View unanswered posts  View active topics
Author 
Message 
Danny Devitt

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 02, 2011 3:04 pm 

Joined: Tue Mar 25, 2008 2:51 am Location: Malibu, California

RubixFreakGreg wrote: I've never seen the 1.1.45, maybe that's the new one, and it looks awesome!  Greg No, that got added a few days ago along with 1.1.44 and 1.2.17. But it does look very cool, with the nonplanar cuts.
_________________ I am taking a break from the forum. You can reach me by PM if needed.


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 02, 2011 3:24 pm 

Joined: Thu Jul 23, 2009 5:06 pm Location: Berkeley, CA, USA

Danny Devitt wrote: The stats page now says /362 puzzles, so apparently one got added. Anyone know which it is? I've been gone too long, so half the puzzles look new to me. I've been keeping track of new puzzles. The 362nd is 3.1.35. GB added it yesterday. Today he added the 363rd and 364th: 2.1.11 and 2.1.12.
_________________ Check out some virtual puzzles I created at http://nan.ma


Top 


Sjoerd

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Jan 03, 2011 3:27 am 

Joined: Mon Sep 29, 2008 9:41 am Location: Spijkenisse, the Netherlands

The new 1.1.45 is quite easy, itÂ´s just a starminx without any centers.
_________________
OlivĂ©r Nagy wrote: 43,252,003,274,489,856,000. Or the full number in Hungarian is: NegyvenhĂˇromtrilliĂłkĂ©tszĂˇzĂ¶tvenkĂ©tbilliĂˇrdhĂˇrombilliĂłkĂ©tszĂˇzhetvennĂ©gymiliĂˇrdnĂ©gyszĂˇznyolcvankilencmilliĂłnyolcszĂˇzĂ¶tvenhatezer )


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Jan 03, 2011 12:49 pm 

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

1.1.44 (slices only 1.1.8 / Megaminx + Pentultimate)
A tricky puzzle! The setups take some getting used to.
Solution Outline
1. Solve 5 corners around a face intuitively. 2. Solve the other corners with (3,1) commutators. 3. Cycle the wide triangles with (3,1) commutators. 4. Cycle the centers with (3,1) commutators. 5. Cycle the thin triangles pure with (6,1) commutators.
I took 14 + 115 + 307 + 41 + 469 = 946 moves.


Top 


Stefn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Jan 03, 2011 4:16 pm 

Joined: Sun Aug 29, 2010 1:56 pm

I was solving 5.1.19 in August 10 before and today I was fighting with it again, and I only slowly remembered my solution steps and the movesequences. My approach differs a bit from Julians solution, wich I did not fully understood. Sjoerd wrote: The new 1.1.45 is quite easy, itÂ´s just a starminx without any centers. Thank you Sjoerd, now even the most stupid guy knows how to solve the small pieces. Yesterday I developed extra a new alg. for it, it was 14 moves long. I wasn't really proud on it, but I used it to solve the puzzle. A (3,1) does it too. schuma wrote: I've been keeping track of new puzzles. The 362nd is 3.1.35. GB added it yesterday. Today he added the 363rd and 364th: 2.1.11 and 2.1.12. Thank you for the notification and thank you for the new puzzles. It is impossible to solve all the puzzles. They faster grow than I can solve. I will continue with some 3.1.x puzzles wich I didn't solve before.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Jan 03, 2011 4:52 pm 

Joined: Thu Dec 31, 2009 8:54 pm Location: Bay Area, California

Stefan Schwalbe wrote: I was solving 5.1.19 in August 10 before and today I was fighting with it again, and I only slowly remembered my solution steps and the movesequences. My approach differs a bit from Julians solution, wich I did not fully understood. I gave Julian's 5.1.19 a testdrive, just pairing up the edges and solving the HM Pyramid portion of it but I ran into a center 3cycle rather than a 22 swap. As far as I know this isn't possible on the HM Pyramid and I never did figure out how to resolve it on the the couple 5.1.x tetrahedra it happens on. I need to think about the issue some more before I work on 5.1.19 again. Stefan Schwalbe wrote: Sjoerd wrote: The new 1.1.45 is quite easy, itÂ´s just a starminx without any centers. Thank you Sjoerd, now even the most stupid guy knows how to solve the small pieces. Yesterday I developed extra a new alg. for it, it was 14 moves long. I wasn't really proud on it, but I used it to solve the puzzle. A (3,1) does it too. Kinda like my (18,1) and (16,1) commutators for the sliceonly 1.1.8 that Julian has a (3,1) and (6,1) for. You and Julian (and schuma and Michael and ...) make me feel like an idiot sometimes. I tried really hard to get a good move count on 1.1.45 and Julian beat me by 10 moves (and 40 minutes faster!) Yesterday I solved 3.2.13 thinking that my reduction to a Dino Cube would beat Julian but nope, I didn't even come close. Stefan Schwalbe wrote: schuma wrote: I've been keeping track of new puzzles. The 362nd is 3.1.35. GB added it yesterday. Today he added the 363rd and 364th: 2.1.11 and 2.1.12. Thank you for the notification and thank you for the new puzzles. It is impossible to solve all the puzzles. They faster grow than I can solve. I will continue with some 3.1.x puzzles wich I didn't solve before. I love all the new puzzles! I'm impressed schuma keeps up so well. I still have no idea for 3.1.35 but schuma got it. It looks like a good candidate for reduction to a 3x3x3 cube but I can't cycle the edge wings or the 4 cornerwedges in the center of each face. I'm encouraged by how helpful and friendly everyone is here though. Even if I can't put up very competitive solves at least this thread is a good community .
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Stefn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Jan 03, 2011 5:32 pm 

Joined: Sun Aug 29, 2010 1:56 pm

bmenrigh wrote: Kinda like my (18,1) and (16,1) commutators for the sliceonly 1.1.8 that Julian has a (3,1) and (6,1) for. You and Julian (and schuma and Michael and ...) make me feel like an idiot sometimes. I tried really hard to get a good move count on 1.1.45 and Julian beat me by 10 moves (and 40 minutes faster!) Yesterday I solved 3.2.13 thinking that my reduction to a Dino Cube would beat Julian but nope, I didn't even come close. I tell you, you must have the right solving order, short movesequences, and a portion of good luck, wich make the records. Today I wondered about my own 5.1.19 result. I was not able to reach it. Until I refound the trick, a 14 movesequence wich cycled 3 pieces pure was not necessary and could be replaced by a dirtier (1,1). And that alone brought the movecount down and I could reach my previous.


