For win - new V-Cube 7

[Markus Pirzer]

The CPC challenge is over. I got already a mail with the right solution.

It was a little bit careless from me to announce 10 Euro to win for that.
I didn't know how easy it would be to get access to my source code.
At the moment I'm looking for a better solution, maybe anyone can give me tips?

Even if it is easier than I thought, I appeal strongly to everyone not to cheat. It would be unfair to every honest competitor.

 

[Markus Pirzer]

You misspelled my nickname, there's an "n" in "Unknown" :rolleyes:

I've just corrected it, sorry

 

[cpt.Justice]

I have a question about the rules:

The competition ends as soon as the visitors counter exceeds 100.000 visitors.

Does this mean that everything achieved before the 100000th visitor can be submitted or that it is to late to submit scores once the counter reaches 100000?

Also, I think this rule is kinda hardcore:

All results which were achieved after reaching 98.000 visitors won't be published before the competition ends.

It will probably drive me crazy

 

[Unknown.soul]

I think sub-4 minutes is very possible and realistic. I just hope people realise it's not really worth the time your spending on it if you take it that far

And I know for a fact that the winning time won't be 8 minutes

I wrote 8 minutes? I typed that on my phone and I wasn't paying attention.

 

[cpt.Justice]

I think sub-4 minutes is very possible and realistic. I just hope people realise it's not really worth the time your spending on it if you take it that far

And I know for a fact that the winning time won't be 8 minutes

I wrote 8 minutes? I typed that on my phone and I wasn't paying attention.

So what is your best time then?

Edit:
I really hope people won't wait too long with submitting their high scores. Right now it looks like I'm the only one who wants to win (which is fine by me btw)

 

[PatrickJameson]

Where's the highscores?

Edit: Nevermind, found it. http://www.rubiks-zauberwuerfel.de/english/highscore_schwer.html

 

[Hunner]

4:40.7!? Ouch. I wonder what the initial state looked like.

 

[cpt.Justice]

4:40.7!? Ouch. I wonder what the initial state looked like.

It wasn't particularily lucky, my new highscore is quite a bit faster

Edit:
I just had a lucky solve, I don't know if I will ever beat this!

 

[Markus Pirzer]

I have a question about the rules:

The competition ends as soon as the visitors counter exceeds 100.000 visitors.

Does this mean that everything achieved before the 100000th visitor can be submitted or that it is to late to submit scores once the counter reaches 100000?

You can submit every score you achieved before the 100000th visitor. Emails that I got 1 day later, are also accepted. As I wrote I will publish the result 3 days after reaching 100000 visitors. This is because I don't know when this will happen and whether I have the time to immediatley update the highscores. But maybe I will publish the winner already 2 days after the end of the competition, I fear otherwise some people will get really impatient.

Also, I think this rule is kinda hardcore:

All results which were achieved after reaching 98.000 visitors won't be published before the competition ends.

It will probably drive me crazy

I added this rule to make the competition a little bit more exciting at the end and to avoid that anyone will be to sure to be the winner. And I wanted to avoid that anyone will be demotivated by a really fast time, but even in this early phase of the competition your time might be for many people very hard to beat.

 

[Markus Pirzer]

4:40.7!? Ouch. I wonder what the initial state looked like.

It wasn't particularily lucky, my new highscore is quite a bit faster

Edit:
I just had a lucky solve, I don't know if I will ever beat this!

Really lucky solves should be very rare, the programm mixes the puzzle randomly with 3200 moves (50 times more than the number of fields).

 

[panyan]

I (also as a wannabe hacker) have started programming a solver.

i was thinking about this also, but then i looked past my monitor to my vcube 7 and thought: "should i waste my time trying to make a programme to solve this or actually improve my v7 times?"

 

[cpt.Justice]

Really lucky solves should be very rare, the programm mixes the puzzle randomly with 3200 moves (50 times more than the number of fields).

By lucky I mean my move count was really low, because I managed to solve multiple pieces at once. It's not like half the puzzle was solved initially or anything.

