I tried 1.2.1. Did the corners within about 5 minutes, but ended up with one corner twisted I think I will try this again later.

Michael, imagine the super x. Now twist one of the 2x2 corners. How would you do it? Apply that to the final corner of 1.2.1

Also it seems Gelatin Brain added a certificate code you can copy just incase your time doesn't register.

But the corner is one piece, isn't it?

I think I figured it out though, I just have to fix the center by turning the vertex and then fix the 3-cycle of the other corners. Silly me

To me it's the same concept really. Rotating a corner and cycling adjacent pieces.


For example, on a Super X:

[R UFL R' UFL]x2

That's not the same at all! You can't rotate a corner by doing a commutator with its 3 parts, if it is only one piece...

Maybe you are not talking about the same puzzle. I changed the numbering after qqref's post at Dec 18. They are now sorted by the depth of the cut. So the puzzle qqref was calling 1.2.1 on Dec 1 is now placed in 1.2.3.
Sorry for causing this confusion.

Well, I thought we were talking about what is now 1.2.2, but it's the same problem. The alg you posted was a 3-cycle (actually a conjugated commutator! cool huh) so it would work if a corner was in 3 parts, but in these puzzles the corner is only in one part and you wouldn't be able to twist a single corner with just commutators.

On 1.2.3:

The White/Blue/Purple corner will be known as U
The Dark Green/Light Blue/White corner will be known as R.

U R U R' U R U R'

How is that not a similar idea to what I was saying? Sure, more pieces than the corners are effected, but right now we're just working on the corners.

It's a similar algorithm, but it flips the corner for a completely different *reason*. Take a look at R U R' U R U R' - that does a 3-cycle of teardrop shaped things (which each include a corner). The final U turn fixes the 3-cycle but turns the corner. That's the reason it works on this puzzle. The reason it works on the Super-X, on the other hand, is because the effect you want is the 3-cycle itself.

Well it twists one corner, doesn't it?

When I get some free time I need to catch up to you guys. I've been organizing several big things, including the Newsletter and perhaps opening my own website.

I solved 1.2.2. Don't read the rest of the post if you don't want spoilers.

The easiest 4-move commutators seem to cycle three corners and three triangles, so I used them to cycle triangles at first, and then to construct corner 3-cycles afterwards. So my method was:
1) Solve edges (intuitive, flip two with a commutator)
2) Solve triangles (4-move commutators plus some interesting setups)
3) Permute corners (8-move corners commutator)
4) Orient corners (16-move commutator to orient two corners, plus parity fix to orient the last one)
It took 41:29, which isn't too bad, but I think that with a better method (and a better parity fix, it took me about 8 minutes because I turned the corner and then resolved the three edges, then the triangles and finally the corners that that messed up) I could get a significantly quicker time.

5d5bcebd226cf9dc23a585f60bd926b3
4ced86f20db946738c1d16d22c6798ce
31ed1acb34ec19dc32b6585db03f6239
c497682fd09b6437c89e613dc27689ec

Solved the 3x3 maze. I also did all of the puzzles that have been done by other people besides the following:

1.1.5
1.1.7 (Pentultimate)
and 1.2.2

Accidentally cut off the stats. 51 minutes. 1200+ moves.



Stupid last center. I couldn't get it, so I just ended up solving everything else. I had the center edges (2 color 'wing') solved, but with a 4 cycle on the center corners (1 color rhombus). Since that was the case, and I knew it could be considered a 'parity', or so I assumed, I just left it in a 3 cycle of both Center Corners and Center Wings.

This puzzle has parities coming out of the wazoo.

Edges. Edge Wings, Center Corners.

This is getting pretty frustrating.

fe920d39a596002fd0fa643cc89e613d
c25a89ea135ea1bc439a91dd237a85f4
0bd126b34c7486e70db846738ceb16d2
2c6798ce31e51acb34e619cd32a7584f

Finished 2.2.16.

1580 moves. 48:33.

You've been doing pretty well lately, Doug and Noah. Time for me to get back on my feet a bit.

I noticed you got less than half of my time and movecount in 4.2.7, so I spent some time trying to find a much more efficient way to do that one (I ended up with 8-move 3-cycle commutators for those pieces we used 24-move 5-cycles to solve before; I'm guessing that's pretty much what you did too). Then I did 2.2.15, which took a LOT less time than it would have before.

I noticed there's no delay in pressing Shift now. Thanks gelatinbrain, this should make quite a few puzzles faster to solve

The new puzzles 2.2.8c and 2.2.8d seem interesting. I tried 2.2.8d but got a weird parity error after doing the centers, so I didn't feel like trying to fix it.