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Jan 03, 2011 8:10 pm 

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

bmenrigh wrote: I gave Julian's 5.1.19 a testdrive, just pairing up the edges and solving the HM Pyramid portion of it but I ran into a center 3cycle rather than a 22 swap. As far as I know this isn't possible on the HM Pyramid and I never did figure out how to resolve it on the the couple 5.1.x tetrahedra it happens on. I need to think about the issue some more before I work on 5.1.19 again. The center 3cycle is what I meant when I said in my summary: "You might come across an orientation issue where you need to make a single vertex move and then fix the broken edges at the end of the stage. To fix broken edges, three type 2 pieces can be cycled with a (5,1) commutator (5.1.9 knowledge with a tweak)." I probably should have worded it better, but I was thinking of it as an overall orientation issue: "Oops, I'm solving the puzzle 120 degrees twisted!" bmenrigh wrote: I tried really hard to get a good move count on 1.1.45 and Julian beat me by 10 moves (and 40 minutes faster!) I'd say being within 10 moves is a tie, but we are probably just keeping the medal positions warm until Elwyn or Michael solves it, because they solved 1.1.5 quicker than we solved 1.1.45, and 1.1.5 has centers too! 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! I made some intuitive moves then 5move algos to swap the 2sticker pieces, (3,1) algos to swap the squares, then solved the reduced Dino Cube, in 48 + 80 + 12 = 140 moves.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Jan 03, 2011 8:30 pm 

Joined: Thu Dec 31, 2009 8:54 pm Location: Bay Area, California

Julian wrote: I probably should have worded it better, but I was thinking of it as an overall orientation issue: "Oops, I'm solving the puzzle 120 degrees twisted!" I figured that's what you meant. Sadly, I don't know how to handle that situation. When I solved 5.1.9 I just hit scramble when I ran into it and resolved hoping to avoid it. I'm holding myself to a higher standard of understanding now though so I won't accept "scramble and try again" as an algorithm. I need to revisit 5.1.9. Julian wrote: bmenrigh wrote: I tried really hard to get a good move count on 1.1.45 and Julian beat me by 10 moves (and 40 minutes faster!) I'd say being within 10 moves is a tie, but we are probably just keeping the medal positions warm until Elwyn or Michael solves it, because they solved 1.1.5 quicker than we solved 1.1.45, and 1.1.5 has centers too! Yeah I took a peek at their move counts and was surprised that they beat us AND solved the centers. I'm sure they are beating us on the edgesolving phase. I used 128 moves even though I was trying to be efficient. I was as efficient as I could be on on the the star points / petal cycling though. Michael did 2.2.11b in 65 moves which is CRAZY and would suggest that's where the win is. 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! I made some intuitive moves then 5move algos to swap the 2sticker pieces, (3,1) algos to swap the squares, then solved the reduced Dino Cube, in 48 + 80 + 12 = 140 moves. Yeah I used a 5move routine to cycle 3 2color pieces too. Just a (1,3) where you leave off the last 3 moves. I got super lucky though because without any setup moves I was able to solve 3 pairs in the first 5 moves, then with 2 setup moves I solved another 3 pairs. I'm sure I paired up the edges in much less than 48 moves. My Dino solve was probably ~22 moves so your 12 moves is a huge improvement. Your big win is in the square cycling. I used a (5,1). Now I'm going to have to search for that (3,1). Edit: Also, take a look at 3.3.13. I was pretty proud of my count until you beat it. I was overly liberal with my use of setup moves though so I think that could pretty easily account for the ~30 move difference. I used a (5,3) for the circlebitten + centers. boublez got a low enough score though that I think there must be a more efficient method.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Jan 04, 2011 4:23 pm 

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

bmenrigh wrote: Sadly, I don't know how to handle that situation. When I solved 5.1.9 I just hit scramble when I ran into it and resolved hoping to avoid it. A 3cycle of centers can't happen with 5.1.9 when the corners are solved, just as a Skewb or Master Skewb cannot have the corners of one orbital solved with a 3cycle of corners needed in the other orbital. You are probably thinking of 5.1.6 and/or 5.1.7, where the centers can be moved around independently from the corners. It's a single move fix with 5.1.19. I have improved the wording of that part of my outline. bmenrigh wrote: 1.1.45 ... I'm sure [Michael and Elwyn] are beating us on the edgesolving phase. I used 128 moves even though I was trying to be efficient. I was as efficient as I could be on on the the star points / petal cycling though. Same here! I took 118 moves on the edges, which is our 10move difference comes from. bmenrigh wrote: Edit: Also, take a look at 3.3.13. I was pretty proud of my count until you beat it. I was overly liberal with my use of setup moves though so I think that could pretty easily account for the ~30 move difference. I used a (5,3) for the circlebitten + centers. boublez got a low enough score though that I think there must be a more efficient method. I can't remember how I solved 3.3.13! I'll have to try it again sometime. Edit: I managed to squeeze a new LM record for 3.3.13 by 1 move (334). My weakness is solving the 3.3.9 pieces to start with, which took me 132 moves. Then I cycled the thin pairs of circle pieces (5,1) non pure, the outer centers (3,1) non pure, and finally the centers pure (5,1). I need to try the first stage a few times until I do it sub100, then when I continue I'll have a chance at sub300 overall.
Last edited by Julian on Tue Jan 04, 2011 5:41 pm, edited 1 time in total.


Top 


Stefn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Jan 04, 2011 5:31 pm 

Joined: Sun Aug 29, 2010 1:56 pm

my aproach for 5.1.19 is Attachment:
5.1.19.png [ 10.74 KiB  Viewed 3653 times ]
For your information: I corners + face parity (3 cycled faces are wrong) II (1,1) III face (12) IV (1,1) V (3,1) VI (3,1) VII (4,4) I needed For I, II, III: 29 For IV: 34 For V: 55 For VI: 46 For VII: 53 moves = 217 moves, wich hit the movecount record my time was about 36 min. This puzzle is a real challange. I dont think its optimal.


