Rubicon Forum
Welcome, Guest. Please login or register.
October 23, 2014, 01:45:02 PM

Login with username, password and session length
Search:     Advanced search
Forum users can use the crate icon (or [level] and [/level] tags) when writing a post, to make a direct link to a level.
5463 Posts in 238 Topics by 2430 Members
Latest Member: Shreyas
* Home Help Search Login Register
+  Rubicon Forum
|-+  Playing the Game
| |-+  Custom Puzzle Levels (Moderator: Bucky)
| | |-+  jf's Puzzles
« previous next »
Pages: 1 ... 27 28 [29] 30 31 Print
Author Topic: jf's Puzzles  (Read 164264 times)
Rene
Director
*****
Posts: 715


View Profile Email
Re: jf's Puzzles
« Reply #420 on: February 28, 2012, 12:14:15 PM »


For Tell me (gusymyf): rynuxeb
For Enigma II (bokycub): dykafuz (5 ticks) - works for Enigma as well

Logged
jf
Director
*****
Posts: 656


View Profile
Re: jf's Puzzles
« Reply #421 on: February 29, 2012, 04:05:39 AM »

For Tell Me (gusymyf) : xozabib

82 ticks

Nice work. You are learning the tricks pretty fast.
Logged
jf
Director
*****
Posts: 656


View Profile
Re: jf's Puzzles
« Reply #422 on: February 29, 2012, 04:15:20 AM »

:-:-:-:-:-: New Puzzle

Well, well, well. Pretty easy puzzles there, huh? Now I'll give you the real puzzle. I hope this one gives you all as much trouble as it gave me.  Wink

Come one and all. Listen to her lovely voice. Direct from the City of Light, I present to you the beautiful, the one and only, La Chanteuse: nodolur
Logged
Werbad
Director
*****
Posts: 422

jonas_88_rehn@hotmail.com
View Profile Email
Re: jf's Puzzles
« Reply #423 on: February 29, 2012, 07:12:08 AM »

I'm going to assume that there is a way to deterministically solve it, without resorting to a statistical solution. Still have no clue how one would do that tho, this one's gonna take a while to figure out.
Logged
Werbad
Director
*****
Posts: 422

jonas_88_rehn@hotmail.com
View Profile Email
Re: jf's Puzzles
« Reply #424 on: February 29, 2012, 10:11:26 AM »

That was a tough one. Great puzzle!
For La Chanteuse (nodolur): punigev

It took me quite a while to figure out a way to do this, and if it wasn't for imagethe F crate already on the matcher I might've still been sitting here trying to figure it out.

The solution I came up with was basically:
imageI take a random crate and generate a sequence of 16 crates using this and dropping it on the matcher, if I get a sequence of crates greater than the target value followed by a sequence of lower crates the first lower value will be the target value, assuming the crate picked was an F.

There are some false flags which are handled by taking the potential target, subtracting the random crate and checking if that value is greater than the target. If the random crate was an F the resulting crate will be Target+1, which will be greater in all cases but F, for which it's already solved. If the crate was anything other than F the result will be 0, which obviously won't ever be greater than the target.

If no false flag was triggered I deliver the crate to the matcher, twice to lock it in place since I don't halt the system.

Logged
jf
Director
*****
Posts: 656


View Profile
Re: jf's Puzzles
« Reply #425 on: February 29, 2012, 06:55:50 PM »

That was a tough one. Great puzzle!
For La Chanteuse (nodolur): punigev

Great work, Werbad! It's one thing to figure out how it can be solved, and another thing to build the moving parts to do it.

My solution works differently:

imageI depend on a sequence of crates starting at 1 and counting up, instead of starting at F and counting down, and I only use a sequence of 15.

imageThe first crate is duplicated, which is harmless. The sequence I am hoping for then is 1, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F. Note that there is no zero crate. I drop these onto the matcher and divide the result into left crates, L, and right crates R.

imageIf the output is (a series of) L, followed by (a series of) R, then it must have started with a 1 crate, and the last of the L series is the target value.

imageIf the output is L, followed by R, followed by at least one more L, then it could not have started with a 1 crate.

imageIf the output is only L, and no R crates at all, then the results are inconclusive. It could be 0, 8, 0, 8, ... dropping on a 9, for example, or 4, 8, C, 0, 4, 8, ... dropping on a C. Or, the target crate is F.

imageIf the output is only R, and no L crates at all, then the target crate must be 0. Odd starting crates will drop sixteen non-zero values.

imageIf the output is R, followed by at least one L, then it did not start with a 1 crate.

imageAfter coming up with this plan, the hard part was making a machine that would do this. It looks for either a sequence of L followed by a sequence of R, capturing the last L, or for 16 R values. Of course, in the case of a target F, it won't find either and remains solved from the start.

