotating around that tiny little kite shaped piece, acting like the center of a face-turning puzzle :shock:
All of the puzzles so far are face, vertex, or edge turning platonic solids (Spheres & 2D excluded) so that would make implementing it a big step. But the hardest part about implementing the Celticube / Edge-Turning Rhombic Dodecahedron / Face-Turning Deltoidal Icositetrahedron is that they only jumble. With the exception of the spheres, nothing else jumbles and spheres are a weird kind of jumbling because they still use rational angles while the jumbling on these puzzles and the helicopter cube jumble at irrational angles.

_________________
Real name: Landon Kryger


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 5:14 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Stefan Schwalbe wrote:
3.1.27

I first solve the corners like a 2x2. Then I solve quarter-faces with a (3,1) until they go out. Then I solve the rest of the (1/8) face pieces with a (7,1). Julian, do you have a shorter alg. for the last step, it's not on your list.
I solved 3.1.27 exactly like you describe, but since then I have found a [6,1] pure cycle for the 1/8 pieces. Hint: the 6 moves are a well-known Rubik's algo.

Edits: 3.1.28-3.1.30 are x=5 puzzles, and 3.1.31-3.1.34 are x=6 puzzles. Thanks, Stefan, for your 3.1.28 & 3.1.29 solution outlines! :)


Last edited by Julian on Sun Jan 23, 2011 2:18 pm, edited 5 times in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 6:08 pm 
Offline

Joined: Sat Jun 13, 2009 7:02 am
Katja wrote:
Gelatinbrain, would it be possible to add pirsquared's/Eitan's new Celticube? It's got to be one of the strangest looking face-turning puzzles I've seen! At first it seemed to be vertex-turning, but then I realized it was rotating around that tiny little kite shaped piece, acting like the center of a face-turning puzzle :shock: Edit: never mind, Brandon is right :lol:
3.6.4
luke1984 wrote:
All in all it wasn't a very difficult puzzle, but it was a new challenge for me.
Well done. I don't really have a lot of steps for this puzzle. I solve the corners using intuition and a mix of Skewb and Helicopter Cube moves. Then I use a Helicopter Cube (3,1) cycle for the x-centers, which is easier to set up thanks to the addition of Skewb moves!


Thanks! I finished another 3.6.4 solve a few minutes ago, got a much better time and move count, 10min/211 moves. I used the same method as before.
Your solution sounds interesting as well, I'll give it a try sometime.

I also tried to solve the Super-X. I got pretty far by intuition, but solving the last two pieces gave me a headache! I'm sure I can figure it out though.

I still haven't tried any of the bigger puzzles, besides the dino dodecahedron, which is rediculously easy.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 6:14 pm 
Offline
User avatar

Joined: Thu Dec 31, 2009 8:54 pm
Location: Bay Area, California
luke1984 wrote:
I also tried to solve the Super-X. I got pretty far by intuition, but solving the last two pieces gave me a headache! I'm sure I can figure it out though.
Did you get down to just two triangle wedges swapped? If so you ran into an interesting parity :wink: . Here is a hint in invisible ink: "the parity happens when you have an odd number of 2x2x2 face quarter-turns. Turn one face a quarter turn and then re-solve all of the broken triangle wedges".
luke1984 wrote:
I still haven't tried any of the bigger puzzles, besides the dino dodecahedron, which is rediculously easy.
If you like the big vertex-turning puzzles you should give 1.2.16 a solve. You have to access it from the "File" menu in the applet.

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 6:28 pm 
Offline
User avatar

Joined: Thu May 31, 2007 7:13 pm
Location: Bruxelles, Belgium
GuiltyBystander wrote:
Katja wrote:
Gelatinbrain, would it be possible to add pirsquared's/Eitan's new Celticube? It's got to be one of the strangest looking face-turning puzzles I've seen! At first it seemed to be vertex-turning, but then I realized it was rotating around that tiny little kite shaped piece, acting like the center of a face-turning puzzle :shock:
All of the puzzles so far are face, vertex, or edge turning platonic solids (Spheres & 2D excluded) so that would make implementing it a big step. But the hardest part about implementing the Celticube / Edge-Turning Rhombic Dodecahedron / Face-Turning Deltoidal Icositetrahedron is that they only jumble. With the exception of the spheres, nothing else jumbles and spheres are a weird kind of jumbling because they still use rational angles while the jumbling on these puzzles and the helicopter cube jumble at irrational angles.