Top 


Sjoerd

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Jan 04, 2011 6:03 pm 

Joined: Mon Sep 29, 2008 9:41 am Location: Spijkenisse, the Netherlands

5.1.19 is much more difficult then 5.1.18. I still can't figure it out.. I DID solve 1.5.3, a mix between the starminx and the 'starminx II' (edge turning), and was a bit disappointed with it. The mix between the starminx and the 'starminx III' (corner turning) was much more difficult, since the vertex moves split the edges, so you first have to reduce the whole puzzle to a starminx. With 1.5.3 I did not need to make any edgeturning moves, it solved almost exactly like a regular starminx. The only difference being the edgeturning helps with setup moves, and I can imagine some parity problems may arise, but I didn't confront them Conclusion: 1.5.3 = StarminxI really miss the algorithm bar in the downloadable version. I would have used it for the starminx triangles, since it would be substantially easier then doing my (8,1) pure algorithm all the time. Also, on most combination puzzles, I used to figure out beforehand the algorithms for a certain puzzle. On combination puzzles, I mostly use a mix of known algorithms fron the 2 puzzles it's based on. For the 3.4.x series I could reuse a lot of the same found algorithms over and over again. The algorithm bar is very useful, since these combination algorithms are mostly very long. And I don't care about move count anyway There was just one puzzle I ever needed to be careful with, the circlestarminx and family. I came up for a couple of great (read: lengthy) pure algorithms for the inner circle pieces. Unfotunately the move counter only went to 4 digits, and when I came to move 9999, I couldn't turn anymore! Now that 'I'm back', I really wish I could use the algorithm bar...
_________________
OlivĂ©r Nagy wrote: 43,252,003,274,489,856,000. Or the full number in Hungarian is: NegyvenhĂˇromtrilliĂłkĂ©tszĂˇzĂ¶tvenkĂ©tbilliĂˇrdhĂˇrombilliĂłkĂ©tszĂˇzhetvennĂ©gymiliĂˇrdnĂ©gyszĂˇznyolcvankilencmilliĂłnyolcszĂˇzĂ¶tvenhatezer )


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Jan 04, 2011 6:19 pm 

Joined: Thu Dec 31, 2009 8:54 pm Location: Bay Area, California

I find the algorithm bar quite useful for exploring puzzles. I'm thinking I might start using macros during solving. I'm pretty sure you can get the applet version working by tweaking Java's Security. See http://twistypuzzles.com/forum/viewtopic.php?f=8&t=7830&start=2067. Just be warned that you shouldn't browse the web with that setting  just Gelatinbrain's applet. Once you set that setting, close out your browser, etc, give the Applet at least 5 minutes to load (seems to be a Java bug).
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


gelatinbrain

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Jan 04, 2011 6:33 pm 

Joined: Thu May 31, 2007 7:13 pm Location: Bruxelles, Belgium

Sjoerd wrote: Now that 'I'm back', I really wish I could use the algorithm bar... Your request is registered, but unfortunately I have no time for a while. It's more complicated than adding puzzles. You'd better troubleshoot the java version. I don't know why it doesn't work on your computer, but you should have changed something since it was working before. Post more details on this thread. Then maybe I can help you... Just added some more overlapping cubes(3.1.36...). This line is still a terra incongnita.
_________________ Virtual Magic Polyhedra Applet(Online) Executable Jar Installer Win32 Executable(Download) troubleshooting


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Jan 04, 2011 7:18 pm 

Joined: Thu Jul 23, 2009 5:06 pm Location: Berkeley, CA, USA

Sjoerd wrote: Conclusion: 1.5.3 = Starminx
Interestingly, I treated 1.5.3 as 1.4.2, using the edge turns in the core commutator algorithms, and the face turns only in the setup moves. So my conclusion is, 1.5.3 = 1.4.2. Sjoerd wrote: There was just one puzzle I ever needed to be careful with, the circlestarminx and family. I came up for a couple of great (read: lengthy) pure algorithms for the inner circle pieces. Unfotunately the move counter only went to 4 digits, and when I came to move 9999, I couldn't turn anymore!
In the webpage applet version, the limit of 9999 moves was removed several months ago. According to the stats page, Johannes Philp solved 1.1.4 using 67244 moves! You might want to try the applet version for this reason. gelatinbrain wrote: Just added some more overlapping cubes(3.1.36...). This line is still a terra incongnita. Thank you for these cubes! These are exactly what I was thinking these two days, when I found that 3.1.35 is essentially a 5x5x5, with the new definitions like (new R) = (1st and 2nd layer R) and (new R&2) = (2nd, 3rd and 4th layer R). One should note that 3.1.38 can be solved as a regular 5x5x5, because a single layer U can be realized by D,U'&2 followed by reorientation. According to this post, a 5x5x5 with (new R) = (1st and 3rd layer R) is also very interesting and related to 3.1.31 and 3.1.32, but harder than them. GB, could you add it? Thanks.
_________________ Check out some virtual puzzles I created at http://nan.ma


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Jan 04, 2011 7:26 pm 

Joined: Thu Dec 31, 2009 8:54 pm Location: Bay Area, California

schuma wrote: Interestingly, I treated 1.5.3 as 1.4.2, using the edge turns in the core commutator algorithms, and the face turns only in the setup moves. So my conclusion is, 1.5.3 = 1.4.2. The orientation parity of 1.5.3 is tied to a edge twist. Once you resolve that parity by solving all of the edges like a standard Starminx / 1.4.2, the centers and Star points should be easily solvable exactly like a Starminx. I bet Michael and Elwyn could solve it in 350 moves. schuma wrote: Thank you for these cubes! These are exactly what I was thinking these two days, when I found that 3.1.35 is essentially a 5x5x5, with the new definitions like (new R) = (1st and 2nd layer R) and (new R&2) = (2nd, 3rd and 4th layer R).
One should note that 3.1.38 can be solved as a regular 5x5x5, because a single layer U can be realized by D,U'&2 followed by reorientation. Hmm, I tried this with 3.1.35 without any luck. I must have given up too quickly.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Jan 04, 2011 9:21 pm 

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

Stefan Schwalbe wrote: I needed For I, II, III: 29 For IV: 34 For V: 55 For VI: 46 For VII: 53 moves = 217 moves, wich hit the movecount record my time was about 36 min. Thanks for sharing your 5.1.19 method. Yes, it's a tough puzzle, because wherever you save moves, there seems to be no way to avoid paying a price later; every road has a toll station. The gamble of my method is that pairing the HM Pyramid edges and then solving 4 piece types at the same time outweighs the cost that the later algos are longer. Our methods probably work out about the same on average. [Edit: But yours could be more consistent, e.g. if both methods average 210 moves, over 2030 solves yours might cover 190230 and mine might cover 180240?] I got a lucky solve of 179 moves tonight: V & VII pieces paired: 29 (about average) I, III, V, VII: 10 (solved like 5.1.3 and had 1 in 12 luck; after finishing a face, with one twist the rest was solved) II: 30 (average  3 cycles) IV: 65 (lucky  only 4 cycles needed) VI: 45 (lucky  only 3 cycles needed) [Edit: This puzzle can be reduced further. When moving only the faces of 5.1.19, piece types IV and VI always stay together in groups of 3. The facesonly 5.1.19 is the equivalent of the Skewb Ultimate, which can be solved at the very end.
Last edited by Julian on Sat Jan 08, 2011 5:42 pm, edited 2 times in total.


