View unanswered posts  View active topics
Author 
Message 
Sjoerd

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Mar 09, 2010 10:58 am 

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

did 1.1.37 with reduction to kilominx with a simple (3,1) commutator and then kilominx solve. about 380 moves, and 12.something minutes. very easy. As for the other new cicleminxes: I can probably do a couple of them with the algorithms I saved for the first couple of circleminxes, but most of them will probably take me ove 45 minutes, and about 9000 moves, so I'm not gonna do that right away.
_________________
OlivÃ©r Nagy wrote: 43,252,003,274,489,856,000. Or the full number in Hungarian is: NegyvenhÃ¡romtrilliÃ³kÃ©tszÃ¡zÃ¶tvenkÃ©tbilliÃ¡rdhÃ¡rombilliÃ³kÃ©tszÃ¡zhetvennÃ©gymiliÃ¡rdnÃ©gyszÃ¡znyolcvankilencmilliÃ³nyolcszÃ¡zÃ¶tvenhatezer )


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Mar 09, 2010 1:40 pm 

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

Elwyn wrote: know i could have saved a fair few moves if i had Brandon's... wait did you mean 4 cycle or 22 swap because i didn't think you could have 4 cycles on puzzles like this just double swaps 3 cycles and 5 cycles. There is no way you'll save moves or time using any algorithm I come up with. I'm the very definition of inefficient Using Julian's notation for dodecahedra back on page 24, my base routine was (U face, u slice): (l u' l') U' (l u l') UBut I generally added a f' setup move or even DR f'I didn't check very carefully but there may have been 3 pieces moving on the U face rather than just 2. That is, what I thought was a 4 cycle could be a 5 cycle. I was using it like it was a 3cycle though much like you can generally think of the bigcubes center threecycle as a swap. Before I started solving I spent about 20 minutes looking for a simple 3cycle that I could use to reduce to a Kilominx but I didn't find anything. Can (or Sjoerd!) you share the steps you went through to find a routine to reduce/pair with the corners? I sure as heck didn't find any (3,1) madness that did anything useful
_________________ 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 Mar 09, 2010 6:31 pm 

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

1.1.37
I saw Brandon's post and thought, "Either that's a typo or Gelatinbrain has suddenly added a lot of new puzzles!" Thanks again, Gelatinbrain. My method for 1.1.37:
1) Build the 5 corners with white stickers and solve those 5 Kilominx corners.
2) Rather like pairing edges of a 4x4x4 cube, use an equator "working area" and equator slice moves to build another 4 kilominx corners, storing completed corners on the face opposite the solved face. (I build the 5 corners that are closest to the white face, both for easier recognition and to keep the unsolved Kilominx corners closer together to reduce setup moves, in step 4.)
3) Solve the remaining corners like a Kilominx.
4) Cycle the remaining center triangles using pure (3,1) commutators. (The 3 moves are a mixture of slice and face moves.)
I took 314 moves on my second attempt. I don't think it will be long before someone (most likely Michael) gets a sub250 solve for this one.


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Mar 09, 2010 9:21 pm 

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

Julian wrote: 4) Cycle the remaining center triangles using pure (3,1) commutators. (The 3 moves are a mixture of slice and face moves.)
I took 314 moves on my second attempt. I don't think it will be long before someone (most likely Michael) gets a sub250 solve for this one. I too took 314 on my seccond attempt except i used a (4,1) not (3,1) 3 cycle but i just found the (3,1) so i might do it again. I know i can atleast get sub 300 but i'll leave it to Michael to get sub 250 hahaha.
_________________ 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 Mar 10, 2010 1:41 am 

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

I finally finished 1.1.6 Attachment:
1.1.6_solved.png [ 29.94 KiB  Viewed 4966 times ]
I ran into just two corners twisted and I spent a miserable 1.5 hours fixing it. The trick ended up being putting a twist into another corner so that I could untwist all three at a time in the same 3cycle. I'm quite impressed at how good you guys are at these things. I figured out a method 1.1.20 but that monster will have to wait until this weekend.
_________________ 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: Wed Mar 10, 2010 5:47 pm 

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

bmenrigh wrote: I finally finished 1.1.6 I ran into just two corners twisted and I spent a miserable 1.5 hours fixing it. The trick ended up being putting a twist into another corner so that I could untwist all three at a time in the same 3cycle. I'm quite impressed at how good you guys are at these things. I figured out a method 1.1.20 but that monster will have to wait until this weekend. Danny came up with a neat (12,2) commutator to twist 2 corners pure in 28 moves; double corner swap, setup a corner to the same position in a different orientation, undo double corner swap, undo setup: [F'2,C2,F2,C'2]x3, E,D, [C2,F'2,C'2,F2]x3, D',E' Useful for the Pentultimate too, if you end up with 2 twisted corners.


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Mar 10, 2010 5:59 pm 

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

1.1.18 is connected to some of the newly added puzzles. 1.1.18 can be solved using reduction or cycling or a mixture. Every piece type can be cycled with (3,1) commutators, some pure, some not.
1.1.31 = 1.1.18, solving the noncircle parts of pieces, then circle pieces exactly like 1.1.26.
1.1.32 = 1.1.18, solving the circle/inner parts of the diamonds, then bitten diamond crescents pure (10,1).
1.1.33 = 1.1.18, solving the noncircle parts of pieces, then cornercircle "wedges" nonpure (8,1), then edgecircle "keystones" pure (10,1).
Some of the others look seriously tricky!
1.1.35 = Pentultimate/1.1.7, then solve the circle pieces pure (10,1).