 

[Markus Pirzer]

Really lucky solves should be very rare, the programm mixes the puzzle randomly with 3200 moves (50 times more than the number of fields).

By lucky I mean my move count was really low, because I managed to solve multiple pieces at once. It's not like half the puzzle was solved initially or anything.

880 moves for 63 pieces seems realistic. This means more than 10 moves for every piece on average.
I have a challenge for all mathematicians: Find "god's number" (the minimum number of moves which are enough to solve every possible state) for the 8x8 slide puzzle.

Your speed is also very amazing. You have an average of more than 3 moves per second, no other has more than 2 moves per second.

By the way, I've just updated the highscores.

 

[cpt.Justice]

Your speed is also very amazing. You have an average of more than 3 pieces per second, no other has more than 2 pieces per second.

By the way, I've just updated the highscores.

That's not true, is it?
3 pieces per second, 64 (or 63) pieces results in a total of 21 seconds
3 seconds per piece is more accurate

Edit: Or do you mean 3 moves per second? That makes more sense ^^ Sorry for my ignorance...

 

[Markus Pirzer]

Of course I mean moves per second.
I should read my messages better before posting them.

 

[coinman]

The Mystical crystal ball was fun, i'll use it to as a prank on all my non cubing friends. It will be fun to se how long it takes for them to figure it out
I will admit that I also hade to think for a while.

 

[Markus Pirzer]

Since it turned out that the CPC isn't safe, I made an additional cheat prevention. For a better safety I will not publish what it is and how it works. If anyone thinks he knows what it is and how it can be cracked, please let me know (per mail or PN).

I will also change my CPC from time to time. Therefore I've added an additional rule in my participation terms:
"Results must be sent within 24 hours after achievement."

Today's the last change to send older results, in future they will not be accepted.

Please send all your results immediately. Please send all your personal records, even if they're not winning times. It makes the competition more interesting to see your progress and to see if someone can achieve times which are at least in the near of the current best time.

 

[Stefan]

It is allowed to have a computer program solve it, right? I mean fully automated, i.e., recognizing the case and then computing and executing a solution. Depending on the computer/applet speed, this could maybe lead to a solve in under one second.

Really lucky solves should be very rare, the programm mixes the puzzle randomly with 3200

Why no random state generator? Rather easy for this puzzle and would lead to better quality scrambles.

 

[Markus Pirzer]

It is allowed to have a computer program solve it, right? I mean fully automated, i.e., recognizing the case and then computing and executing a solution. Depending on the computer/applet speed, this could maybe lead to a solve in under one second.

Of course this is not allowed (I should add this to the regulation), but it would be almost impossible to recognize that someone uses such a programm (as long as it reaches not a time which is absolutely impossible like the mentioned 1 second). This was one thing I didn't think about when I started this competition. How difficult do you think would it be to write such a programm?

Really lucky solves should be very rare, the programm mixes the puzzle randomly with 3200

Why no random state generator? Rather easy for this puzzle and would lead to better quality scrambles.

Because this many moves are (I think) much more than the moves which are needed to solve every state, on a average PC they are calculated very fast and on my tests I got every time a good scramble.
On a random state generator there would be the problem how to make sure that the puzzle is solvable. Do you know how I could solve this problem?

 

[Stefan]

Of course this is not allowed

So if I've been working on this I've been wasting my time? (I haven't, but suspect others have)

How difficult do you think would it be to write such a programm?

Not much, I think, though I only have little experience and only from a few years ago. The hardest part would of course be the interaction with the applet. I wouldn't run it in a browser, instead I'd write my own Java application, open a Java Frame, and run your applet in that. So I'd provide my own environment for your applet. Then I should have easy access for reading pixels and executing "mouse clicks". I've been wanting to do this with Ryan's cube simulator, too, I just don't have the time.

On a random state generator there would be the problem how to make sure that the puzzle is solvable. Do you know how I could solve this problem?

Rather simple parity check, have a look here:
http://en.wikipedia.org/wiki/Fifteen_puzzle#Solvability