Teri and Sandy 1LLL Sheet WIP

[teri2769]

me and sandy are making a 1LLL sheet is is going to be very good soon i hope we are like 1/3 of the way through i think??????

long link: https://docs.google.com/spreadsheets/d/1rOYlKT4PWPVodZAQnWv6OWmBvxGLcjgrEPA-Ca9aEf0/edit?usp=sharing
short link: https://bit.ly/TeriSandy1LLL

submissions for cases not yet on the sheet: https://bit.ly/1LLLSubmissions
list of submissions: https://bit.ly/1lllsubmissionsheet

We are genning manually for every case

Comment on cases if you think you have a better alg

 

[tungsten]

are you planning on memorizing these? or is this purely for reference for cubers

 

[IsThatA4x4]

This will be awesome! All the 1LLLs in 1 place is pretty cool. I'll definitely be learning some easy cases from this sheet at some point in the future.

 

[GRVigo]

@teri2769 This is a lot of work!!!

I can help you. If you send me a text file with an algorithm per line, I can use a special tool I made to audit algorithms for my solver. I can compare your list with @Jack314 1LLL algorithms list to assign automatically algorithms to each case and check if they work.

With that text file it would take me less than a minute to check the complete algset!

Ask me if you want more details.

 

[teri2769]

@teri2769 This is a lot of work!!!

I can help you. If you send me a text file with an algorithm per line, I can use a special tool I made to audit algorithms for my solver. I can compare your list with @Jack314 1LLL algorithms list to assign automatically algorithms to each case and check if they work.

With that text file it would take me less than a minute to check the complete algset!

Ask me if you want more details.

I use embeded visualcube images generated from the alg in the cell next to it to show the images in my sheets, so I dont neeed to check whether the algs work or not, as the images auto update to whatever alg it is next to

 

[GRVigo]

I use embeded visualcube images generated from the alg in the cell next to it to show the images in my sheets, so I dont neeed to check whether the algs work or not, as the images auto update to whatever alg it is next to

OK, that's nice.

In the past I recopilated as many algorithms as I could, and put it all together in a list (a text file with 6000 or 7000 algs). I have reference algsets with only an algorithm per case, and when I want an algset with as many algorithms per case as possible, I have a software that automatically searchs in the big list for algorithms that have the same functionality for each case, and add them to the algset.

I don't know if something like this could be interesting in your project. You can send me a text list with all your algorithms and I can merge it with my algorithms list, to get an even more complete algset.

Perhaps generatig the txt file could be a lot of extra work, in which case it's not worth it.

I think it is an incredible job that you are doing. I hope to seeing it finished soon.

 

[teri2769]

are you planning on memorizing these? or is this purely for reference for cubers

Probably will learn some, but not all. Mainly for a reference, a better resource for people with the time and motivation to actually learn it

 

[trangium]

If you aren't already, you might want to use the Movecount Coefficient Calculator to speed up the process of picking the best algs. Of course, it's not perfect, but it usually does a pretty good job of sorting the generated algs such that the fastest algs are near the top.

 

[qwr]

How do these compare to the CubeRoot 1LLL pages

 

[teri2769]

If you aren't already, you might want to use the Movecount Coefficient Calculator to speed up the process of picking the best algs. Of course, it's not perfect, but it usually does a pretty good job of sorting the generated algs such that the fastest algs are near the top.

I do not like this program and do not use it, in my experience my eye has been way better at finding good algs

 

[teri2769]

How do these compare to the CubeRoot 1LLL pages

I have not been checking his sheet at all during this process, but im pretty sure most of his algs came from Jack Love's sheet and i know that sheet is not very good. I have been using basically no reference for this process, with the exception of zbll which i based on and heavily modified juju zbll

 

[tungsten]

how are you generating these algs? are you using a program like ksolve, or are you using puzzle theory, like conjugations, commutators, etc..
ksolve will get you the shortest algorithms, but they might not be very finger-trickable. I tried implementing an alg (found with ksolve) where I could insert f2l pairs where ever was convenient, and then swapping them afterwards, but I quickly found that the algs took longer than just inserting them in regularly, and probably just isn't a good idea.

 

[qwr]

how are you generating these algs? are you using a program like ksolve, or are you using puzzle theory, like conjugations, commutators, etc..
ksolve will get you the shortest algorithms, but they might not be very finger-trickable. I tried implementing an alg (found with ksolve) where I could insert f2l pairs where ever was convenient, and then swapping them afterwards, but I quickly found that the algs took longer than just inserting them in regularly, and probably just isn't a good idea.

I don't understand the use of working on 1LLL algs except as a proof of concept or for completion's sake, but I guess people can spend their time on what they want

 

[Timona]

I don't understand the use of working on 1LLL algs except as a proof of concept or for completion's sake, but I guess people can spend their time on what they want

Like spending your time learning full VLS, who would do such a thing

 

[teri2769]

how are you generating these algs? are you using a program like ksolve, or are you using puzzle theory, like conjugations, commutators, etc..
ksolve will get you the shortest algorithms, but they might not be very finger-trickable. I tried implementing an alg (found with ksolve) where I could insert f2l pairs where ever was convenient, and then swapping them afterwards, but I quickly found that the algs took longer than just inserting them in regularly, and probably just isn't a good idea.

I mass gen algs in cube explorer and then manually edit them to be more fingertrickable in algtrans

 

[qwr]

I mass gen algs in cube explorer and then manually edit them to be more fingertrickable in algtrans

Do you think it would be useful to have a cube explorer program that could weight different moves as more or less costly? So ex. RU is considered low cost moves

 

[teri2769]

Do you think it would be useful to have a cube explorer program that could weight different moves as more or less costly? So ex. RU is considered low cost moves

No, as the way shorter algs are printed out of cube explorer may not be ergonomic the way they come out, but with modification just about any of them can be made to be ok, and using a movecount coefficient program would ignore good solutions that would just require modification from the way they come out of cube explorer.

 

[teri2769]

No, as the way shorter algs are printed out of cube explorer may not be ergonomic the way they come out, but with modification just about any of them can be made to be ok, and using a movecount coefficient program would ignore good solutions that would just require modification from the way they come out of cube explorer.

for example, L' R' B L B' R B U B' U' B' U B U' B U B' may seem not very good but a few buttons in algtrans and x' U'D' R U R' D x R U R' U' R' U R U' R U R' is very nice and quick

 

[teri2769]

x' U r U r' F' R U R' F' r U' r' F' R U R' U' x
NO ONE ELSE COULD FIND THIS