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Mar 10, 2010 10:21 pm 

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

Julian wrote: 1.1.18 is connected to some of the newly added puzzles. 1.1.18 can be solved using reduction or cycling or a mixture. Every piece type can be cycled with (3,1) commutators, some pure, some not.
1.1.31 = 1.1.18, solving the noncircle parts of pieces, then circle pieces exactly like 1.1.26.
1.1.32 = 1.1.18, solving the circle/inner parts of the diamonds, then bitten diamond crescents pure (10,1).
1.1.33 = 1.1.18, solving the noncircle parts of pieces, then cornercircle "wedges" nonpure (8,1), then edgecircle "keystones" pure (10,1).
Some of the others look seriously tricky!
1.1.35 = Pentultimate/1.1.7, then solve the circle pieces pure (10,1). Good to know that they are related. But I haven't solved 1.1.18... Actually I haven't worked on any puzzle on gelatinbrain's page that is not a platonic solid. By the way, there are at least two hidden puzzles, 1.1.35b and 1.1.39. One can choose them through the menu.
_________________ Check out some virtual puzzles I created at http://nan.ma


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Mar 12, 2010 3:42 pm 

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

The new 1.1.x puzzles have lots of interesting connections. More solution outlines without actual algos: 1.1.34 = solve everything except the inner circles like 1.1.18, then macaroni pieces (8,1) non pure, then little triangles (10,1) pure. 1.1.36 = solve corners and inner circles like 1.1.8, then little triangles non pure (7,1), then curvy/bitten trapezoids (5,1) and macaroni pieces (8,1) pure in either order. 1.1.38 = solve corners and inner circles like 1.1.37 then curvy/bitten trapezoids and macaroni pieces using exactly the same algos as 1.1.36. Currently hidden puzzles, available from the File menu at the top left of the applet: 1.1.35b = Circle SuperPentultimate. Solve the SuperPentultimate then the circle pieces (10,1) pure. 1.1.39 = 1.1.3 + 1.1.6 = . I'd imagine that reducing to 1.1.3 and then solving the reduced Pyraminx Crystal would be an efficient method, but I haven't played with it much yet. Edit: I had forgotten about the little triangles of 1.1.36, so I've added those, and also shortened the curvy trapezoid algo from (6,1) to (5,1).


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sat Mar 13, 2010 1:49 pm 

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

With all of this talk about 1.1.18 I really want to solve it. My solution for 1.1.20 applies pretty nicely to 1.1.18 but I keep getting 20%50% of the way through a solve and Java crashing on me. It isn't a Java code issue, the java.exe binary itself crashes. It has been happening when I click and drag the view around. 1.1.18 is the only puzzle I've run into that during a drag the whole applet will randomly hang for about a second. It is during these hangs that Java occasionally crashes. I'm guessing 1.1.18 is causing one of the JOGL DLLs to do something different and crash.
I guess I'll have to go solve 1.1.20 first. The thing is going to take me like 30003500 moves and 34 hours of time.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sat Mar 13, 2010 4:11 pm 

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

1.1.20 turned out to be quite a bit less tedious than I was expecting. I ran into an edgepair position parity but that only took me about 5 minutes to figure out (standard tricks): Attachment:
1.1.20_solved.png [ 34.18 KiB  Viewed 4829 times ]
1) Pair edges. I used a 3cycle commuted commutator of the form (1, (1, 1)) 2) Solve edgepairs like 1.1.3 3) Solve center pentagons like 1.1.4 4) Solve center diamonds exactly bigcube center 3cycle 5) Solve center trapezoids/tip triangles independently like center triangles from 1.1.4 1.1.18 is actually less work than this so if I can figure out what is causing Java to crash it should be faster.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sat Mar 13, 2010 11:26 pm 

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

During the last little bit of solving for 4.2.3 the applet detected the puzzle as solved prematurely Attachment:
4.2.3_bug_solved.png [ 8.71 KiB  Viewed 4802 times ]
The clock is stopped and the results already submitted. It took me a few more moves and another 20 seconds or so to solve this but I still feel like I cheated.
_________________ 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: Mon Mar 15, 2010 7:45 am 

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

I just checked the total puzzles solved reccords and it seems Brandon has solved 71 new puzzles in the last 43 days I think i've solved 10 hahaha. Not only that but of all the circle cubes that have been solved he has solved all of them. So much for your goal of getting to 40 puzzles solved hey. And Schuma is only one puzzle away from 200 solved puzzles!!! do thesee people do nothing but solve it is amazing.
_________________ 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: Mon Mar 15, 2010 4:55 pm 

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

Elwyn wrote: I just checked the total puzzles solved reccords and it seems Brandon has solved 71 new puzzles in the last 43 days I think i've solved 10 hahaha. Not only that but of all the circle cubes that have been solved he has solved all of them. So much for your goal of getting to 40 puzzles solved hey. It looks like I started solving somewhere around the 27th or 28th of Jan 2010. I did the SuperMegaminx on my second or third day. It isn't fair to compare my 71 with your 10 because I had all of the easy puzzles to fly through while you've had a bunch of hard puzzles to do. The more I solve the more impressed I am with some of your very low move counts. That and I'm especially impressed with your solutions to 1.1.(16,17,21) as well as 2.2.(5,6). As for circle cubes, well, they are all based on roughly the same principles. 3.1.15 throws the 5x5x5 centers into the mix and 3.1.19 has some interesting inner 3x3x3 pieces. Once you know the tricks solving them is kinda tedious. I've kinda been wanting to write up a Circle Cube "tricks/solutions" post describing the basics of the cubes. I haven't wanted to piss you guys off though revealing too much info and allowing people that would not solve them on their own to get an unearned big boost in solutions. Thoughts? Elwyn wrote: And Schuma is only one puzzle away from 200 solved puzzles!!! do thesee people do nothing but solve it is amazing. Yeah Schuma is a machine. I wish I had nothing to do but solve. Work and school have been kicking my butt so I've been solving on the weekends to escape from regular duties.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Mar 15, 2010 6:16 pm 

