ACube4x4 question

[qqwref]

Lucas, does this program allow counting in qtm?

I suggest searching:
- Rw, Lw, U
- r, l, U
- r, l, U, F

 

[reThinking the Cube]

Yeh, maybe allowing some B turns (=x'), and a few others would help shorten the solution, and still get something nice.

Okay, please be more specific and non-contradictory.

Sorry for being vague. Meant - Why not allow <B2> face turns too? Can be turned as "U2" with x' cube rotation, so the alg will still be easy to execute. I just thought you would see that without my input anyway. The program may be taking longer to solve with only 2-gen to work with. Give it a few more turns to play with, and then see if shorter algs come quicker.

Anyhow, the <U, 2L, L> search returned algs not using L:
U2 l U2 l' U2 l U2 l U l U l U2 l2 U2 l U2 l U l' U' l2 (22)
U2 l U2 l' U2 l U2 l U' l U' l U2 l2 U2 l U2 l U' l' U l2 (22)
U2 l2 U l' U' l U2 l U2 l2 U2 l U' l U' l U2 l U2 l' U2 l (22)
U2 l2 U' l' U l U2 l U2 l2 U2 l U l U l U2 l U2 l' U2 l (22)

No surprises I guess, that this search will get mostly the same results (mirrored) as the other one.

I'm not going to try brute-forcing a lot of cases. Unless maybe I can run things remote without staying logged in.

I understand, but even a couple might find something interesting.

 

[masterofthebass]

I'm not going to try brute-forcing a lot of cases. Unless maybe I can run things remote without staying logged in.

lrn2screen. You can run things forever w/o staying connected.

 

[Lucas Garron]

Are you claiming that the pure OLL parity is the shortest possible way to change edge parity?

If only an odd permutation exists (not double parity, ok), then yes, the pure odd permutation alg is the shortest possible way.

cmowla, I have no doubt you have some mental abilities, but I don't think anyone here believes your proofs and statements, because a lot of them have been uneducated, false, contradictory, or falsely substantiated.
At this point, anything you say that you can't have concrete evidence for is like the boy who cried wolf. I believe you can count enough to know you have a 22q alg, but that's about it.

Would you mind restating your "proof" for this?

I seem to recall you saying it always had to be a setup of some sort, but I don't even understand what you meant. You called some alg by Stefan "different" when it really wasn't. What about an alg like l' U2 r' D2 r' B2 r D2 r' D2 r F2 r' D2 r F2 r?

 

[Lucas Garron]

Well, the way it looks to me, Stefan's alg itself:
x' (Uu)' R2' (Uu)' (l2r2R2) (Uu)' L2 (Uu) (l2r2R2)' (Uu)' (l2r2R2) (Uu) L2' (Uu)' L2 (Uu) L2' (Uu) z

looks very different than:

(Rr)2 B2 U2 (Ll) U2 (Rr)' U2 (Rr) U2 F2 (Rr) F2 (Ll)' B2 (Rr)2

Let me lend you some glasses:

Stefan's Alg: x' (Rr)' D2 (Rr)' U2 (Ll)' U2 (Ll) U2 (Rr)' U2 (Ll) U2 (Ll)' U2 (Ll) U2 (Ll)

And I didn't know there was a "school" that I could go to be "educated" enough to make "educated" comments. 99% of all I know has been self-learned and not from outside sources either (the cube solver program I mentioned in reThinker's thread too).

Everyone here is self-educated. Sure, we learned more from other sources than figuring out ourselves, but that doesn't matter. We just don't make claims about the nature of cubes unless we actually know what we're talking about. Intuition and hunches are fine, as long as they're qualified. Everyone even makes mistakes, it's just about the encouraging, progressive, collaborative attitude.

 

[rachmaninovian]

would clement gallet's solver be as powerful as ACube, less, or the same?

 

[deadalnix]

To answer that, I need to know where to find ACube.

Clement solver is very powerful, but require IT competenties. I know he recently made a new version, but havn't tested it yet.

 

[reThinking the Cube]

Lucas, does this program allow counting in qtm?

I suggest searching:
- Rw, Lw, U
- r, l, U
- r, l, U, F

<R*,r*,U*,u*,F*,f*> would be my 1st choice, to just be able to find "something" a.s.a.p. Afterwards, searching for easy-turning optimizing algs, with qqwref's turn suggestions is a BIG+1 for me.

Also - don't specify any corner placement for the 5 corners: UFR ULF UBL URB DRF. Those turns <R*,U*,F*> will get the maximum benefit if you set it up that way.