Noah - since the tips on 2.2.16, 2.2.17, 2.2.18, 4.2.8, 4.2.9, and 4.2.10 behave a lot like supercube centers, for the last two you might want to just solve the "+centers", and then use commutators for the "xcenters". That's generally what I do. Fortunately edge pairing is so easy on these puzzles

I solved Gigaminx two more times, trying out a modified version of Arnaud's Edge Pairing for 5x5x5 (see the thread in the tutorials section of speedsolving.com). I saved about 100 moves, and I beat my best time by a little on the second attempt (although I think this is just savings on the centers from the faster Shift+click). I love a nice-turning Gigaminx Still haven't done a movecount-optimized solve on it, only speed solves. I'll be more likely to give it a shot if someone else can do under 1000 moves on that thing

gelatinbrain - could you make some kind of leader board, in the high scores list, which just lists the users in order of the number of puzzles solved (along with a percentage)? I think this would be interesting to see. Thanks again for such an addictive program!

I think you guys should spend more time on fewest-moves solving; we could have a lot of competition on the more intuitive puzzles, and you don't have to worry about fast clicking, just thought. It's also really really fun


At some point I'm going to have to try some of the more obscure puzzles Doug has mentioned solving - 1.1.4, 1.1.5, 1.1.6, 1.2.3, 3.3.3, 3.3.5. And 1.1.10 (I don't like the centers on that thing ). They seem very forbidding at the moment, though. I'm always amazed at Doug's patience with these beasts

I actually was just about to mention that, but I didn't want to make a quadruple post.

I was thinking...

1st place on time is 10 points, second is 9, third is 8, so on and so forth. Then you could have 3 categories. Speed, Moves and Overall.

Yeah, I figured out the problem you mentioned by the way.

And for the one where I halved your time, I used a method similar to solving the tips on the Dogic 10/20 color.

Oh! That's a very different way than what I did. Sounds like an efficient way to do it though.

I did Gigaminx in 12:42 just now Splits were about 3:45 and 10:40. Arnaud's edge pairing is fast

EDIT: I did Gigaminx in 592 moves (trying for fewest moves). Edges were the hardest to optimize, although I could have probably done a little more for centers. I tried to always find three-cycles with two setup moves or less. Then the megaminx LL was easy. I think the limit for a human doing a fewest moves solve with reduction is over 500.

It's funny that I just discovered this puzzle applet a few days ago.
So far, I've gotten 1 layer on the Pyraminx Chrystal in 10:37 (second try), and the dino cube first try ever in 2:01.

Last night I walked a drunk Doug through one of the 20 sided ones. Which ever one took him 4 hours and 2000+ moves.

Any way we could get bump cubes?

Also I finished a lot of the vertex turning 20 sided puzzles.

Requesting a way to see which puzzles have not been solved on the main page. Maybe an asterisk on the ones that have been solved linking to the high scores list.

Finished one of the odd colored ones.



It didn't register though. I'm 99% sure it's good though.

Also, for the move count, I messed up so many times that I could get at least 500 less if I needed to. Not going to though. Too much of a hassle.

Noah - if something doesn't register (or you are not sure) it's always to your advantage to ask for the certificate and either post it here or mail it to gelatinbrain. He told me that the certificates contain your name, the move count, and the time taken, so if you give him a certificate he'll have all the relevant data.

It didn't prompt my name or let me ask for a certificate. I finished it and time was still counting.

This must be a bug. I'm sorry. I will register your record manually. Tell me if you remember the exact time when you finished.

A few seconds before I took the screen shot. Just use that time. It's all good. I just wanted to finish the puzzle. Time or turn count didn't really matter to me on that solve.

I fixed the bug, but I'm not 100 % sure, because I have no capacity to solve it myself.
I hope also that someone makes a physical version of this sticker variation of Dogic.

I don't see my score up there, does it mean I have to solve it again?


Also, 4 moves, 12 seconds on the Halpern-Meier Tetrahedron.

f2420192bd271285f407d926b34c7986
f209b946738ce916d3d86799ce31e51a
cb3ce618cd34a7544f959d6e3b42976a
2fd09b6437c89e613dc27689ec135ea1

5.1.2

Your score is there. Only the place is not in order, 2.2.8d comes after 2.2.10b. I will fix this later.

Is there a way to save a solve in progress?

Leave the computer on?

Generally, if I run out of time (going/leaving class, etc), unless I got past a section by luck that I normally wouldn't be able to get past, I just close it and start over. It's just more practice.

Then there's the question whether that could be considered cheating. One could for example, recreate a 2x2 scramble while paused and find the quickest solution, then perform in on the unpaused puzzle throwing the scoreboard off a bit.


Generally I'm ok with it anyways. I'm just trying to keep up with Micheal anyways.

It should be made so the time would disappear when you hit SAVE.

I usually don't have enough time to finish a puzzle once I've begun one, unfortunately.

Which puzzles are you doing? I don't think I've ever spent more than 30 minutes on some of the simple puzzles. (Props if you're trying one of the harder ones)

Most of them I'm figuring out for the first time. I'm not using commutator (for some reason). That would probably speed me up a lot. Anyway, it's hard for me to stare at the screen for 30 minutes.