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

bmenrigh wrote: Elwyn wrote: Elwyn wrote: And Schuma is only one puzzle away from 200 solved puzzles!!! do thesee people do nothing but solve it is amazing. Yeah Schuma is a machine. I wish I had nothing to do but solve. Work and school have been kicking my butt so I've been solving on the weekends to escape from regular duties. Actually I am a human. I confess that I have spent a lot of time playing these puzzles. Fortunately the grad school is like a kindergarden. Check this comic strip to see why: http://www.phdcomics.com/comics/archive ... micid=1286Sometimes I solve them in working hours, if my advisor is not pushing me.
_________________ Check out some virtual puzzles I created at http://nan.ma


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Tue Mar 16, 2010 12:54 am 

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

bmenrigh wrote: I'm especially impressed with your solutions to 1.1.(16,17,21) hmmm 1.1.17 is that the one i solved so fast and so efficiently that i don't even remember it and niether does the score board As for 1.1.21 i'm certain you could solve it as soon as you realised that it is just a different puzzle (that you have solved) missing some corners and in disguise. Infact have a closer look at 1.1.16 too and it will hopefully apear very eazy. It was 1.1.15 that i found hard. I am still verry proud of my 2.2.5 solve though hahaha but i think i need to resolve it concidering i now know a pure (3,1) alg for the pyraminx crystal like edge pieces instead of a (4,1). Won't save manny moves but i might be able to get sub 700 which would be nice. I think that's why i haven't solved as manny puzzles as you hahaha because i get too caught up in getting lower and lower move counts.
_________________ 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: Tue Mar 16, 2010 2:06 am 

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

Elwyn wrote: bmenrigh wrote: I'm especially impressed with your solutions to 1.1.(16,17,21) hmmm 1.1.17 is that the one i solved so fast and so efficiently that i don't even remember it and niether does the score board Oops, haha, not even Michael can beat a move count of 0 ! Of course, I meant 1.1.15 Elwyn wrote: As for 1.1.21 i'm certain you could solve it as soon as you realised that it is just a different puzzle (that you have solved) missing some corners and in disguise. Hmm, I've turned that thing for at least 15 minutes and not spotted it. I need to think about those pieces some more. Elwyn wrote: Infact have a closer look at 1.1.16 too and it will hopefully apear very eazy. It was 1.1.15 that i found hard. I haven't played with 1.1.16 it much yet but I already see how to solve about half of it. Maybe I'll tackle the rest of it in a few days. Elwyn wrote: I am still verry proud of my 2.2.5 solve though hahaha but i think i need to resolve it concidering i now know a pure (3,1) alg for the pyraminx crystal like edge pieces instead of a (4,1). Won't save manny moves but i might be able to get sub 700 which would be nice. I'm not very comfortable with the view I get on the deepcut icosahedron duels of the dodecahedra. I was able to solve 2.2.1 and 2.2.2 with their nice shallow cuts but I still have trouble spotting the correspondence between the pieces in 2.2.3 and pieces in 1.1.4. Plus those supercenters are going to be annoying. I hope once I start solving them they will become as natural as their dodecahedra counterparts. Elwyn wrote: I think that's why i haven't solved as manny puzzles as you hahaha because i get too caught up in getting lower and lower move counts. Yeah for sure, you'd be killing me in solves if that was your goal. I need to work on techniques for developing shorter routines for pure cycles. After about 30 minutes of playing with 1.1.39 I think I now have a solution that would work in theory. I could never actually pull it off. though. I basically have 4 nonpure cycles that work my way in to the center of the puzzle, each time solving the side effects of the previous nonpure piece movement and at the same time scrambling new, more interior pieces. The only pure cycle I have would finally fix the most interior pieces. I bet Julian already has a bunch of pure, short 3cycles for that thing . I feel dumb when I can't even come up with a pure 3cycle for either type of the center pieces in 1.1.8 I can move them around together so a pure cycle for either would take care of the mess. I guess I need a healthy dose of patience .
_________________ 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: Tue Mar 16, 2010 3:44 am 

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

bmenrigh wrote: I feel dumb when I can't even come up with a pure 3cycle for either type of the center pieces in 1.1.8 I can move them around together so a pure cycle for either would take care of the mess. I guess I need a healthy dose of patience . 1.1.8 is to me the most anoying puzzle on here just because i like doing things on order and it stumped me and put a gap between allot of solved puzzles so i have a gap but after reading your post i decided to have another look at it and i now have a (3,1) for the small center triangles non pure but i think i will use reduction for most of those and a (4,4) pure 3 cycle for the wide triangles. It's a little long and it would be better if it was for the small triangles because then i'd try and solve it like 1.1.37 then just cycle them last. hmmm perhaps i'll keep looking for something better because that sounds like it could be a good method. Edit: found a (4,3) pure cycle for the small triangles and a (3,1) non pure for the wide triangles. Oh i forgot to mention that when i said i had a (3,1) for the centers of 1.1.37 i was yet again being fooled by a 5 cycle looking like a 3 cycle but i'd say this new one would work just fine.
_________________ 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: Tue Mar 16, 2010 10:40 pm 