After that:

The 4-cycle dbl parity (UFl->UBl->UFr->UBr) should also be looked at (rather than just limiting this to the 2-cycle transposition UFr<->UFl), but there are others that could produce even better OLL parity algs. Edge placement (for all 5 remaining dedges) can be even-cycle commutes to change parity, but as was previously pointed out (by Lucas), there is a problem with not being able to force software solvers to keep the dedges paired up, into their permuted locations. Reduction redux. Maybe you could use another program that parsed the Acube 4x4x4algs, and then tested the positions generated to see if the dedges are all still paired up. Worry about that later.

Lucas, I'm STOKED! Maybe some results will be coming soon?

http://www.youtube.com/watch?v=gq_5QZDpY1Y

 

[qqwref]

<R*,r*,U*,u*,F*,f*> would be my 1st choice, to just be able to find "something" a.s.a.p. Afterwards, searching for easy-turning optimizing algs, with qqwref's turn suggestions is a BIG+1 for me.

You don't seem to understand anything at all about solving programs - try playing around with one sometime. The more moves you allow, the longer it takes to get up to a certain movecount (since it has to check more possibilities). Allowing six moves (and since your notation makes no sense I can't even tell what moves you intended) is probably too much to get any solution reasonably quickly. The setups I suggested were attempts to minimize the moves used while having a high probability of finding short solutions.

 

[sz35]

Send me one copy please:
[email protected]

 

[reThinking the Cube]

<R*,r*,U*,u*,F*,f*> would be my 1st choice, to just be able to find "something" a.s.a.p. Afterwards, searching for easy-turning optimizing algs, with qqwref's turn suggestions is a BIG+1 for me.

You don't seem to understand anything at all about solving programs - try playing around with one sometime. The more moves you allow, the longer it takes to get up to a certain movecount (since it has to check more possibilities).

I already know this quite well, but you just went ahead, and assumed otherwise. Teamwork Baby! It is not just the #of turns allowed that increase the solve time. Having to restore everything, or being able to leave some pieces unresolved also makes a big difference. Finding PLL algs for LL. Makes a difference in alg length, and solve time, depending on BOTH #of turns allowed, and how many other pieces are unsolved (greyed out). What I wanted to emphasize though, is that this case has a whole bunch of pieces that are not solved yet, and Acube should be able to take advantage of that.

Allowing six moves (and since your notation makes no sense I can't even tell what moves you intended) is probably too much to get any solution reasonably quickly. The setups I suggested were attempts to minimize the moves used while having a high probability of finding short solutions.

I wasn't finding fault with your suggestion. Did you see the "BIG+1"?. My turn notation is standard enclosed in angle brackets like I think it should be. "*" means any, and that shouldn't have been too hard to decipher. No need to try all the turns from that set at the same time, but I think this set of turns will better take advantage of the unsolved corners that were specified, so what I am suggesting is - let's try 3, then 4,5,..... and more from that set first. All of this seems very reasonable to me.

 

[Lucas Garron]

What I wanted to emphasize though, is that this case has a whole bunch of pieces that are not solved yet, and Acube should be able to take advantage of that.

And Stanford should be able to set up networks that don't run out of IPs in dorms, ever. But saying that somebody should be able to implement something doesn't magically make it happen.

Ignore means ignore. For most solvers, including all decent ones I've ever heard of, this means you can't use constraints like "ignore the position of this piece, and let it split up from its partner, but make sure they come together at the end." (ksolve does bandaging, though.)
Most solvers use prune tables based on permutation, and I'm not sure those stay nearly as efficient for that purpose.

My turn notation is standard enclosed in angle brackets like I think it should be. "*" means any, and that shouldn't have been too hard to decipher.

No, it's not. The standard is from group theory, where * means something else entirely. If that's what you meant, just leave out the stars, and you're fine.

No need to try all the turns from that set at the same time, but I think this set of turns will better take advantage of the unsolved corners that were specified, so what I am suggesting is - let's try 3, then 4,5,..... and more from that set first. All of this seems very reasonable to me.

With "let's" do you mean "someone else should interpret my vague request and do it for me"?



qq: ACube 4x4x4 works by leaving DBR fixed. Allowed moves:
U F L / u d f / b l r / Uu Ff Ll / ud' fb' lr' / Uud' Ffb' Llr'

- Rw, Lw, U
Impossible.

