Simultaneous Redge and Ledge -- Waterman-ish Question

[BadgerSex]

Before I actually ask a question, I'm going to tell you, I ALWAYS search for the answers to my questions before posting.

What I wanted to know was, in Waterman's description of his method, he says there was a table of algorithms that he used to simultaneously solve a redge and a ledge. Where can I find this table, or how can I generate this table myself? I know about the algorithm finders, but how do you find all the possible cases that occur?! I would have to do this before I generate any algorithms!

Thanks in advance!

 

[Kenneth]

In the Waterman documentation at Jelineks site there are scans of the original Waterman description from the 80's. I'm not sure but I guess the tables can be found there.

 

[BadgerSex]

In the Waterman documentation at Jelineks site there are scans of the original Waterman description from the 80's. I'm not sure but I guess the tables can be found there.

Thanks... It's table 8 of the algorithms section. But I really want to know how you would find this out yourself!

 

[rubiksfriend]

Many of Waterman's moves are based off commutators. This was a human-made method (unlike Fridrich LL). No computer searches, just really smart math guys. My guess would be to study cube principles and then try(?) to make up some.

 

[BadgerSex]

Many of Waterman's moves are based off commutators. This was a human-made method (unlike Fridrich LL). No computer searches, just really smart math guys. My guess would be to study cube principles and then try(?) to make up some.

Couldn't we optimise waterman with computer-generated algs?

 

[somerandomkidmike]

I made up some of my Waterman method algorithms with commutators. For me the algorithms I made up are faster than on Watermans site.

 

[somerandomkidmike]

Many of Waterman's moves are based off commutators. This was a human-made method (unlike Fridrich LL). No computer searches, just really smart math guys. My guess would be to study cube principles and then try(?) to make up some.

Couldn't we optimise waterman with computer-generated algs?

Yes we could do that. For it to fit in the URM group, you'd have to have a solver that could use M slice moves, and restrict what moves you can do. Doing this could improve it quite a bit.

 

[BadgerSex]

I made up some of my Waterman method algorithms with commutators. For me the algorithms I made up are faster than on Watermans site.

Maybe you could post some of these algorithms? Pretty please?

 

[qqwref]

Yes we could do that. For it to fit in the URM group, you'd have to have a solver that could use M slice moves, and restrict what moves you can do. Doing this could improve it quite a bit.

Give JACube a try. It handles slice turns (but not double turns) and you can restrict moves. You can also choose to optimize by face turns or quarter turns (or slice turn metric).

 

[BadgerSex]

Yes we could do that. For it to fit in the URM group, you'd have to have a solver that could use M slice moves, and restrict what moves you can do. Doing this could improve it quite a bit.

Give JACube a try. It handles slice turns (but not double turns) and you can restrict moves. You can also choose to optimize by face turns or quarter turns (or slice turn metric).

Yeah... But I can't stand entering all the cube configurations like that!!

EDIT... Um, yeah... I did find some remarkably more efficient algs using JACube. But you might have to learn more because of the lack of R/R' situations. I guess I'll just find algs for symmetrical cases...

 

[Kenneth]

Try Ron's java solver at Speedcubing, it is slow but can do slices.

 

[fanwuq]

Yes we could do that. For it to fit in the URM group, you'd have to have a solver that could use M slice moves, and restrict what moves you can do. Doing this could improve it quite a bit.

Give JACube a try. It handles slice turns (but not double turns) and you can restrict moves. You can also choose to optimize by face turns or quarter turns (or slice turn metric).

Yeah... But I can't stand entering all the cube configurations like that!!

EDIT... Um, yeah... I did find some remarkably more efficient algs using JACube. But you might have to learn more because of the lack of R/R' situations. I guess I'll just find algs for symmetrical cases...

You know for JAcube, you can just enter !! if you don't want to change any of the rest of the cube.
UB ! UF !! is equivalent of UB ! UF ! ! ! ! ! ! ! ! !

also, you can restrict to URM metric. What do you mean by "lack of R/R' situations. I guess I'll just find algs for symmetrical cases..."
it's just 1 to turn something on, 0 to put something off. then, you convert that to 0-7 numbers. This you put right at the beginning before you input the position
Read the Acube instructions file, there are many other things you can do.

What I don't understand is how to set it at the beginning so that you search for optimal QTM algs rather than HTM algs (the default). I see the instructions, but can't understand what to do.

 

[Lucas Garron]

What I don't understand is how to set it at the beginning so that you search for optimal QTM algs rather than HTM algs (the default). I see the instructions, but can't understand what to do.

You have to call the Java from command line, you can't do it inside run.bat. (Yeah, I don't know why, either. :rolleyes

 

[qqwref]

I set up a tiny little batch file (run-qao.bat) which goes:

@java -cp ACube3.jar ACube q a o

So I just double-click it and ACube starts in qtm. Very convenient. (I also made ones for htm, stm, and sqtm.)