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

I don't undersatnd this pariy. Attachment:
parity.jpg [ 56.73 KiB  Viewed 4743 times ]
The pices that are wrong are equivelent to the (edge) centers on a 5x5x5 and there are no identicle pieces for this piece type just mirror images that don't work in the same spot....I know i should swap some identicle pieces but i don't know what when it is normaly obvious. It's 4.2.9 if that wasn't clear.
_________________ 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 Mar 17, 2010 12:16 am 

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

Sorry for the poor reply, I'm on a phone right now. That pairity comes from an odd number of quarter turns. If you put a counterclockwise quarter turn into that tip you will have a 3cycle to finish the edges. You would do better to put the pairity into the next layer down so that the quarter turn doesn't twist the 4color tip.
_________________ 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: Wed Mar 17, 2010 4:50 pm 

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

1.1.8
Psst... small triangles (3,1) non pure, then big triangles (3,1) pure.


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Thu Mar 18, 2010 5:20 pm 

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

bmenrigh wrote: After about 30 minutes of playing with 1.1.39 I think I now have a solution that would work in theory. I could never actually pull it off. though. I basically have 4 nonpure cycles that work my way in to the center of the puzzle, each time solving the side effects of the previous nonpure piece movement and at the same time scrambling new, more interior pieces. The only pure cycle I have would finally fix the most interior pieces. That can be a good way to solve a puzzle! I often use the dirtiest algos possible, with only a single pure algo at the end. bmenrigh wrote: I bet Julian already has a bunch of pure, short 3cycles for that thing . I didn't look at 1.1.39 properly until tonight. Have you tried Ctrlclick with this one? It looks like it's supposed to move just the very outer slice, the same as Shiftclick for the 1.1.6. I'm guessing that Gelatinbrain is aware of the bug and has deliberately kept it off the main webpage until it's fixed? With working Ctrlclick moves, we could solve the corners and edges like a Pyraminx Crystal, then kite pieces (3,1) non pure, trapezoid (3,1) non pure, and finish with large and small triangles in either order (3,1) pure. But without working Ctrlclick moves I would solve in a different order, cycling corners (4,1) near the end and finishing with the small triangles (5,1) pure. (The kicker is the 1.1.6 trapezoid pieces. If you want to cycle them without moving corners, I can't see faster than (3,4), where the 4 are simulating a Ctrlclick move.)


Top 


Mindstormscreator

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Thu Mar 18, 2010 8:14 pm 

Joined: Wed Mar 11, 2009 3:33 pm Location: Pennsylvania, USA


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Mar 19, 2010 7:30 am 

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

Julian wrote: I didn't look at 1.1.39 properly until tonight. Have you tried Ctrlclick with this one? It looks like it's supposed to move just the very outer slice, the same as Shiftclick for the 1.1.6. Weird. 1.1.40 the circle version of this puzzle also has this weird movement.
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Mar 19, 2010 10:03 am 

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

I also noticed the buggy move on 1.1.39. I made that move not by ctrl+click, but with the command like "A&4". But it was so weird that I decided not to use it in my solve.
_________________ 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: Sat Mar 20, 2010 3:15 am 

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

Julian wrote: 1.1.8
Psst... small triangles (3,1) non pure, then big triangles (3,1) pure. Madness I finally have a pure 3cycle for the big triangles but it is (10,1). I'll probably give it a solve with this routine. I'm amazed at how easily (quickly) you find pure (3,1) or (4,1) cycles. For 1.1.8 I just don't see how to isolate a piece in only 3 moves. Thanks for your help so far . Edit: 5 minutes after I wrote the above I thought of a way to do it in (3,1) pure .
_________________ 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: Sat Mar 20, 2010 8:25 pm 