- r, l, U

Spoiler

l' U2 l r2 U2 r' U2 r U2 l' U2 r U2 r' U2 r U2 l (18)
l' U2 l2 r U2 l' U2 l U2 r' U2 l U2 l' U2 l U2 l (18)
l' U2 l' U2 l U2 l' U2 r U2 l' U2 l U2 l2 r' U2 l (18)
l' U2 r' U2 r U2 r' U2 l U2 r' U2 r U2 l' r2 U2 l (18)
r U2 l U2 l' U2 l U2 r' U2 l U2 l' U2 l2 r U2 r' (18)
r U2 l2 r' U2 l U2 l' U2 r U2 l' U2 l U2 l' U2 r' (18)
r U2 l' r2 U2 r U2 r' U2 l U2 r' U2 r U2 r' U2 r' (18)
r U2 r U2 r' U2 r U2 l' U2 r U2 r' U2 l r2 U2 r' (18)
Done. (8 solutions found)

- r, l, U, F

Spoiler

U2 l2 U2 F2 l F2 l' F2 l2 U2 r U2 r' F2 l (15)
U2 l2 U2 F2 r U2 l' U2 r2 F2 r F2 l' F2 l (15)
U2 r2 U2 F2 l' U2 r U2 l2 F2 l' F2 r F2 r' (15)
U2 r2 U2 F2 r' F2 r F2 r2 U2 l' U2 l F2 r' (15)
F2 l2 F2 U2 l' U2 l U2 l2 F2 r' F2 r U2 l' (15)
F2 l2 F2 U2 r' F2 l F2 r2 U2 r' U2 l U2 l' (15)
F2 r2 F2 U2 l F2 r' F2 l2 U2 l U2 r' U2 r (15)
F2 r2 F2 U2 r U2 r' U2 r2 F2 l F2 l' U2 r (15)
l U2 l' U2 r U2 r2 F2 l' F2 r U2 F2 l2 F2 (15)
l U2 r' F2 r F2 l2 U2 l' U2 l U2 F2 l2 F2 (15)
l' F2 l F2 r' F2 r2 U2 l U2 r' F2 U2 l2 U2 (15)
l' F2 r U2 r' U2 l2 F2 l F2 l' F2 U2 l2 U2 (15)
r F2 l' U2 l U2 r2 F2 r' F2 r F2 U2 r2 U2 (15)
r F2 r' F2 l F2 l2 U2 r' U2 l F2 U2 r2 U2 (15)
r' U2 l F2 l' F2 r2 U2 r U2 r' U2 F2 r2 F2 (15)
r' U2 r U2 l' U2 l2 F2 r F2 l' U2 F2 r2 F2 (15)
Done. (16 solutions found)

Disappointing. Time for ksolve.

 

[reThinking the Cube]

Ignore means ignore. For most solvers, including all decent ones I've ever heard of, this means you can't use constraints like "ignore the position of this piece, and let it split up from its partner, but make sure they come together at the end." (ksolve does bandaging, though.)
Most solvers use prune tables based on permutation, and I'm not sure those stay nearly as efficient for that purpose.

I agree that this is a challenge, but I have already suggested some possible workarounds. Just because it is a challenge, doesn't make it impossible.

My turn notation is standard enclosed in angle brackets like I think it should be. "*" means any, and that shouldn't have been too hard to decipher.

No, it's not. The standard is from group theory, where * means something else entirely. If that's what you meant, just leave out the stars, and you're fine.

Come on Lucas. Are you serious! <---- Flagrant, improper, confusing, and vague use of the factorial symbol. BTW - I can site more than a few examples where you use <r,Rw,Rr> all to mean the same thing. Even though the dissertation from M.G. claims otherwise.

Lucas usually uses the SiGN notation, which is not in any way ambiguous or confusing (although you do have to know it is being used to follow it, since it's not the same as the other notation). Unlike the notation you're familiar with, it is actually consistent with the 3x3 notation, in that r is a double-layer turn.

No it's not. It is the abbreviation for "rat".

I CAN do it this way too, but prefer to post respectfully, OK.

So from now on, IMO - ONLY Stefan Pochman should be allowed to make posters look less than adequate using his patented Pochman™ postpicky methods.

No need to try all the turns from that set at the same time, but I think this set of turns will better take advantage of the unsolved corners that were specified, so what I am suggesting is - let's try 3, then 4,5,..... and more from that set first. All of this seems very reasonable to me.

With "let's" do you mean "someone else should interpret my vague request and do it for me"?

OK. You got me on the use of "let's", which implies then that I (incorrectly?) assumed that this was at least somewhat of a team effort here. Making suggestions, sharing ideas, results, feedback, problem solving ..... whatever. My bad. Let's make sure to not let that happen again.

But how am I being "vague"? Look at these posts where I have tried, again, and again, and again, to make the idea behind this request as clear as possible, without lashing out at anybody.

http://www.speedsolving.com/forum/showpost.php?p=290107&postcount=1

http://www.speedsolving.com/forum/showpost.php?p=290797&postcount=20

http://www.speedsolving.com/forum/showpost.php?p=292339&postcount=58

http://www.speedsolving.com/forum/showpost.php?p=294116&postcount=90

http:///www.speedsolving.com/forum/showpost.php?p=296127&postcount=99

Lucas: Have I been wrong to assume that you enjoy searching for, and finding these types of algs?