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Jan 05, 2011 2:00 am 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

Since my name got mentioned i thought i should have a go at 1.1.45. I got 368. At the beginning i tried something different which both saved moves and wasted them (because i got greedy). I was trying to see how well solving some star points/petals with block building at the start would work. The first five (as in all the white ones) I solved whilst solving the first 5 edges. I am pretty sure this saved moves but i then got the next 10 (the ones closest to the white side) whilst solving the next 5 edges which took a few moves and turned out to be a complete failure due to them getting scrambled again when solving the rest of the edges. Due to that i can't say how many moves just the edges take but i know i can get just a few moves under 100 if i concentrate. bmenrigh wrote: Michael did 2.2.11b in 65 moves which is CRAZY and would suggest that's where the win is. Those are megaminx edges not pyraminx crystal edges. Very different edges that don't really relate to 1.1.45 An edges only pyraminx crystal might be a fun challenge to see how low people could get without it taking over an hour to solve like this puzzle did. All gelatinbrain would need to do is "uncurve" the cuts of this puzzle.
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Jan 05, 2011 3:13 am 

Joined: Thu Dec 31, 2009 8:54 pm Location: Bay Area, California

Elwyn wrote: Since my name got mentioned i thought i should have a go at 1.1.45. I got 368.
At the beginning i tried something different which both saved moves and wasted them (because i got greedy). I was trying to see how well solving some star points/petals with block building at the start would work. The first five (as in all the white ones) I solved whilst solving the first 5 edges. I am pretty sure this saved moves but i then got the next 10 (the ones closest to the white side) whilst solving the next 5 edges which took a few moves and turned out to be a complete failure due to them getting scrambled again when solving the rest of the edges.
Due to that i can't say how many moves just the edges take but i know i can get just a few moves under 100 if i concentrate. I tried that on 2.2.20 but for me it was definitely a netloss. I also ended up breaking a number of the pieces I placed. If you did the edges in ~100 moves (~30 better than me) then you did the petal pieces ~20 better than me. That's really good, I don't think I could improve that much even given days of practice. When I go for cycling pieces efficiently, I first try to fix 2 at a time with no setup moves and then when those are all done, 2 at a time with 1 setup move, then 2 at a time with 2 setups, and so on. I only get 3 at a time once or twice when I'm lucky rather than searching for ways to cycle 3 with very few setups. Do you go for 3 at a time instead? If so, how many setups to you usually use? Elwyn wrote: bmenrigh wrote: Michael did 2.2.11b in 65 moves which is CRAZY and would suggest that's where the win is. Those are megaminx edges not pyraminx crystal edges. Very different edges that don't really relate to 1.1.45 I knew that. I dunno why I remembered them as Pyraminx Crystal edges rather than Megaminx edges. It seems my idiot syndrome is flaring up again. Elwyn wrote: An edges only pyraminx crystal might be a fun challenge to see how low people could get without it taking over an hour to solve like this puzzle did. All gelatinbrain would need to do is "uncurve" the cuts of this puzzle. Yeah it would be the same style of cut as 1.1.12. It might be fun to see what you guys could do with it.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Jan 05, 2011 4:35 am 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

bmenrigh wrote: If you did the edges in ~100 moves (~30 better than me) then you did the petal pieces ~20 better than me. That's really good, I don't think I could improve that much even given days of practice.
When I go for cycling pieces efficiently, I first try to fix 2 at a time with no setup moves and then when those are all done, 2 at a time with 1 setup move, then 2 at a time with 2 setups, and so on. I only get 3 at a time once or twice when I'm lucky rather than searching for ways to cycle 3 with very few setups. Do you go for 3 at a time instead? If so, how many setups to you usually use?
I should probably mention i got more pure cycles than normal on this solve. I look for all possible perfect cycles, i have some kind of method for looking for them which works alright so i just look over the whole puzzle till i find one. Then as i mentioned before if i run out of perfect cycles (which normally happens earlier than it did in this solve) i look for a cycle that will solve 2 pieces and have the third piece that moves move to a place that will make a perfect cycle possible. It's a lot of looking and takes a couple more setups too. I'd say i average 23 setups per cycle. Rarely 1 but rarely more than 4, what alg do you use out of curiosity? You should resolve the starminx, i'm sure you could get sub 500, perhaps aim for julians count. You're definitely getting better at fewest moves. Also when i got half way through my 1.3.12 solve (reduction to starminx but had a blackout) I tried looking at what triangles moved when i solved the centers and cycling them in a way that if a triangle was next to the center it belonged with, it would stay with it and go to the correct face... hard, but i think it worked alright.
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Jan 05, 2011 8:29 am 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

I just got the record for the starminx (1.1.5) with a 357 move solve beating Michael by 25 moves and beating my 1.1.45 move count as well... So i'm guessing that could be brought down a few moves. It was the first starminx solve i have done in ages. 1.1.5 is a record i have wanted for a long time Pretty sure getting the first five triangles at the start and solving as many as possible whilst solving centers saves a few moves as i wasn't exceptionally lucky with perfect cycles this time. If anyone wants to know i had done all the edges along with the first 5 triangles (and the white center to go with them) at 92 moves.
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


gelatinbrain

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Jan 05, 2011 4:58 pm 

Joined: Thu May 31, 2007 7:13 pm Location: Bruxelles, Belgium

schuma wrote: One should note that 3.1.38 can be solved as a regular 5x5x5, because a single layer U can be realized by D,U'&2 followed by reorientation.
This is not what I meant. I meant that the slice is triple instead of double(equivalent to 3.1.35 + centers), so I modified. You have to resolve.
_________________ Virtual Magic Polyhedra Applet(Online) Executable Jar Installer Win32 Executable(Download) troubleshooting


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Jan 05, 2011 5:23 pm 

Joined: Thu Dec 31, 2009 8:54 pm Location: Bay Area, California

Elwyn wrote: [While discussing 1.1.5 / 1.1.45 star points..] What alg do you use out of curiosity? I use (R'2 L2 R2, l) where R and L are separated by one face and l is a slice about L. I use any of the 4 possibilities for the slice (l, l2, l', l'2) which allows for more flexibility in the third piece to be cycled. Of course I also use the inverse, mirror, and mirrored inverse for 16 total variations on the routine. I assume you use the same? Also, do you use something better than (4,1) for the centers? Also, congrats on your 1.1.5 record. I thought that record was untouchable. I bet Michael thought the same I'm going to give 1.1.5 another solve this weekend to see how I stack up.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Jan 05, 2011 7:21 pm 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

bmenrigh wrote: I use (R'2 L2 R2, l) where R and L are separated by one face and l is a slice about L. I use any of the 4 possibilities for the slice (l, l2, l', l'2) which allows for more flexibility in the third piece to be cycled. Of course I also use the inverse, mirror, and mirrored inverse for 16 total variations on the routine. I assume you use the same? Also, do you use something better than (4,1) for the centers?
No i use one i found myself. I may have made a longer than necessary post about it Here.... Not sure if i counted all the variants correctly (128 seems a little too many) but i am sure i don't use as many of them as i should to save setups. I probably only use about 8 as i never really got used to the positions that get cycled for the other ones. As for the centres i use a (3,1) because i actually find it easier to use than the (4,1). Using Julian's notation from ages ago (i hope you remember it, if not it's pretty Similar to normal cube notation U is up D is down and R is on the right adjacent to the U face) It is (R U R', D) where D can have 4 variants and U can have 4 variants and R can be either R or R'. The mirrored version using L is just actually the same as using R' instead off R to start with but there is all the inverses to count as well. As i said i find it much easier to use than the (4,1). If you want it in gelatin notation it's C, B, C', K, C, B', C', K',
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