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

bmenrigh wrote: I'm amazed at how easily (quickly) you find pure (3,1) or (4,1) cycles. For 1.1.8 I just don't see how to isolate a piece in only 3 moves. Edit: 5 minutes after I wrote the above I thought of a way to do it in (3,1) pure . The "Eureka!" moments are great though, aren't they? A few people have asked for tips on how to find useful algos, so here goes... Sorry for the lack of screenshots, but work is killing me at the moment! If you label two moves L and R, try all possibilities of L' R L, (L',R), and (L,R), all distances apart, double turns or single turns if you have 72 or 90 degree turns, and regular or slice if you have a choice. For L' R L with slice or hybrid puzzles, it usually makes more sense for the L move to be the one that affects more pieces. Look out for any piece types that have been isolated, or almost isolated. If you see a situation where you think, "Apart from this group of pieces here being messed up, this would be a good algo", try to find 1 or 2 moves that push the messed up pieces conveniently out of the way. Then reset the puzzle, and try the inverse of those moves, the almostuseful ago, then those moves. So if A helps (L',R), the next try is A' (L',R) A. Maybe that doesn't quite do the job, but another move B seems to fix everything up. Reset again, and try B' A' (L',R) A B. I did this with 1.2.9, one of the most difficult Gelatinbrain puzzles, and after some experimentation found an algo (4 moves) (L',R) (4 moves)' to isolate a triangle in 12 moves. In general, with deeper cut or deep cut puzzles, it's a case of "Try to push all the scrambled junk together on one side, and hope that just one piece gets left behind." It might sound crude, but it's a very effective strategy! The algos built up from 3 central moves use a conjugate sequence to isolate a piece. The formula is a short sequence of moves to isolate a piece from its neighbors, a move to swap that piece with another (but with side effects, other pieces in that part of the puzzle are moved too), then undo the initial short sequence to get back to a position where you have isolated a single swapped piece with no side effects. "Don't be cleaner/purer than you have to." Some really useful short algos can look horrible at first glance, until you look carefully and see that although lots of pieces move, only one of a particular piece type is cycled. That can be your algo to finish off that piece type near the beginning of the solve. One algo of the type (1 setup) (L,R) (1 undo)  F L R L' R' F'  is incredibly versatile. Always give it a try, spacing out the turn points of F, L, and R, and trying both symmetic and assymetric arrangements. Sometimes it's worthwhile making a commutated commutator, for example, a (3,1) commutator isolates a piece in a turning region, so you commutate it again for a ((3,1),1) or (8,1) algo overall. Or perhaps the (3,1) needs a setup move either side to isolate a piece, leading to a (10,1) algo overall. With shallower cut edge turning puzzles , conjugate algos are often efficient, like (ABA,C) or (ABCBA,D) or (ABA,CDC) in addition to (A,B). Also, check out every possible distance apart for (LR)*2 and (LR)*3 as starting points for algos. The goal is to find the right depth of cut so that (LR)*2 cycles 3 groups of pieces, one of which you slice up further; or with deeper cut puzzles, to flip over 2 groups of pieces with (LR)*3 and then push all except one out of a turning region.With some deep cut hybrid puzzles (e.g. 1.3.2, 2.4.1, 3.6.2, 4.5.1) you'll find that if you take a move of each type as far apart as possible and repeat a commutator 3 times, you'll nearly have a pure algo: just one setup move either side of (L',R)*3 will give you a pure (14,1) algo. Next, don't forget that cycling isn't always the best way to solve a particular puzzle. You might be able to get a long way blockbuilding, gather groups of pieces together, in the early stages. Also there is reduction, reducing a complex puzzle to a simpler type you already know how to solve. For example, with 4.1.24.1.4, you can reduce to a Pyraminx or Jing Pyraminx solution once you have solved 4 non adjacent faces. Let's say you normally finish by solving the center triangles with (L R L, r), where L and R are non adjacent faces. If you invert this and write out all the moves, that's r' L R L r L' R' L'. But wait  those last 3 moves do nothing to spoil the reduced puzzle; they just move the bigger groups of pieces around. So with r' L R L r you have with just 5 moves instead of 8 moves done a cycle of little triangles, not caring where the Pyraminx edge groups go, to finish up with a Pyraminx solve at the end, saving loads of moves. Similarly with 3.4.x puzzles, if you have an algo mixing face moves and corner moves, you can leave out all the face moves at both ends of an algo to get a shorter algo suitable for reducing the puzzle to a cube. Edit: Added more detail on edge turning puzzles in bluish text.
Last edited by Julian on Mon Mar 22, 2010 7:04 pm, edited 1 time in total.


Top 


Elwyn

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sat Mar 20, 2010 9:46 pm 

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

Julian wrote: For example, with 4.1.24.1.4, you can reduce to a Pyraminx or Jing Pyraminx solution That's how i solved 4.1.24.1.4 after reading something about it a while ago and it was a strange but very fun method Julian wrote: 1.1.8 Psst... small triangles (3,1) non pure, then big triangles (3,1) pure. I still havn't found a pure big triangles alg smaller than (4,4). I do have a pure small traingles (4,3) and a non pure (3,1) so i could solve it but i'd rather solve it with a shorter alg but uni just got rather hectic so i don't really have time to look for one...... actulay i should be studying maths right now hahaha
_________________ 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: Sat Mar 20, 2010 10:13 pm 

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

Elwyn wrote: Julian wrote: 1.1.8 Psst... small triangles (3,1) non pure, then big triangles (3,1) pure. I still havn't found a pure big triangles alg smaller than (4,4). I do have a pure small traingles (4,3) and a non pure (3,1) so i could solve it but i'd rather solve it with a shorter alg I ended up solving it with a (3,1) pure for the wide triangles and a ((1,1),1) pure for the small triangles. In hindsight, I tried all sorts of stupid, overly complicated ways to do the wide triangles pure. When I found the (10,1) routine I realized that it was easy to isolate the wide triangle in the same way without that whole long sequence of moves. Of course, I only place one piece with each 3cycle so my move count was still awful compared to what you and Julian typically get.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sat Mar 20, 2010 10:24 pm 

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

I have been wanting to solve the circle 1.1.4 for a long time (1.1.25) so I decided to save it for my 100th solve : Attachment:
1.1.25_solved_100th.png [ 33.1 KiB  Viewed 4336 times ]
1) Solve corners, 1.1.3 edges, and centers just like 1.1.4 2) Solve circle pizza wedges like the center triangles on 1.1.4 (7,1) nonpure 3) Solve center triangles pure. I used a commuted Megaminx PLL routine (15,1) I took several 5 minute breaks in this solve so I could cut about 1520 minutes off this solve.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Mar 21, 2010 12:23 am 

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

I took a stab at 3.4.5 but I ended up with a parity I don't understand: Attachment:
3.4.5_parity_1.png [ 15.98 KiB  Viewed 4411 times ]
It is easy to turn into a position parity instead: Attachment:
3.4.5_parity_2.png [ 15.93 KiB  Viewed 4411 times ]
Normally I'd solve this by parity with halfturn in the top but in this case, that will put a parity into the 3x3x3 edges instead. I've been in this situation for more than an hour and so far I haven't made any useful progress. I have tried swapping two identical pieces but I don't think that can be done without flipping their orientation. Ideas?
_________________ 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 Mar 21, 2010 12:28 am 