Without having a real puzzle, it's hard to imagine the behavior of these jumbling puzzles.
Another problem is the drawing of concave shapes.
Without using perspective and shading, drawing of these complex shape will not be decent. I don't want to waste time for graphical details. And it will considerably slow down the rendering speed. As I know, most graphic cards support the hardware acceleration with only full-screen mode.

3.9.* too have 24 axis, so I don't think this is a major obstacle.

_________________
Virtual Magic Polyhedra
Applet(Online)
Executable Jar Installer
Win32 Executable(Download)
troubleshooting


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Thu Jan 20, 2011 6:33 pm 
Offline
User avatar

Joined: Mon Dec 08, 2008 1:45 am
Location: New Zealand
Hi Gelatinbrain,

Is it possible to have a spherical version of Oskar's Mixup cube? Like a spherical 3x3x3, but the inner slices can turn at 45 degrees?

-Mark- :)

_________________
My Shapeways Shop!
Tony Fisher wrote:
A rare puzzle is one that is only lightly cooked.

Kelvin Stott wrote:
Squiggle is such a funny word to say out loud. Squiggle! :lol:

I am with Frank's Family


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sat Jan 22, 2011 12:19 pm 
Offline

Joined: Sat Jun 13, 2009 7:02 am
bmenrigh wrote:
luke1984 wrote:
I also tried to solve the Super-X. I got pretty far by intuition, but solving the last two pieces gave me a headache! I'm sure I can figure it out though.
Did you get down to just two triangle wedges swapped? If so you ran into an interesting parity :wink: . Here is a hint in invisible ink: "the parity happens when you have an odd number of 2x2x2 face quarter-turns. Turn one face a quarter turn and then re-solve all of the broken triangle wedges".
luke1984 wrote:
I still haven't tried any of the bigger puzzles, besides the dino dodecahedron, which is rediculously easy.
If you like the big vertex-turning puzzles you should give 1.2.16 a solve. You have to access it from the "File" menu in the applet.


Thanks! I'll give that a try.

1.2.16 looks like fun too. The dodecahedron equivalent of a lattice cube.

I've been playing with circle skewb II/3.2.10 a little. I created some commutators to cycle the center pieces, but I haven't solved the whole puzzle yet.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sat Jan 22, 2011 2:38 pm 
Offline
User avatar

Joined: Thu Jul 23, 2009 5:06 pm
Location: Berkeley, CA, USA
luke1984 wrote:
I've been playing with circle skewb II/3.2.10 a little. I created some commutators to cycle the center pieces, but I haven't solved the whole puzzle yet.


Luke, think more before you start solving 3.2.10. It is actually a puzzle that everyone is familiar with, in disguise.

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


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sat Jan 22, 2011 3:51 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
Here are the outlines to some puzzles, I have recently solved. My time was not good, but my move-count was ok. I often made use of x4 moves. (Some things are writen in invisilbe color.)
3.1.28
Attachment:
3.1.28.png
3.1.28.png [ 14.23 KiB | Viewed 3621 times ]

I and II: 3x3
III: reduce half of the III-pieces (triangles) to the corners of the face-centers, wich creates solved face-squares. Use [1,1]'s or even [1:1]'s, at least one setup move is necassary. : [B',R4]
IV: reduce the IV-pieces (wedges) to the rest of the III-pieces (triangles) to build the V-blocks, I use a [[1:1],[1,1]] wich can be shortened: highlite:
[[R:F][B,R4]] =
R,F,R', B,R4,B',R4, R,F',R', R4,B,R4,B' =
R,F,R', B,R4,B', R'3, F', R3, B,R4,B' =
[R:F], [B:R4], [R'3:F'], [B:R4] :shock:
the last [B:R4] is sometimes not necessary.

V: solve V blocks , use [1:1]'s or [1,1]'s, the same alg's as in step III

edit: (as Julian noticed)
III: solve all III-pieces (triangles)
IV: solve the IV-pieces (wedges)
V: no V


3.1.29
Attachment:
3.1.29.png
3.1.29.png [ 14.68 KiB | Viewed 3620 times ]

I-V: like 3.1.28
VI: [[1:1],1] pure : [[B:R4],L'2]


Last edited by Stef-n on Mon Jan 24, 2011 12:39 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Sun Jan 23, 2011 3:03 pm 
Offline
User avatar

Joined: Mon Jul 21, 2008 4:52 am
Location: Brighton, UK
Stefan, thanks for your 3.1.28 and 3.1.29 outlines. You saved me from wasting a lot of moves on the last (VI) pieces of of 3.1.29; without highlighting your algo I found my own [[1:1],1]. With both 3.1.28 and 3.1.29 I cycled all the pieces after solving the 3x3x3. I think that with these two puzzles it doesn't make much difference whether we build groups or not, because reducing probably saves around the same number of moves that need to be used to solve the groups at the end.