Katja

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Thu Jan 06, 2011 8:55 am 

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

1.2.6 Attachment:
Skjermbilde 20110106 kl. 14.42.05.png [ 25.63 KiB  Viewed 3482 times ]
1.) No algorithm; the center pieces can simply be placed using intuition. 2.) I 3cycled the thin triangles using a nonpure (4,1). This was the hardest step in my opinion as you constantly have to keep a close eye on the centers in case you accidentally do the algo wrong or reverse the setups wrong. I spent two hours on this step, and this is the first puzzle I've ever had to write down the setup moves on. 3.) Lastly I cycled the wide triangles pure using a (4,1). This was far easier and only took about 58 minutes. I also came up with all the algorithms to solve 1.2.7 with not too long ago, and that might be the next puzzle I solve.
Last edited by Katja on Thu Jan 06, 2011 10:17 am, edited 1 time in total.


Top 


Katja

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Thu Jan 06, 2011 10:12 am 

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

schuma wrote: Sjoerd wrote: Conclusion: 1.5.3 = Starminx
Interestingly, I treated 1.5.3 as 1.4.2, using the edge turns in the core commutator algorithms, and the face turns only in the setup moves. So my conclusion is, 1.5.3 = 1.4.2. I would have to disagree with both of you I solved it using about the same amount of edge turns and face turns. Finding short algorithms is quite easy this way; I found a pretty useful pure (3,1) for the triangles. However, it did give me a pretty bad move count, but I really solved it this way for the fun of it and it might be one of my favorite puzzles on GB. So, my conclusion: 1.5.3 = 1.1.5 + 1.4.2


Top 


alaskajoe

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Thu Jan 06, 2011 10:59 am 

Joined: Mon Feb 06, 2006 12:52 am

Hm This might be somewhere in this thread already but is there a chance the programm doesn't update some solves? I solved the pentultimate and another of the dodecahdra the first time a couple of days ago. Yesterday I solved yet another one and the time is already there. I have been waiting for the Pentultimate solve for days but it doesn't come. This took soo long solving!! No I am not really mad. I am glad I found back to this. Man had I gotten out of shape. It took so long for even only one of those easy new face turning icosahedron with the curvy cuts like 2.1.0 I am almost back to old power by now though. It might even be over 9000 again. ... Starts solving Pentultimate and 1.1.4 a second time...
_________________ Life is simple. eat. sleep. train. repeat.


Top 


alaskajoe

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Thu Jan 06, 2011 11:53 am 

Joined: Mon Feb 06, 2006 12:52 am

Ha. Pentultimate solved in 52 minutes. (only because I am eating spaghetti meanwhile. )
_________________ Life is simple. eat. sleep. train. repeat.


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Jan 07, 2011 3:43 pm 

Joined: Thu Jul 23, 2009 5:06 pm Location: Berkeley, CA, USA

Recently I've found four hidden puzzles in the web applet that are not yet numbered. (edit: now they are numbered. )I solved them to make the stat 374/370. One of them is the beautiful Curvy Copter (I) by TomZ. Look how pretty it is: Attachment:
curvy.PNG [ 10.34 KiB  Viewed 3407 times ]
How to find them: Go to File menu > new > cubes > edge turn. After 3.3.18, the next two empty slots are actually working links to two unnumbered hidden puzzles (3.3.19 and 3.3.20??). The first one is a circle version of little chop, which has less pieces than 3.3.10. The second one is TomZ's Curvy Copter. Attachment:
33.PNG [ 56.61 KiB  Viewed 3407 times ]
Go to File menu > new > cubes > edge turn. Between 3.4.20 and 3.4.22 there is an empty slot (3.4.21??). It's similar to 3.4.20. After 3.4.24, there is another empty slot (3.4.25??). It's a circle 2x2x2 + master skewb. Attachment:
34.PNG [ 111.3 KiB  Viewed 3407 times ]
_________________ Check out some virtual puzzles I created at http://nan.ma
Last edited by schuma on Fri Jan 07, 2011 5:42 pm, edited 1 time in total.


Top 


Stefn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Jan 07, 2011 4:17 pm 

Joined: Sun Aug 29, 2010 1:56 pm

regarding 3.1.2b, the pieces are: Attachment:
3.1.2b.pc.png [ 18.21 KiB  Viewed 3399 times ]
I used a 3x3 beginner method.


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Jan 07, 2011 5:48 pm 

Joined: Thu Jul 23, 2009 5:06 pm Location: Berkeley, CA, USA

Elwyn wrote: An edges only pyraminx crystal might be a fun challenge to see how low people could get without it taking over an hour to solve like this puzzle did. All gelatinbrain would need to do is "uncurve" the cuts of this puzzle. GB just added the edgeonly pyraminx crystal: 1.1.46, together with 1.1.47~49, some other nonplanar cut dodecahedra.
_________________ Check out some virtual puzzles I created at http://nan.ma


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Jan 07, 2011 5:59 pm 

Joined: Thu Dec 31, 2009 8:54 pm Location: Bay Area, California

schuma wrote: GB just added the edgeonly pyraminx crystal: 1.1.46, together with 1.1.47~49, some other nonplanar cut dodecahedra. Oh whoa, these puzzles look awesome. 1.1.49 looks very innovative. Something an artist would design.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Jan 07, 2011 8:18 pm 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