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

bmenrigh wrote: I have been wanting to solve the circle 1.1.4 for a long time (1.1.25) so I decided to save it for my 100th solve : 100 puzzles congradulations!!! but why not solve it using the (4,1) alg that you know for the triangles for 1.1.4 to solve the large pizza pieces non pure. It cycles two on the U face and one on the DR face. That's how i plan to do it and then a [(4,1),1] or (10,1) comutated version of it for the small triangles. Can't help you with dino plus 3x3x3 because i havn't solved it yet sorry. Edit: this parity is dicussed back on page 13 a little more than half way down and Doug came up with a solution eventualy i think UMichSpeedCubist wrote: This is what I would do. I would do the single quarterturn and then solve all the corners leaving edgepairty. Then make a single quarter turn of an innerslice of 3x3 to convert the edgeparity to centergroup parity. And then resolve centers in some GRUELING way.
Or perhaps not so bad because I would just use the idea of fixing 4x4 wingpairty (r U2 r U2 r U2 r U2 r) where the (r) are appropriate Dino turns, and the (U2) is a pure center180 alg on Super3x3. OKAY, that does work!
_________________ 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 Mar 21, 2010 1:23 am, edited 1 time in total.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Mar 21, 2010 1:23 am 

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

Elwyn wrote: 100 puzzles congradulations!!! Thanks I didn't think I would ever make it to 40, then 60, then 80 I'm starting to get into the "hard" puzzles so progress is going to be slower going. I need to figure out the 5.2.x series but those things just don't make any sense... Elwyn wrote: but why not solve it using the (4,1) alg that you know for the triangles for 1.1.4 to solve the large pizza pieces non pure. It cycles two on the U face and one on the DR face. That's how i plan to do it and then a [(4,1),1] or (10,1) comutated version of it for the small triangles. Mostly bad habit. When I first solve 1.1.4 the setup moves needed to place even one triangle in place seemed very hard to me. When I got down to just two or three unsolved faces I was spending 5+ minutes per triangle thinking and trying setup moves. I found (Sune, U slice) to be easier to work with. Now I'm a lot more comfortable with many setup moves and I could work with the (4,1) routine easily. I really should have... If you destroy my time on 1.1.25 I might have to go back and do it with something that won't take 3000+ moves
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sun Mar 21, 2010 2:04 am 

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

Elwyn wrote: Can't help you with dino plus 3x3x3 because i havn't solved it yet sorry. Edit: this parity is dicussed back on page 13 a little more than half way down and Doug came up with a solution eventualy i think Okay, I considered this (solving the edgeparity like a void cube) but rejected it as too much work (or maybe even impossible) to resolve the centers. Doug's idea of using a Dino turn and then a super3x3x3 180 to swap 2 center pieces at a time is a great idea and would make it almost efficient. I hadn't read to page 13. Doug is quite a good solver it seems. I should work my way through this monster thread reading and learning.
_________________ 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: Sun Mar 21, 2010 7:41 pm 

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

1.1.39 & 1.1.40  Ctrlclick moves have been fixed! Thanks Gelatinbrain!
1.1.39: Solve the corners and edges like a Pyraminx Crystal, then kite pieces (3,1) non pure, trapezoid (3,1) non pure, and finish with large and small triangles in either order (3,1) pure.
1.1.40: Same as 1.1.39, then finish with the edgecircle pieces (5,1) pure, and cornercircle pieces (10,1) pure, in either order.
Edit: Fixed a goof with my suggested outline for 1.1.40.
Last edited by Julian on Mon Mar 22, 2010 6:54 pm, edited 1 time in total.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Mar 22, 2010 1:48 am 

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

Julian wrote: 1.1.39: Solve the corners and edges like a Pyraminx Crystal, then kite pieces (3,1) non pure, trapezoid (3,1) non pure, and finish with large and small triangles in either order (3,1) pure. Thanks for your writeup on solving and the routines you've devised for these puzzles. It keeps me motivated to find the shorter ones. I finally have short enough sequences for 1.1.39 to give it a solve by hand. I hadn't tried CTRL+click before and I wasn't aware some of the puzzles took CTRL in addition to shift. With a slight modification to my (3,1) routine for the wide triangles on 1.1.8 and the nowworking CTRL+click I have a very clean (but not pure) (3,1) cycle for the trapezoid pieces on 1.1.39 Next weekend 1.1.39 is going down
_________________ 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: Tue Mar 23, 2010 9:14 am 

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

I decided to have a look at the dino+3x3x3 and i got the parity i then solved void cube parity and had 4 swaped (whole 3x3x3) centers i did it all using "dino turn RUR'Ux5 undo dino turn" On a different note i updated my solution for 2.2.5 back on page 30 with some new ideas that i'm suprised Julian didn't point out would be more eficient back when i first posted it. I also spent a little time looking for Doug's magic 24 move single center twist for 2.2.32.2.6 and came to the conclusion that it is just that.... Magic.
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Thu Mar 25, 2010 8:49 am 

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

Does anybody know whether gelatinbrain is updating the rankings manually or he has a program to do it automatically? I have noticed that my solves since the day before yesterday have not been shown in the rankings. I have never seen this before. Usually it always shows up within no more than one day.
_________________ Check out some virtual puzzles I created at http://nan.ma