I used a [6,1] algo for the IV pieces but I have now seen that putting a setup and undo move around my algo for the VI pieces gives a [5,1] for the IV pieces, working similarly to your 12-move algo. The extra 2 moves I took per cycle roughly accounts for the difference in our 3.1.28 move counts, assuming we both took 17-18 cycles to solve the IV pieces.

The algo I used for the VI pieces of 3.1.29 cycles pieces from 3 adjacent faces, which makes setups quite easy and short. (Try 270 degrees for the 1st and 3rd moves.) I took 519 moves to solve the 3.1.28 pieces, was lucky and already had 11 VI pieces solved, then took 217 moves to finish in 736 moves overall.

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

3.1.15 -- I can't see a faster way to solve than your outline, apart from going into heavy depth like 3x3x3 methods, and doing something like solving the circle pieces of 2 opposite faces, then solving all the rest of the circle edges in a single algo, then solving all of the rest of the circle corners with a single algo. Definitely I can't see faster than solving all the circle pieces first, then finishing everything else with [3,1] cycles.

3.1.27 -- Something I tried with my last solve was to scoop up 11-12 of the 1/8 pieces for 2 adjacent 2x2x2 corners, then solve the rest of the 2x2x2 without disturbing those 2 corners. I don't know how many moves this saves, but I think it must save some moves, not only by solving some 1/8 pieces quickly, but also by increasing the percentage of remaining 1/8 pieces that are already solved or sitting in 3-cycles. Maybe it could help you get under 200 moves?


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Mon Jan 24, 2011 12:56 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
I was not pleased with my 3.1.29 solve, and I solved it again. Here are some statistics:
I (3x3): 54 moves, 2:00
II all triangles: 144 moves, 11:00 (3 moves, 14s per piece)
III pentagons: 286 moves, 19:45 (6 moves, 25s per piece)
IV outer wedges: 190 moves, 13:19 (4 moves, 17s per piece)
=674 moves, 46:04
Julian wrote:
I think that with these two puzzles it doesn't make much difference whether we build groups or not, because reducing probably saves around the same number of moves that need to be used to solve the groups at the end.
Yes, that's even better. I have added this to my outline.
Julian wrote:
3.1.27 -- Something I tried with my last solve was to scoop up 11-12 of the 1/8 pieces for 2 adjacent 2x2x2 corners, then solve the rest of the 2x2x2 without disturbing those 2 corners.
This is a good idea. I'm curious how it works.


Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Mon Jan 24, 2011 3:37 pm 
Offline

Joined: Sun Aug 29, 2010 1:56 pm
I complete the diff-cubes, and bring a
3.1.30outline, to share it with you.
Attachment:
3.1.30.png
3.1.30.png [ 13.77 KiB | Viewed 3666 times ]

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. : [f', [ U': b ]] or for blocks: [ [ r': d' ], U2] or similar alg's
III: pair up all triangle-pieces, with [1:[1:1]] : [d: [F' : U4] ]
IV: now it is reduced to a 3.1.28 with already solved 3x3 pieces, ( 3.1.28 - method more above on this page)

my last solve - statistics:
I (4x4): 146 moves
II (kites): 237 moves
III (pair up triangle-pieces): 173 moves
IV (solve triangle-pairs): 169 moves
V (solve wedge-pieces): 250 moves

Edit: (as Julian noticed) solve the faces (kites) first, and then the 4x4. For the wing parity look at the next after the next post and the following post (if you need algorithms).


Last edited by Stef-n on Fri Feb 04, 2011 1:25 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread
PostPosted: Mon Jan 24, 2011 4:52 pm 
Offline
User avatar

Joined: Thu May 31, 2007 7:13 pm
Location: Bruxelles, Belgium
I made an installer for executable jar.
viewtopic.php?f=8&t=19728&start=14
I tested it on my Windows/xp, but I'm not sure if it works also in other enviroments (linux, mac and newer versions of windows).
Please test it on your machine and tell me if you got a problem.
Thanks.

_________________
Virtual Magic Polyhedra
Applet(Online)
Executable Jar Installer
Win32 Executable(Download)
troubleshooting


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 3096 posts ]  Go to page Previous  1 ... 46, 47, 48, 49, 50, 51, 52 ... 62  Next

All times are UTC - 5 hours


Who is online

Users browsing this forum: No registered users and 7 guests


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

Search for:
Jump to:  

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