Date: Fri, 18 Aug 89 12:21:29 EDT
From: Richard Pavelle <rp@xn.ll.mit.edu>

I took some time off this week and began playing with the cube to
teach one of my kids how to solve it. I had not tried for perhaps 5
years. To my surprise I had forgotten a few transformations while
recalling a few which are "equally difficult". It took about 10 hours
to get back to the stage where I can solve it in about 3 minutes
except for the flip of two opposite edges. I recall that we discussed,
in this forum, a nice procedure for this move many years ago and I
wonder whether anyone recalls it.

Recall the "Extended Befuddler" language:

B, F, U, D, L, R are counter-clockwise quarter twists.
Lower-case are clockwise.

I, i, J, j, K, k are whole-cule rotations agreeing in sense and axis
with B, F, U, D, L, R, in that order.

We group together sequences that are order-independent. These sequences
almost always correspond to intuitive "moves".

And now, to flip the FD and BU edges:

;;; First monoflip:
f ; Get FD edge into equator.
jUd ; Slice it to the back.
FF ; Turn the vacated slot over
JuDJuD ; and slice the cubie back into the inverted slot.
F ; Move the cubie to the top.
;;; Segue
UU ; Exchange it with the other edge to be flipped.
;;; Second monoflip:
f ; Move the new edge into the equator,
JuDJuD ; slice it to the back the long way,
FF ; turn the vacated slot over,
JuD ; and slice the cubie back into the inverted slot, the short way.
f ; Get it back to the top
;;; Coda
UU ; Un-segue
FF ; and take first edge back to the bottom.
;;; Checksum of whole-cube moves: jJJJJJ = 1.
;;; 26 qtw, 13 "moves" including half-twists and slices.

I doubt if this is minimal, but it is so intuitive that I was able to
type this sequence without a cube in my hands.

Also, to what extent have others shared my experience of forgetting
moves?

Some.