Top 


Sjoerd

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Thu Mar 25, 2010 9:37 am 

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

I heard the score rankings get updates when GB logs in. So he just didn't log in for a couple of days.
_________________
OlivÃ©r Nagy wrote: 43,252,003,274,489,856,000. Or the full number in Hungarian is: NegyvenhÃ¡romtrilliÃ³kÃ©tszÃ¡zÃ¶tvenkÃ©tbilliÃ¡rdhÃ¡rombilliÃ³kÃ©tszÃ¡zhetvennÃ©gymiliÃ¡rdnÃ©gyszÃ¡znyolcvankilencmilliÃ³nyolcszÃ¡zÃ¶tvenhatezer )


Top 


gelatinbrain

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Thu Mar 25, 2010 5:47 pm 

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

Sjoerd wrote: I heard the score rankings get updates when GB logs in. So he just didn't log in for a couple of days. There seems to be a serverside problem. I haven't received any scores for the last two days. It happens sometimes, but since I don't have any serverside permissions, there's no way to detect the exact cause. By cutting off the connection just before clicking "submit", you can get the certificate to mail me. For lost scores, you can PM me screenshots or other "proofs" . I will manually update the scoreboard.
_________________ 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: Thu Mar 25, 2010 6:39 pm 

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

gelatinbrain wrote: Sjoerd wrote: I heard the score rankings get updates when GB logs in. So he just didn't log in for a couple of days. There seems to be a serverside problem. I haven't received any scores for the last two days. It happens sometimes, but since I don't have any serverside permissions, there's no way to detect the exact cause. By cutting off the connection just before clicking "submit", you can get the certificate to mail me. For lost scores, you can PM me screenshots or other "proofs" . I will manually update the scoreboard. This is a bummer. I solved 1.1.15 in about 1 hour, 30 minutes and about 2100 moves as well as 3.4.5 in about 1 hour 15 minutes but I don't know a rough move count. I haven't been taking screenshots of non "hard" puzzles and since I want the time and moves to be accurate, I'll just resolve them. Perhaps you would consider a backup system? I'm thinking something along the lines of me or somebody else running a CGI script somewhere else on the web. When a puzzle is solved, in addition to posting to your script, it posts to the backup script. In the event of catastrophic loss the records of solves can be synced up. Actually in the more general case, I'd be happy to help with some of the support/backend code and maintenance.
_________________ 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: Thu Mar 25, 2010 8:15 pm 

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

Hmmm this makes me kind of glad the only puzzles i've solved have been dino+3x3x3 and resolved the superx for a better movecount. Gelatinbrain do you have any idea when this might be fixed?
_________________ Some PBs 3x3x3 :20.7 seconds, 5x5x5 2:33, gigaminx 16:40, 7x7x7 9:48, pyraminx crystal 3:42


Top 


schuma

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Mar 26, 2010 12:43 pm 

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

I am glad to see that the rankings have already got updated. All my recent solves are there. It seems the problem is solved. Thank you GB.
_________________ 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 Mar 26, 2010 5:01 pm 

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

1.1.39 Attachment:
1.1.39_solved.png [ 36.41 KiB  Viewed 4079 times ]
This puzzle is awesome. I really like 1.1.3 and 1.1.6 so combining them makes for a fun solve. I'd suggest Julian's method if you can find the routines. Mine was: 1) Solve corners and edges like 1.1.3 2) Solve center big trapezoids much like 1.1.8/1.1.37 (3,1) nonpure 3) Solve center small triangles ((1,1),1) nonpure 4) Solve center diamonds/kites (3,1) nonpure (big cube centers) 5) Solve center big triangles ((1,1),1) pure I really struggle with the setup moves for deepcut puzzles like this or 1.1.4 so phase 3 took me 1.5 hours. Before I started I thought I'd have a chance to beat Schuma's time but I think that even with the macro penalty his time was really good.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Sjoerd

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Mar 26, 2010 5:54 pm 

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

I took a look at 1.1.39 and made up all the neccicary(correct spelling?) algorithms. It is recuction to pyraminx crystal, and the algorithms aren't that long. I have 2 of 8 moves, 2 of 10 moves and 2 of 22 moves. I do not have the time to put them to use right now. And just a note partially related to this: All the talk lately about all of these commutators, I found one this long, then someone finds a shorter one, is very nice and makes for a little bit of competition(which is good), but it's not very satisfying to read. I do not want to spoil anyones fun, but can I just show the algorithms? Anyone can then use them, and a puzzle like 1.1.39 wouldn't be that hard anymore, but I would just like to share some stuff. Hmm now that I hear myself saying it it doesn't seem to be a good idea
_________________
OlivÃ©r Nagy wrote: 43,252,003,274,489,856,000. Or the full number in Hungarian is: NegyvenhÃ¡romtrilliÃ³kÃ©tszÃ¡zÃ¶tvenkÃ©tbilliÃ¡rdhÃ¡rombilliÃ³kÃ©tszÃ¡zhetvennÃ©gymiliÃ¡rdnÃ©gyszÃ¡znyolcvankilencmilliÃ³nyolcszÃ¡zÃ¶tvenhatezer )