Thank you gelatinbrain.
Brandon my first solve of 1.1.46 was 82 moves so that's about 40 moves less than the edges of 1.1.45 took you so this is probably what you should try to improve if you want better movecounts for puzzles with these edges. I started getting lower movecounts when i stopped thinking about them as having a last layer and just started cycling them 2 at a time near the end of the solve with setup moves if that's any help to you.
Also 1.1.47 is a funny puzzle, Highlight for spoilers below.
it is the equivalent to a pyraminx crystal with starminx/pentultimate centres. Or another way to look at it is 1.1.4 without the triangles.
The curvy ones are a little scary but 1.1.48 is a nice looking puzzle.
For 1.1.48 spoilers highlight below It is just 1.1.2 with pentultimate/starminx centres... don't try and look for a three cycle for those tiny pieces like i did they are pyraminx crystal edges hahaha.
1.1.49 spoilers below as well.
the odd shaped pieces next to the centres seem to be an entirely new piece. A pure three cycle is hidden below, however i would solve them dirty before the centres with a (3,1) C, B, C', K, C, B', C', K',
And then solve the centres with a (8,1) because there are fewer of them to cycle.
F', C, F, C', F', C, F, C', I', C, F', C', F, C, F', C', F, I, is a pure three cycle for the weird pieces.
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


Sjoerd

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sat Jan 08, 2011 8:08 pm 

Joined: Mon Sep 29, 2008 9:41 am Location: Spijkenisse, the Netherlands

The new 1.1.4649 were quite fun, the used cuts are quite interesting, but the puzzles were quite easy to define. The only thing I really liked solvingwise were the new pieces on 1.1.49, but an algorithm was easily found for that. Solving them all one after another was quite repetitive, so I will move on to some real challenges Further on this weekend I will have a go at two very difficult puzzles: 1.4.6 and 1.1.40. I spent the last couple of evenings figuring out all of the algorithms, so my recipe is almost ready. The only difficulty I may have on 1.4.6 are the central kitepieces that are nearest to the edges (rather than the corners). I have not found an algorithm yet, so the recipe is not complete yet. They will be my first step in solving, so it doesn't need to be pure or anything. Maybe I'll try intuition! The next difficulty will be all of the setup moves... And 1.1.40 shouldn't be a problem. I found 10 and 14 move algs for the circle pieces, and I can solve 1.1.39 already, so I don't think I'll have any difficulties. 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
_________________
OlivĂ©r Nagy wrote: 43,252,003,274,489,856,000. Or the full number in Hungarian is: NegyvenhĂˇromtrilliĂłkĂ©tszĂˇzĂ¶tvenkĂ©tbilliĂˇrdhĂˇrombilliĂłkĂ©tszĂˇzhetvennĂ©gymiliĂˇrdnĂ©gyszĂˇznyolcvankilencmilliĂłnyolcszĂˇzĂ¶tvenhatezer )


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sat Jan 08, 2011 8:16 pm 

Joined: Thu Dec 31, 2009 8:54 pm Location: Bay Area, California

Sjoerd wrote: The new 1.1.4649 were quite fun, the used cuts are quite interesting, but the puzzles were quite easy to define. The only thing I really liked solvingwise were the new pieces on 1.1.49, but an algorithm was easily found for that. Hey Sjoerd, I've played with those new 1.1.49 pieces too and the best routine I have is (6,1) pure. The 6 is just a (1,1) with a setup+undo move. I have tried awfully hard to get that routine shorter without any luck. I'm thinking a better strategy may to be to solve the pieces nonpure and then solve the centers pure instead. Can you comment on what you ended up using?
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 12:51 am 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

bmenrigh wrote: I have tried awfully hard to get that routine shorter without any luck. I'm thinking a better strategy may to be to solve the pieces nonpure and then solve the centers pure instead. Can you comment on what you ended up using? Great minds think alike? If you bother highlighting my 1.1.49 "spoilers" i suggested the exact same thing. It seems by far the better approach.
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 1:22 am 

Joined: Thu Dec 31, 2009 8:54 pm Location: Bay Area, California

Elwyn wrote: bmenrigh wrote: I have tried awfully hard to get that routine shorter without any luck. I'm thinking a better strategy may to be to solve the pieces nonpure and then solve the centers pure instead. Can you comment on what you ended up using? Great minds think alike? If you bother highlighting my 1.1.49 "spoilers" i suggested the exact same thing. It seems by far the better approach. Dang, I don't know why I didn't notice your spoiler. Did you add them later in an edit? I think your idea is a better strategy. Solve 12 pieces inefficiently rather than 60 pieces that way. I had the same piece recognition problem on 1.1.49 as I had on 1.1.17. Something about all of the triangles being nearly the same size makes it hard for me to see which triangles come together to form a single piece and which belong to separate pieces. I tried hard to get a good move count for 1.1.49 but my Megaminx solves are terrible (300+). I never learned any F2L style solving for the Rubik's cube so I solve a Megaminx in a rather naive way. My method is so bad that I really think I could solve a Megaminx much more efficiently if I were to first place all of the edges and then 3cycle corners + orient them second. I gave 1.1.46 a solve and got 125 or 126 moves. I think my problem is that I'm still thinking about them as 1.1.3 edges. Your suggestion of solving the first 510 intuitively and then efficiently placing 23 at a time using 3cycles for the rest is a good idea. I know I can break 100 with this method. 1.1.49 is a beautiful puzzle. I'm wondering how deep those loop cuts can be taken. What would happen if they extended past the corner and into the other side's domain? Can we get a deeperthanPentultimate puzzle using a trick like that or does the geometry not really support that depth of cut? Gelatin brain, have you discovered a whole new class of puzzle similar to circle puzzles? I'm looking forward to what evil things you have in store for us next .
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 1:48 am 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

bmenrigh wrote: I never learned any F2L style solving for the Rubik's cube so I solve a Megaminx in a rather naive way. Wow, i really suggest learning intuitive F2L on a normal cube. It helps so much on puzzles like the megaminx and really got me used to block building in general. bmenrigh wrote: Your suggestion of solving the first 510 intuitively and then efficiently placing 23 at a time using 3cycles for the rest is a good idea. I usually go for more like 15 intuitively and also don't expect many perfect three cycles, they're a rare occurrence with these edges. It's also annoying if a piece is in the correct spot but misoriented or two edges are in each others spots as then you have to settle for solving just 1 in a cycle. bmenrigh wrote: I'm wondering how deep those loop cuts can be taken. What would happen if they extended past the corner and into the other side's domain? I'd say something like 3.1.31 would happen if the cuts extended past the deep cut of the solid as that is what has happened there. Strange puzzles indeed. Oh and as for cycling the centres pure on 1.1.49 i have a (4,3) which is equivalent move wise to a (6,1) but can't find anything shorter, can you? Actually i have a normal (6,1) as well and now a (6,1) for the weird pieces pure but cent find anything shorter. I also have a 14 move pure cycle for the 1.1.44 (slice only 1.1.8) thin triangles which is the same length as Julian's In fact i now have the same length algs for every step... not sure if i want to solve it yet, it looks hard
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 4:18 am 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

