(I assume the scramble is preformed in solving orientation).
The edge cycle is BU-DF-(breaking to a new cycle)-UR-RD-BD-LB-RB-FL-LU-DL-UR.
The corner cycle is FRD-BRU-DBL-DLF-LUF-RFU-DRB.
[D R2 B L: R2 U R U R' U' R' U' R' U R'] // setting up BU and DF to UL and UR, then alg and undo...
I realized that I execute pretty slow for no reason, so I just increased my tps during executions. Got a 1:35 and 1:36 which is good since I wanted some good sub-2's. (Next two solves were DNFs, so maybe I shouldn't rush that much)
Fourth 4BLD attempt. Another fail, but an improvement over the former solves - I totally messed up the centers but wings and corners were (almost) fine. Maybe I should execute centers before wings. Memo was 12ish, total 26.
Third 4BLD attempt ever (in my second attempt earlier today I just forgot some of my wing memo) , the whole cube was a mess (except for some correct wings). At least my times are getting a bit better, 23 mins instead of the 30 of my first attempt :\
Just a thought I had yesterday - how about automatically checking the algorithms? For example if each case contains not just an image but also a description of the cube and so a code can just check all added algorithms.
Not sure if it's puzzle theory or software, but - how are scrambles generated?
I looked at the code at Mark 2, and it seems that 2x2, 3x3 and Pyra are random state (which seems reasonable since optimally solving them is not that hard). For NxNxN, I think it just randomly generates moves - is...
Got back to doing BLD after like half a year (and cubing in general, though I mostly do BLD). Took me a few days, but now I got a sub3, yay!
I'm again aiming for a NR, which shouldn't be too hard to get. (I think I could've done it in the last comp but I couldn't come)