Cipher thread

[Duckie]

Is anyone trying to solve the cipher and want me to list the numbers I have? I only have a bit over 200


BlaXpirit [UA]: We're collecting all the patterns here http://blaxpirit.pythonanywhere.com/lyne/. Check out the information if you want to help.

[ThomBowk]

I think you may have to wait for the regular Steam users to catch up, in the meantime this Twitter user started a pretty cool thing. https://twitter.com/WillKirkby/status/440924042040590337

[Bilut]

Thanks for that link, I've just finished regural levels so there is only that cipher on my mind now :)

[louismax1]

So a couple interesting things I've managed to determine about the ciphers:
1. Most of them start with A and consist of a 3x2 and 3x3 configuration, but I have a few that start with Z and have only a 3x2. I've also seen one that starts with X and has a 3x3 configuration consisting of a single point.
2. It may appear that these correspond to end configurations of LYNE puzzles, but this is not so; many of them have vertices with 3 edges coming out of them, which is never possible in a LYNE puzzle, as the only odd-degree vertices are the endpoints, which have degree 1. Everything else has edges that come in and out in pairs, and thus has even degree.
3. Dev Tomeh (Thomas Bowker) refers to the triangles earned as "trytes." According to wikipedia, "Some ternary computers such as the Setun defined a tryte to be 6 trits, analogous to the binary byte." (A trit is a ternary digit.) The theme of math and triangles in particular is not new to Thomas; he's on the team that made the beautiful and long-needed Flatland game in which you play as a triangle, as well as the subatomic particle-inspired Particulars, and his first game Phi is also numerically based.
Anyone want to write code to allow us to more easily share and compile a complete database of these things? It's especially hard to look at since I'm on iOS. I also find it strange the the diamond which doesn't fade and spins out when you start the game varies, but I don't know if that's relevant.

[louismax1]

Oh yeah. Here's the bizarre X one.

[Bilut]

Quick observation - compiling my 150+ unlocked configurations - I have never seen few specific connections.
What I mean is in 2x3 part of the puzzle there was never a top diagonal going from left top corner and accordingly there was never a bottom diagonal going from left bottom corner.
So the only allowed diagonals begins in left middle point.

In 3v3 part there are only diagonals starting from left middle point or right middle point.
In conclusion - it limits the amount of configurations allowed - and excludes conflicting diagonal paths.

[louismax1]

Hm. Nice observation!

[Bilut]

(+)-0-(+)
 |    /|
 1  2  3
 |/    |
(+)-4-(+)
 |\    |
 5  6  7
 |    \|
(+)-8-(+)

(+)-0-(+)-1-(+)
 |    /|\    |
 2  3  4  5  6
 |/    |    \|
(+)-7-(+)-8-(+)
 |\    |    /|
 9  A  B  C  D
 |    \|/    |
(+)-E-(+)-F-(+)

I mapped all my unlocked stuff onto those diagrams.
It seems there are no 1, 2, 4, 6, 7 and F links in a 3x3 graphs.
Also no 4 in 2x3 graphs.

So it is looking that those are all possible links:

(+)-0-(+)
 |    /|
 1  2  3
 |/    |
(+)   (+)
 |\    |
 5  6  7
 |    \|
(+)-8-(+)

(+)-0-(+)   (+)
      / \     
    3     5   
  /         \ 
(+)   (+)-8-(+)
 |\    |    /|
 9  A  B  C  D
 |    \|/    |
(+)-E-(+)   (+)

[Bilut]

To clarify - I'm talking about all 'A' sets, not Z or X. I have barely an example of them so i wont draw any conclusions about them yet.

[Mist]

A couple friends and I have compiled what we have together. We have just over 300 together. All of the 3x3's are unique whereas we only have 65 unique 2x3's. We are currently using a google document to compile our thoughts. I don't know a good way to open it up to others to avoid someone trolling and changing a bunch of values.

[Mist]

I realized I can let people view it without problems. https://docs.google.com/spreadsheet/ccc?key=0AqFYbSQjyz1ddDNjR1NlM2RTUzVQQ01fVkE5MWdIRWc&usp=sharing

[BlaXpirit]

Mist, it's very nice to store that way. But did you do all that manually? D:
Also, please specify which line each bit corresponds to. Then I can import and visualize and maybe somehow analyze your data...

Anyway, I've been working on a graphical web interface to this stuff. Not sure if it's worth anything. Gotta finish and possibly optimize it... (Submit only shows the bit string for now and doesn't actually save anything).
Oh, and it seems to work only in Chrome :/

Then if we have the data in a useful format, we can try to make some more sense of it.

[SethLios]

For anyone that's interested, I'm building up a collection of screenshots for each one I unlock:

http://steamcommunity.com/id/SethLios/screenshots/?appid=266010&sort=newestfirst&browsefilter=myfiles&view=grid

[Mist]

At BlaXpirit, there's a couple sheets to the spreadsheet. One of them numbers the lines. Then there's a 1 in a given position if a line is there and 0 otherwise. Also, is there a way to get the data not just by looking at it in the hidden menu? If there was I would much prefer that from entering it manually.

Also we've started some data analysis. For example, if the 3x3 ones are numbers we have narrowed down what some of the groups of digits have to be.

[Bilut]

my quick calculations on possible coverage of 2x3 grid

lines:  combinations
0:       1
1:       8
2:       28
3:       56
4:       70
5:       56
6:       28
7:       8
8:       1
sum:   256

Seems to me that removing few possible paths (2 diagonals and a middle bar) had lead to strangely familliar number here (max different values coded on 8bits).

Here goes 3x3 board:

lines:  combinations
0:       1
1:       10
2:       45
3:       120
4:       210
5:       252
6:       210
7:       120
8:       45
9:       10
10:      1
sum:   1024

Another nicely rounded power of 2 here :)
I hope im on to something here.

Would be nice if someone confirmed my results.

[Mist]

Yes bilut that is correct. We should get a power of 2 no matter what though. Another way to think about that calculation is that we can choose each line to be there or not to be there. So there should be 2^8 and 2^10 combinations respectively.

Also made a kinda bulky way for people to submit data to the spreadsheet I currently have going. https://docs.google.com/forms/d/1tWpUXskqtFS2f0Is3tULLMkKaHK7yPlpkJPMb1PiEsc/viewform