For La Chanteuse (nodolur): dakocar
« Last Edit: March 01, 2012, 06:00:00 PM by jf » Logged
Rene
Director
*****
Posts: 715


View Profile Email
Re: jf's Puzzles
« Reply #426 on: March 01, 2012, 06:30:46 AM »

For La Chanteuse (nodolur): gosapur

Interesting puzzle. As Werbad, it took me a while to find an algorithm that works as well. I was trying to figure out faster ways to solve it (imagetrying to reliably fabricate a 1 crate), but could not get them to complete for all values. Once I hit on this slower algorithm, the solution was actually not that hard.
EDIT: I actually use the same algorithm as Werbad, but different solution
« Last Edit: March 01, 2012, 03:04:56 PM by Rene » Logged
jf
Director
*****
Posts: 656


View Profile
Re: jf's Puzzles
« Reply #427 on: March 01, 2012, 05:58:28 PM »

For La Chanteuse (nodolur): gosapur

That's an elegant design, Rene. I haven't worked out exactly how the trap works -- I'm guessing that imagethe timing is such that a crate lands on the left platform under the double ++ only on the correct output stream, but haven't verified that detail yet.

I also forgot to comment on your Ice Box solution dupubyv, which also has that clean aesthetic.
Logged
Rene
Director
*****
Posts: 715


View Profile Email
Re: jf's Puzzles
« Reply #428 on: March 01, 2012, 06:32:50 PM »

For La Chanteuse (nodolur): gosapur

That's an elegant design, Rene. I haven't worked out exactly how the trap works -- I'm guessing that imagethe timing is such that a crate lands on the left platform under the double ++ only on the correct output stream, but haven't verified that detail yet.

I also forgot to comment on your Ice Box solution dupubyv, which also has that clean aesthetic.

Thanks. The trap on the bottom right does all the matching work. imageThe left dozer selects L crates and converts them into their original value, the right dozer selects R crates. The trap retains the first L crate. It holds it under the door key, if and only if it has been preceeded by at least one R crate, and is not followed by another R create. Otherwise, it pops up into the door key, and bocks the mechanism. That is, it recognizes the sequence: one or more R crates, followed by only L crates; exactly the sequence generated by an F initial crate.
Logged
jf
Director
*****
Posts: 656


View Profile
Re: jf's Puzzles
« Reply #429 on: April 29, 2012, 09:55:18 AM »

*NEW PUZZLE*

There was a slip of paper taped to it with image"1 7 6 0" written on it. Other than that, I don't have anything more to tell you: Ring Ring! hanicim
« Last Edit: April 29, 2012, 10:18:33 AM by jf » Logged
Werbad
Director
*****
Posts: 422

jonas_88_rehn@hotmail.com
View Profile Email
Re: jf's Puzzles
« Reply #430 on: April 29, 2012, 11:19:55 AM »

For Ring Ring! (hanicim): haculil
Logged
jf
Director
*****
Posts: 656


View Profile
Re: jf's Puzzles
« Reply #431 on: April 29, 2012, 05:12:43 PM »

For Ring Ring! (hanicim): haculil

Well done! I wasn't expecting this puzzle to be solved so fast.

Now let's see if you can find the solution in the dark:

Ring Ring Ring kezirin
Logged
Werbad
Director
*****
Posts: 422

jonas_88_rehn@hotmail.com
View Profile Email
Re: jf's Puzzles
« Reply #432 on: April 29, 2012, 07:25:43 PM »

For Ring Ring Ring (kezirin): cypodab

Had to use a slightly slower approach there.
I think there might be a way to handle the imageF and 7 crates that I ended up with in my other solution, but I didn't manage to find a solution without gates.
Logged
jf
Director
*****
Posts: 656


View Profile
Re: jf's Puzzles
« Reply #433 on: April 29, 2012, 08:38:51 PM »

For Ring Ring Ring (kezirin): cypodab

Had to use a slightly slower approach there.
I think there might be a way to handle the imageF and 7 crates that I ended up with in my other solution, but I didn't manage to find a solution without gates.

Another nice solution -- I was thinking of an intended solution using imagemodular exponents and so I didn't even consider this possibility.

My solution to Ring Ring and Ring Ring Ring: dazahar

imageI started with x^phi(16) = x^12 = 1 in Z*16, and found that already x^4 = (x^2)^2 = 1, and all odd x mod 16 are in Z*16.
« Last Edit: April 29, 2012, 08:47:40 PM by jf » Logged
curly
Engineer
****
Posts: 120


View Profile
Re: jf's Puzzles
« Reply #434 on: April 29, 2012, 10:57:03 PM »

For Ring Ring (hanicim) : mubadyd image(didn't use your mechanism)

For Ring Ring Ring (kezirin) : satubyd image(tries all odd numbers until it finds the one that goes through the mechanism quickly)
« Last Edit: April 30, 2012, 01:29:55 AM by curly » Logged
Pages: 1 ... 27 28 [29] 30 31 Print 
« previous next »
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.0.22 | SMF © 2006-2011, Simple Machines LLC Valid XHTML 1.0! Valid CSS!