how can I swap 2 edges or 2 corners on the U layer?

[cookingfat]

I don't really understand commutators, so please could someone tell me how I can swap say either the UBR and UFR corners, or the UF and UR edges without changing the rest of the cube?

 

[iainfs]

If we're talking 3x3x3, you can't do either on it's own. You must make an even number of piece swaps -- see Ryan Heise's Laws of the Cube at http://www.ryanheise.com/cube/cube_laws.html. Both of the states you describe are impossible to solve without disassembly.

If you want to swap both UBR/UFR and UF/UR, that's what known as a Jb permutation. The one I use is

(R U R' F') (R U R' U') R' F R2 U' R' U'

but there are plenty more.

Now someone wiser than myself will come along and tell you how to do it with commutators (either that or they'll call you a n00b). Or, you could learn yourself at Joel van Noort's excellent site -- http://solvethecube.110mb.com/commutators.html

 

[cookingfat]

it's a 3x3 letter cube like here

All the pieces are solved except 2 edges, I can do a PLL perm, but that obviously leaves 2 corners swapped.

There's got to be a way to do it cos the cube was solved when I bought it.

 

[blade740]

Try to find 2 identical pieces. If you only have 2 pieces that you KNOW to swap, swap those 2 and 2 identical pieces at the same time.

 

[iainfs]

The 'solution' that you're trying to reach is likely an impossible state of the cube. There must be another solution that is a legal cube state.

I've never tried one of these, but I'd image that the way to do it is to attack it with a piece of paper before getting your hands dirty. You should be able to solve the puzzle multiple ways, at least one of which will be physically possible to replicate on the cube.

I'd suggest that, unless you're really good at match the scramble, the easiest way to actually recreate your theoretical solution on the cube would be a BLD method - maybe orient first 3OP or Classic Pochmann. Don't forget you can easily rotate centres once you've solved the rest of it.

Good luck!

 

[cookingfat]

thanks, I've got to have made some mistake somewhere then.

The way I've been doing it is LBL method for F2l (I'm using pictures of some of the sides of a solved cube for reference)

I made a paper cube so I could work out where each piece goes and I've managed to complete all the letters, so I have a paper cube of how it should look solved. Then I've been trying to use classic pochmann to move the last layer pieces about. . This cube's pretty hard.

edit - tried again and now 2 opposite corners need swapping. mmmmm....

 

[iainfs]

I would think that you need to find another solution. The one you're trying to achieve is clearly not a legal state. I'd suggest you go back to your paper cube and try to figure out another arrangement that would work

 

[Vulosity]

Since your center cubies have letters on them that mark their orientation, I think that you have to rotate them using PLL's. Tis will probaly make the 3x3 solvable.

 

[cookingfat]

ok, I finally figured it out (with help from CAT13). My paper cube was correct, it had to be as there is only one solution for this cube, there are no identical pieces, no two corners have the same letters, only two pairs of edges share the same letters but they are different orientations, so swapping them wouldn't work.

By doing this algorithm >

U2 R2 D R' L B' R2 F2 U D' R' F' B U

it changes 2 edge pieces and also swaps the top and bottom centres which were both letter O, front and right centres which were both A, back and left centres which were both E.

try it on a solved cube.

then I just rotated the centres so they all aligned.

Man, this cube was certainly a challenge.

 

[blade740]

Doesn't that sorta mean "identical pieces"?

 

[cookingfat]

well the centres are two As, two Es and two Os, so I suppose they are the only identical pieces on the cube. The edges and corners are all unique.