I don't want you to just "do it for me".
I wish more posters would chime in, and make it their request too.
That way this becomes the forum's desire to discover, and not just mine.

 

[aronpm]

No, it's not. The standard is from group theory, where * means something else entirely. If that's what you meant, just leave out the stars, and you're fine.

Come on Lucas. Are you serious! <---- Flagrant, improper, confusing, and vague use of the factorial symbol. BTW - I can site more than a few examples where you use <r,Rw,Rr> all to mean the same thing. Even though the dissertation from M.G. claims otherwise.

The symbol '!' has been in use since, I believe, the 15th century. The notation 'n!' for factorials, has been in use since 1808. I fail to see how that compares. Also, cite.

4 > 12 ' 16, and you better know my notation is standard.

 

[reThinking the Cube]

The symbol '!' has been in use since, I believe, the 15th century. The notation 'n!' for factorials, has been in use since 1808. I fail to see how that compares.

So you failed to see the point of my wity sarcasm, got a little brain KRAMP that took you back to 1808, and felt compelled to post anyway. What where you going to teach me about math? You even failed to notice in your rush to show your stuff, that I didn't even use the term "n!", but rather the seriously factorial term: "serious!". So tell me, since when, has "serious!" been in use? Will you also fail to see how that factorial compares?

I will let you in on a little secret aron, shhhh...i...will...whisper..it...so nobody will hear.....shh............"you have failed to see many things, but you just don't know it yet".

http://www.speedsolving.com/forum/showpost.php?p=291558&postcount=33

There is no need for a parity algorithm to be used before the last slot. It is pointless.

So if you think that it is pointless, why are you posting what amounts to insults, in this thread?

How about bringing some positive contribution next time, instead of trying to make yourself look smart - by just being a SMART*S*?

 

[Swordsman Kirby]

<R*,r*,U*,u*,F*,f*>

What the, Kleene star? I think you just mean:

<R, r, U, u, F, f>

 

[rubixfreak]

could you upload the programm on rapidshare please?
or copy and paste the source code so that we can compile it ourselves?

 

[reThinking the Cube]

<R*,r*,U*,u*,F*,f*>

What the, Kleene star? I think you just mean:

<R, r, U, u, F, f>

http://www.robertnowlan.com/pdfs/Kleene,%20Stephen%20Cole.pdf

excerpts:

In the 1960s I was privileged to attend a course in metamathematics taught by Stephen Kleene, although at the time, I wasn’t certain it was a pleasure. In an eight-week summer session, he took us for a brisk romp through his Introduction to Metamathematics. The material is on three levels, mathematics, logic, and metamathematics. We had to keep track of a great number of symbols. To distinguish them Kleene used three different pieces of colored chalk, one for each of the levels. Even so, on one occasion he said, “We will denote … by a star.” Someone pointed out that he had already used a star to represent something else. Without a moment’s hesitation, he said, “OK, we’ll represent …by a huge star.” On several occasions Kleene would explain some theorem or share an insight, and muse to himself, “Who first thought of this?” Then brightly say, “Oh, yes, I did.”

In the first chapter of Mathematical Logic, Kleene gives students a fair warning: “It will be very important as we proceed to keep in mind this distinction between the logic we are studying (the object logic) and our use of logic in studying it (the observer’s logic). To any student who is not ready to do so, we suggest that he close the book now, and pick some other subject instead, such as acrostics or beekeeping.”

Quotation of the Day: “Metamathematics must study the formal system as as system of symbols, which are considered wholly objective. This means simply that those symbols are themselves the ultimate objects, and are not being used to refer to something other than themselves. The metamathematician looks at them, not through and beyond them; thus they are objects without
interpretation or meaning.” – Stephen Cole Kleene.

 

[rachmaninovian]

gah, just use the notation that everyone uses.
<R, r, U, u, F, f> is perfectly understandable and needs no Kleene star. moreover, the cube wasn't born in the 1960s...

 

[reThinking the Cube]

gah, just use the notation that everyone uses.
<R, r, U, u, F, f> is perfectly understandable and needs no Kleene star. moreover, the cube wasn't born in the 1960s...

Yeh, I agree. I wasn't trying to change notation with "*", just trying to convey a clearer meaning, and for that reason, I will refrain from using it. What get's my goat, is this syntax nitpicking, that is being used to degrade and demean what would otherwise be a good/great post. If it is done in jest, that is fine, but what I see here mostly are trivial syntax and terminology definitions, being used as cheap putdowns directed at the poster. An Ape like me can only take so much.

http://www.youtube.com/watch?v=AMag1h8Rc88

Yeh, I agree. I would much rather post about a new software breakthrough that has yielded better parity algorithms.