In regards to 1.1.35 Julian said Julian wrote: I tried the method schuma describes before I solved 1.1.35 but I couldn't work out a short algo that way. Here are hints to make a (10,1) pure algo for the circle pieces: I hadn't seen Schuma's post about the method but i just was doing something similar and came up with a (8,1) pure 3 cycle for the circle pieces and then went looking to see how short your alg was. Now it's one of the rare occasions i have a shorter alg than you I'm starting to think hiding algs like this isn't really necessary because if written in gelatinbrain notation it's not like anyone who doesn't really want to know it is going to glance at it accidentally and instantly see what it is/how it works so here it is A, B2, A', K, B', C, B'2, C', K', C, B2, C', K', B, A, B'2, A', B, It's odd as after the alg a whole puzzle rotation/two different 5 cycles on the top and bottom faces has occurred however this doesn't matter at all on this puzzle, on 1.1.35B/C however this would be a problem. This can be gotten around with some (rather funny) trickery to form an alg of the same move length A, B2, A', K, B', C, B'2, C', K', C, B2, C', K', B, A, B'2, A', K, Possibly the funniest case of moves canceling i've had since it was accidental that i found it I was going to just add two moves to the end of the alg to undo the problem but then realised this would involve undoing the last move, why i didn't find this alg first i do not know . So there's a pure (8,1) for the circle pieces of all colour variants of 1.1.35. It's a strange alg indeed. I thought about just posting the final alg but thought my redundantly complex way of finding it was amusing. So Julian, another puzzle (or three if you count colour variants) crossed off your list of puzzles requiring commutators longer than (8,1) Edit: I changed the orientation of the alg so it is easier to perform. Also thought i'd add the K which is the 1 in the (8,1) can be any of the 4 possible turns of that side k, k', ect and the B2 can be B'2 and vice versa to cycle from different positions.
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42
Last edited by Elwyn on Sun Jan 09, 2011 5:29 am, edited 1 time in total.


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 4:43 am 

Joined: Thu Jul 23, 2009 5:06 pm Location: Berkeley, CA, USA

Elwyn wrote: bmenrigh wrote: I'm wondering how deep those loop cuts can be taken. What would happen if they extended past the corner and into the other side's domain? I'd say something like 3.1.31 would happen if the cuts extended past the deep cut of the solid as that is what has happened there. Strange puzzles indeed. I think this is not the only way to get "superdeep" cuts. Bandaging several layers also leads to something equivalent. For example, in 1.1.8, if I define a new type of face turn that affects the first layer of that face together with the second layer of the opposite face (like A,L'&2), its effect goes farther than the the pentultimate. So it's "superdeep". There are many new types of pieces, which cannot be found in any planarcut puzzles or circle puzzles. Just like the pentultimate, every turn affects half of all the pieces. Another example is, in Megaminx, if each face is bandaged to its opposite face (like A,L'), then the puzzle is essentially a slice only Megaminx. I haven't seen any discussion about it yet and I don't know how hard it is to solve. Of course one can think of opposite faces rotating in different direction, (like A,L). Oskar's Gear Cube behaves as this, thanks to a differential gear mechanism. I expect the Megaminx analog to be solved just like a regular Megaminx, because when you solve one hemisphere, the other is solved automatically by symmetry. There are so many ways to bandage layers (like in 1.1.10, bandaging first + second layer of one face and the first layer of the opposite face) or to take slice only puzzles (like sliceonly Pyraminx Crystal, sliceonly Starminx etc). All of the possibilities provides new pieces and new challenges. The pieces are not tiny, so the puzzle looks still OK. The only question is, along this endless road, when to stop.
_________________ Check out some virtual puzzles I created at http://nan.ma


Top 


Katja

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 6:58 am 

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

Elwyn wrote: I hadn't seen Schuma's post about the method but i just was doing something similar and came up with a (8,1) pure 3 cycle for the circle pieces (8,1)? I spent forever finding a pure (10,1) for those pieces. Great find! Sjoerd wrote: And 1.1.40 shouldn't be a problem. I found 10 and 14 move algs for the circle pieces, and I can solve 1.1.39 already, so I don't think I'll have any difficulties. Nah, you sholdn't have too much trouble solving 1.1.40 with your track record I solved it a while back, when I had a lot less experience than I do now and it was a great feat for me. I'm certain you'll beat my time and move count too. I had a pretty short algo for one of the circle pieces, but I recall using a pretty long and clumsy for the other; the ones connected to the trapezoids which again is connected to the corners. I'm sure I could find a better one now if I wanted to, but at the moment I'm not really interested. As I mentioned in one of my above posts, I'm planing on solving 1.2.7  vertex turning Master Brilic  I have all the algorithms and surprisingly they're all pretty short. The longest one is an 18move pure cycle. I have attempted to solve it once, but one hour into the solve I got stuck. I'll try again today, but this is one hard puzzle! EDIT: Elwyn, I just noticed that you stole my time record for 1.1.47! No worries, I'll just steal it back Also, Gelatinbrain, I love the new dodecahedra, thanks a lot!


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 8:27 am 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

Katten wrote: Elwyn, I just noticed that you stole my time record for 1.1.47! No worries, I'll just steal it back Comments like this kind of bring out my competitive nature... so I just did a 5:21 solve. I don't normally go for time records though, purely because i know Michael could beat me for every single one if he wanted to
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 9:44 am 

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

1.1.35Elwyn wrote: A, B2, A', K, B', C, B'2, C', K', C, B2, C', K', B, A, B'2, A', K,
So there's a pure (8,1) for the circle pieces of all colour variants of 1.1.35. It's a strange alg indeed. Brilliant! I see your main sequence as (3 face moves) (simulated slice move) (inverse of 3 face moves). If shiftclick existed for 1.1.35x this conjugate would be 7 moves instead of 8. I tried to find such a slicing algo for quite a while, then I gave up and arrived at my (10,1), which was an (8,1) for a different circle puzzle with the main sequence sandwiched inside a single setup/undo. I doubted that I would be able to find better than ((3,2), 1) = (10,1) with simulated slices anyway... but I failed to take account of The Elwyn Factor. You can probably cut 100+ off my move counts for 1.1.35 and 1.1.35b with this algo. I will use it when I get around to 1.1.35c. By the way, thanks for reminding us of your tips for the wing pieces of 1.1.41.1.6 ( here). I wasn't solving much over the summer (northern hemispherically speaking) so it slipped under my radar. In addition to the clever algos, things like solving 2 pieces so the 3rd piece is set up for a perfect 3cycle the next time, and trying to cycle wings into the correct position at the same time as cycling the centers. They seem fairly obvious now that you've pointed them out, but they never occurred to me, and they give valuable insight into the wide variety of tools we need in our toolbox for really good least move solving. Elwyn wrote: So Julian, another puzzle (or three if you count colour variants) crossed off your list of puzzles requiring commutators longer than (8,1) Yes, there aren't many left. When the next page starts I will do an update. Algo lengths are shrinking so much, I should probably switch to listing puzzles believed to require longer than ( 6,1) rather than (8,1)!
Last edited by Julian on Sun Jan 09, 2011 9:54 am, edited 1 time in total.