I usually have a TV show playing in the background and/or some Instant Messaging windows open. Even sometimes I'm posting on this forum or watching YouTube at the same time. Multitask if you can. It makes it less dull.

Timer didn't stop for the Super Megaminx.

http://users.skynet.be/gelatinbrain/App ... ca_f12.htm

Not going to bother with my score, just reporting it so it can be fixed.


Regards,
Noah

Hi, this new look&feel is very cool.



You can use the stand-alone version instead. The state in progress is automatically saved into the registry and restored when you relaunch. You can save in to a file too. It's only for Windows and, as I'haven't updated for a while, some recent puzzles are lacking, and it's has no score registration function.

I was goofing around and thought of this pattern.

Nice - Star Trek fan?

Nope. I was just thinking about swapping the Corner-Edge-Corner block around, and I came up with this on my Meffert's Minx. Originally I has the Yellow and White edges cycled as well, which looks better on the Meffert's Minx, but looks terrible on ones were the corners meet at a point instead of allowing some space for the edge. So I just put it like this and thought it looked nice.

I like the Star Trek pattern you have going on there. I am too intimidated by these puzzles to attempt them (still haven't figured out the last layer on the regular megaminx yet ).

A few puzzles that don't seem too hard to add.

http://twistypuzzles.com/cgi-bin/puzzle.cgi?pid=773
http://twistypuzzles.com/cgi-bin/puzzle.cgi?pid=771
http://twistypuzzles.com/cgi-bin/puzzle.cgi?pid=33
http://twistypuzzles.com/cgi-bin/puzzle.cgi?pid=397
http://www.asahi-net.or.jp/~WQ6H-TKJ/bb ... ation1.jpg

Also, could we have a news page too? Just to know when puzzles are added, and when there's updates and stuff?

I can only do the Dino Cube.. hah.. maybe I'll try the Helicopter Cube later tonight.

I don't understand how to function the third puzzle. How its solved state is defined?
For the rest, there are already functionally equivalent puzzles for each (If I understand well...).
1. Impossiball
2. 2x2x2
4. Pyraminx
5. 12-color Alex's Star

I'm rather interested to find new puzzles, and like to keep those shape variations for later.
And there are still enough unsolved puzzles for the rest of your life.
Why not try them? Are they too easy for you? Then try just to make me happy.

I will make a log for future updates(as far as I remember...).
Thank you for those suggestions.

Would it be possible to make puzzles that have irregular cuts, so that when scrambled they have that kind of chaotic look? Like the Fisher Cube, for example.

The 3rd puzzle is a 2x2. You can add many color variations stickering wise to establish permutation and orientation.

As for the unsolved puzzles. Aside from a few I know I can do, and never got around to actually spending a few hours working on them, I'm fairly sure I've solved all that I can possibly do.

My todo list
1.1.7 (I know I can do, I even wrote a tutorial about it)
2.2.8c
2.2.9c
2.2.10c
2.2.10d

I was just throwing out ideas for shape variants because I find them interesting, and I can no longer finish other puzzles. And of the ones I have yet to get around to, (all of the 20 sided ones) are more or less the same puzzle or technique.

I hope you don't think I'm complaining or anything. I'm just trying to help you by suggesting new things to do. You've already been extremely helpful and I'm very appreciative of that.

As for the updates, I don't care if it is much, and it doesn't have to be for every single thing. Even if it were a simple post here, it would be fine. I was just shocked (good way) to find the Impossiball puzzle, and had wondered how long it had been up there.

I don't know if this is dabbling into shape variations or new puzzles (or both), but if you are looking to make some new challenges, how about introducing a new shape. If any, I suggest the Rhombic Dodecahedron.
http://en.wikipedia.org/wiki/Rhombic_dodecahedron

Thank you very much, and happy puzzling,
Noah Hevey

P.S. How about you take a try at the puzzles.

I solved the Master Skewb.

2090b3211b879c0bd957b3447986f20d
b902738ee916d32c676cce30e51acb34
e611cd30a7564fb49d643be397c42fd3
9b6437c89e613dc27689ec135ea1bc43

580 Turns. 36:01. I wish I could get one of these in real life. Man I wish I had the cash to commission Okamoto if I could.

I wasn't going for speed or post count. A lot of it was experimenting with algorithms.

If anyone wants a hint, I feel special, so I'd rather not blurt it out. If you're interested in what I have to say, feel free to drop a PM.

Finished the Fat Dino Cube.

7ae376cf53a5304fb0d46239c497682f
d07d6437c89e613dc28289ed135ea1bc
436691de237485e40bfa26874ca386f3
0db946738ce916d32c6798ce31e51acb

230 moves. 9:45.

Once again, much room for improvement.

The SuperX is driving me insane.. I keep getting that edge parity.. I can't solve this thing for the life of me! Can anyone help? I've gotten every parity known to man on this thing.