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Mar 26, 2010 6:13 pm 

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

Sjoerd wrote: I took a look at 1.1.39 and made up all the neccicary(correct spelling?) algorithms. It is recuction to pyraminx crystal, and the algorithms aren't that long. I have 2 of 8 moves, 2 of 10 moves and 2 of 22 moves. I do not have the time to put them to use right now. It's necessary but since English isn't your native language I don't think it is necessary that you spell anything correctly When Julian first suggested reducing to the Pyraminx Crystal I investigated that strategy but I didn't think I would be able to do it. Sjoerd wrote: And just a note partially related to this: All the talk lately about all of these commutators, I found one this long, then someone finds a shorter one, is very nice and makes for a little bit of competition(which is good), but it's not very satisfying to read. I do not want to spoil anyones fun, but can I just show the algorithms? Anyone can then use them, and a puzzle like 1.1.39 wouldn't be that hard anymore, but I would just like to share some stuff. Hmm now that I hear myself saying it it doesn't seem to be a good idea Personally I would prefer to share too routines but I don't want to upset others. There is a certain amount you can learn by figuring things out yourself and then there is all the things you can learn from others. I'm not going for fewest moves or fastest time so if I come up with a really cool routine I want to share it too. I suppose we should share routines when we want to and we know it won't wreck other's fun. For 1.1.39 the only routine I have that is "valuable" is the trapezoid/1.1.8/1.1.37 cycle.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Sjoerd

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Fri Mar 26, 2010 7:22 pm 

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

Went ahead and solved it anyway Attachment:
1.1.39.JPG [ 81.37 KiB  Viewed 4058 times ]
Attachment:
1.1.39.2.JPG [ 15.56 KiB  Viewed 4058 times ]
Order: Step 1: pair 1 and 2. (1,3) with 1 being a ctrl slice move. Step 2: pair 1 and 2 with 3. Same (1,3) but with 1 being a shift slice move. Step 3: pair 4 and 7. (1,4) 1 being a shift slice move. Step 4: pair 5 and 6 with 4 and 7. ((4,1),1) or something like that. 2 separate algorithms for righthanded (5) and lefthanded (6) pieces. Step 5: solve like Pyraminx Crystal (1.1.3). Easy. Anyone who's interested in the algorithms, PM me.
_________________
OlivÃ©r Nagy wrote: 43,252,003,274,489,856,000. Or the full number in Hungarian is: NegyvenhÃ¡romtrilliÃ³kÃ©tszÃ¡zÃ¶tvenkÃ©tbilliÃ¡rdhÃ¡rombilliÃ³kÃ©tszÃ¡zhetvennÃ©gymiliÃ¡rdnÃ©gyszÃ¡znyolcvankilencmilliÃ³nyolcszÃ¡zÃ¶tvenhatezer )


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Sat Mar 27, 2010 6:57 am 

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

1.1.39
Thinking about it, the best of both worlds would be reducing to a Pyraminx Crystal like Sjoerd, using Brandon's and my shorter algos. My (3,1) algo for cycling the kite pieces non pure (small triangles move too) can be shortened to 5 moves if we let the 1.1.3 pieces move to new positions. My other algos are surrounded by slice moves so they can't be shortened. However, reducing still saves quite a few moves when solving those piece types, for example:
We have a choice between setting up a cycle with slice face slice face face face or face face face face face face face slice. If we solved the 1.1.3 pieces at the beginning and want to leave them solved, the first choice is better, as it's shorter. However, if we are reducing to 1.1.3, the second option is better. 8 moves to setup the cycle, but only 1 undo move: the slice. We don't care about the 1.1.3 moves because we're going to solve the 1.1.3 at the end anyway. So when reducing, the second option is 9 moves in addition to the algo, instead of 12 moves. These savings will really add up. The trick with efficient setups will be to make the smallest number of 1.1.3 moves to either set up the pieces for cycling already, or leave them in a position where a slice or two at the end puts them into position.
Combined with solving 2 or 3 pieces at a time as often as possible, I this one can be solved in under 2,000 moves.
Edit: I originally said, "My other algos are 100% made from slice moves..." (3rd sentence), which I've since noticed is incorrect. One of them includes a face move.


Top 


Julian

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Mon Mar 29, 2010 4:46 pm 

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

1.2.13
Apart from the single color circle pieces, solves like 1.2.2. The 2color circle pieces show the location and orientation of the 1.2.2. corners. Then cycle the circle wedges (7,1) pure.
1.2.14
Start by solving the big 2color diamonds, moving the bitten diamonds to join the 2color circle pieces. Towards the end the bitten pieces can be cycled (3,1). Then single color bitten diamonds (1,1), pentagon centers (3,1), single color circle pieces (7,1), and finally triangles (5,1). Only the last algo is pure.


Top 


Brandon Enright

Post subject: Re: Gelatin Brain's Applet Solutions Discussion Thread Posted: Wed Mar 31, 2010 12:18 am 

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

1.1.13: Attachment:
1.1.13_solved.png [ 38.62 KiB  Viewed 3891 times ]
This thing seems very daunting at first and is visually very chaotic when scrambled. It is quite simple though. 1) Solve 3color hexagons and 2color >< edge pieces like 1.1.3 2) Solve pentagon centers like centers on 1.1.4/Pentultimate 3) Solve 3diamond hexagons (looks like 3D cube) and wide triangles like + and x centers of 1.1.10 respectively 4) Solve inner narrow triangles that touch pentagon centers like inner small triangles on 1.1.4 There definitely should be a lot more solvers for this puzzle than just 8. Campbell's time is beatable but Michael's time is inhuman as always.
_________________ Prior to using my real name I posted under the account named bmenrigh.


Top 


Who is online 
Users browsing this forum: Google Adsense [Bot] and 11 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