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 9:51 am 

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

1.1.49 Elwyn wrote: bmenrigh wrote: I'm thinking a better strategy may to be to solve the pieces nonpure and then solve the centers pure instead. Can you comment on what you ended up using? It seems by far the better approach. Yes, and the centers can be solved pure (6,1). The 6 move sequence is a (1,1) commutator of face moves enclosed in a slice setup and undo: (B&2 (F',C) B'&2, J).


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 12:33 pm 

Joined: Tue Sep 08, 2009 8:41 am Location: The Blue Mountains, Australia

Julian wrote: Yes, and the centers can be solved pure (6,1). The 6 move sequence is a (1,1) commutator of face moves enclosed in a slice setup and undo: (B&2 (F',C) B'&2, J). Mine doesn't use a slice A, C, B2, C', B'2, A', K, A, B2, C, B'2, C', A', K', It's similar to the (6,1) for pentultimate corners but requires the third move to be of 144 degrees rather than 72. Julian wrote: You can probably cut 100+ off my move counts for 1.1.35 and 1.1.35b with this algo. I had a look at your record and it really looks like i'd need a lot of perfect three cycles to beat it by that much, that move count is a lot better than i realised Julian wrote: By the way, thanks for reminding us of your tips for the wing pieces of 1.1.41.1.6 (here). That link goes to me explaining how many variants of my three cycle there are... I think you might meant this post.
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 1:52 pm 

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

Elwyn wrote: Julian wrote: Yes, and the centers can be solved pure (6,1). The 6 move sequence is a (1,1) commutator of face moves enclosed in a slice setup and undo: (B&2 (F',C) B'&2, J). Mine doesn't use a slice A, C, B2, C', B'2, A', K, A, B2, C, B'2, C', A', K', Excellent! To minimize setups, if the centers to be cycled are spread out I'll use your algo, whereas if they are close together, I'll use mine. I posted mine for info because your invisible ink hint algo was (and currently still is) an (8,1). Elwyn wrote: Julian wrote: By the way, thanks for reminding us of your tips for the wing pieces of 1.1.41.1.6 (here). That link goes to me explaining how many variants of my three cycle there are... I think you might meant this post. I should have given both links, because I meant both: the clever algos, and the extra tips. Hand and glove: the shortest possible and most varied/flexible algos, and being as efficient as possible using those algos to cycle pieces.
Last edited by Julian on Sun Jan 09, 2011 3:02 pm, edited 1 time in total.


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 2:33 pm 

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

5.1.19 Stefan  I think your solution has the best order when solving by piece type, and probably has optimal algos too. Full reduction does save moves, but not that many. I estimate it averages a total of 195 moves: 30 moves each for the edges and centers, 50 moves each for the small center pieces, then around 35 moves to solve the... Skewb Ultimate. Just a couple of days ago I suddenly noticed what the 5.1.19 with vertex turns only is. So I pulled my Skewb Ultimate off the shelf and started trying to improve my solution! I reduced 5.1.19 to a Skewb Ultimate in 151 moves and then solved the SU in 27 moves for a total of 178 moves. On average 3 out of the 12 pairs of circle pieces (IV in your diagram) belong to the correct Skewb Ultimate corner (tetrahedron face center), which saves a cycle or two. We also save moves by not having to undo face setup sequences, and sometimes for type II we can shorten 8 moves to 7 moves, for type IV we can shorten (5,1) to 9 or 7, and for type VI we can shorten (6,1) to 11. These savings aren't too common because vertex setups are often needed. Still, it was interesting and fun.


Top 


Stefn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 4:44 pm 

Joined: Sun Aug 29, 2010 1:56 pm

Julian wrote: 5.1.19
Stefan  I think your solution has the best order when solving by piece type, and probably has optimal algos too. It's a type of standard method, wich I always use first to solve an unknown puzzle. But your movecount is 40 better than my solve had, and so I recommend using your method, rather than mine. One thing I have to tell you, I didn't find out, how you solve the reduced (big) edges against the corners. Vertex moves seem to be tabu. I tried to find it, but I couldn't. Do you use repeatation  one cycle against an other? I really have no clue. On 5.1.3 I can use vertexmoves. On 5.1.19 not ?! Didn't I understand something. I really want to try your method.


Top 


gelatinbrain

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 5:16 pm 

Joined: Thu May 31, 2007 7:13 pm Location: Bruxelles, Belgium

bmenrigh wrote: I'm wondering how deep those loop cuts can be taken. What would happen if they extended past the corner and into the other side's domain? I'd say something like 3.1.31 would happen if the cuts extended past the deep cut of the solid as that is what has happened there. Strange puzzles indeed.
Analogues of 3.1.31 with other shapes(puzzles with two overlapping layers) is geometrically impossible. Because if the cuts intersect with the equator, you cannot turn it without collision. Attachment:
temp0.gif [ 7.22 KiB  Viewed 3098 times ]
Of course, unlike physical puzzles, virtual puzzles don't need to respect the law of geometry. If you ignore the collision, any imaginable puzzle is possible. Attachment:
temp1.gif [ 14.65 KiB  Viewed 3098 times ]
schuma wrote: For example, in 1.1.8, if I define a new type of face turn that affects the first layer of that face together with the second layer of the opposite face (like A,L'&2), 1.1.50
_________________ Virtual Magic Polyhedra Applet(Online) Executable Jar Installer Win32 Executable(Download) troubleshooting


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Jan 09, 2011 7:47 pm 

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

Stefan Schwalbe wrote: One thing I have to tell you, I didn't find out, how you solve the reduced (big) edges against the corners. Vertex moves seem to be tabu. I tried to find it, but I couldn't. Do you use repeatation  one cycle against an other? I really have no clue. On 5.1.3 I can use vertexmoves. On 5.1.19 not ?! Didn't I understand something. I really want to try your method. The reduced 5.1.19 is like solving a Skewb Ultimate where you are only allowed to turn the 4 corners of one orbital. Any time I would like to turn a vertex of 5.1.19, I turn the opposite face and do a puzzle rotation instead. Watch out for a single twisted Skewb Ultimate corner when reducing the small circle pieces! Once you have at least one pair of pieces around each 5.1.19 face in valid relative positions, you are safe and don't need to think about this any more.


Top 


Who is online 
Users browsing this forum: No registered users and 3 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


