An idea similar to EJF2L (mainly for petrus users)

[Robert-Y]

Do the petrus method normally up to the last F2L pair.

1. Solve final F2L edge + permute all corners.

2. Orient all corners (can be done using 2-gen).

3. Permute all edges (can be done using 2-gen).

This could be a good idea for OH, not so sure about two hands.

Even if this is a good idea, does anyone know how to find algs using a program for the first part?

 

[fanwuq]

How to do step 1? That is the missing link for ZZ-d.

David, step one can be LUR or RUD, which are OK for OH, but recognition would be almost impossible.

 

[DavidWoner]

Do the petrus method normally up to the last F2L pair.

1. Solve final F2L edge + permute all corners.

2. Orient all corners (can be done using 2-gen).

3. Permute all edges (can be done using 2-gen).

This could be a good idea for OH, not so sure about two hands.

Even if this is a good idea, does anyone know how to find algs using a program for the first part?

Step 1 cannot be done 2gen, and thus would ruin edge orientation, or just have really ugly algs. I still don't see how this is like EJF2L either, other than that it has petrus start.

 

[byu]

Acube can help you find algs

 

[deco122392]

cls? (message too short)

edit:wait...... wouldnt cls also eliminate step 2? and then... hasnt that ll system been thought up already?

 

[Robert-Y]

How to do step 1? That is the missing link for ZZ-d.

David, step one can be LUR or RUD, which are OK for OH, but recognition would be almost impossible.

How bad is it?

 

[JLarsen]

How to do step 1? That is the missing link for ZZ-d.

David, step one can be LUR or RUD, which are OK for OH, but recognition would be almost impossible.

How bad is it?

Yeah basically this can be done in a different sort of order if you just use full MGLS with Petrus once you get to step four. You could insert the edge alone, and then you get 108 cases i think where you can solve oll in one alg. It's the same kind of thing. Besides the point of EJF2L is to add ease to blockbuilding. Not just to restrict it to step 4.

 

[ErikJ]

there would be 120 cases for step 1. not worth it.

 

[Robert-Y]

Ah shoot... to add to that, the only good part is the last step which isn't much of an advantage to normal PLL (time-wise)

 

[JLarsen]

Ah shoot... to add to that, the only good part is the last step which isn't much of an advantage to normal PLL (time-wise)

lol sorry we didn't mean to come in a decimate your ideas I swear.

 

[Robert-Y]

Ah shoot... to add to that, the only good part is the last step which isn't much of an advantage to normal PLL (time-wise)

lol sorry we didn't mean to come in a decimate your ideas I swear.

sometimes I should think more about my ideas myself before posting about them

 

[JLarsen]

Ah shoot... to add to that, the only good part is the last step which isn't much of an advantage to normal PLL (time-wise)

lol sorry we didn't mean to come in a decimate your ideas I swear.

sometimes I should think more about my ideas myself before posting about them

That and I should spell check more >_<

 

[Johannes91]

there would be 120 cases for step 1.

How did you calculate that?

 

[Robert-Y]

Hmm... I just realised that for step 1, the U corners don't necessarily have to be in place. e.g. they could be left in a position where doing a U face turn will solve the permutation.

 

[Lofty]

This is not the same as the missing link. However I don't think that it is any better. Missing link just puts the cube into <R,U> this actually solves the permutation of the 5 remaining corners. Especially after Johnannes questioned it I doubt that its 120 cases. With 4 corners you only get 2 unique cases why would it jump all the way to 120 just because you add one more corner? You could probably make up a recognition system similar to how COLL does it. I dont really know why that won't work for missing link. I guess its because you have to check more information. With COLL its only 4 stickers on the LL beyond recognizing the OLL case which is trivial (Some could argue that over time recognizing the permutation becomes trivial) But the missing link you also would have to check corners on the D layer. I don't know if this would be as bad as missing link recognition because you don't actually have to look at the 5th corner gathering all the information about the 4 corners on the U layer will tell you about the 5th on the bottom.

 

[blindalchemist]

After inserting the last edge and AUF, there are 20 corner permutations(1 solved, 6 single swaps, 3 double swaps, eight 3-cycles, and two 4-cycles), each with 3 orientations for the last slot corner. This gives 60 cases, which can be solved by 31 algorithms and 28 mirrors. The last layer can then be solved by petrus 6+7, or by 49 2-gen algorithms and 31 mirrors, 80 cases(+ 1 solved) averaging 11.71 moves.