« first day (156 days earlier)      last day (1068 days later) » 

03:54
@PhiNotPi That's neat.
@PhiNotPi It's really easy to see the changing shape with this one because of the overlaid grid.
With a physical puzzle it's basically impossible to spot considering there's gaps between the pieces anyways.
Incidentally, I happened to notice that the ratios are related to the Fibonacci sequence. I wonder if those ratios give the best illusions.
I wonder if Fibonacci numbers are more likely to be involved because the illusion involves "numbers that add up in multiple ways"
Yeah, that's plausible too.
04:11
...challenge idea?
 
8 hours later…
12:18
@orlp oh that is binoxxo=)
@orlp Oooh, neat, haven't seen this game before.
I think I posted a version (what I know as binoxxo) here before, with x and o instead of blue/red
12:49
@orlp I like the hints!
Just finished 12x12! Thankfully just before I needed to leave for work!
I wonder how it would function as a two-player game. Players place their tiles one at a time and any forced tiles are automatically filled in. The last person to make a move wins or loses, and the game can end when either 1) the board is full or 2) neither player can place a tile.
can you place it everyhwere? (in your suggested game?)
13:50
@El'endiaStarman the problem with that game is that the board is largely balanced. Getting more of a single color requires a good amount of pieces, and when you have another player doing the opposite, I'm pretty sure the entire game will be even
@flawr Anywhere that is legal, yes.
@flawr That's a surprisingly hard game.
14:39
@El'endiaStarman But usually the boards have unique solutions,s o you'd need to know the whole board before you could start playing which would make it trivial, wouldn't it?
@flawr Ah, but the unique solution is enforced by the placement of a few key tiles at the start.
ahs oy ou mean starting out with an empty grid?
15:08
It would be cool if the grid automatically updates to show wich are possible=)
15:18
Yeah, that's what I mean by "forced tiles are automatically filled in".
Actually, that would mean it's impossible to make illegal moves.
15:58
@El'endiaStarman brb writing a solver :)
 
1 hour later…
17:00
does anyone know how to make box-selections in sublimetext2? (using windows, preferably using the keyboard)
@El'endiaStarman thanks, unfortunately [crtl]+[alt]+[up|down] change my screen orientation XD
@flawr Yeah, I have a similar problem on Mac.
Doesn't change my screen orientation but it does do a slow zoom-out/layout of the windows you have open.
ah, that is unfortunate=)
That said, Sublime has a file with all of the keyboard shortcuts. You could probably set it to whatever you want.
17:10
I quite liked the shortcut in notepad++, where you could normally select using shift+arrows, and box-select using shift+alt+arrows
Preferences -> Key Bindings. Look for ctrl+alt+up in there.
oh thank you!
gonna get dinner, I'll report later whether that worked=)
Alright. :)
18:16
Whatssss thisss, Precious?
I'm writing code that generates basal ganglia networks given some parameters (number of channels, numbers of cortex populations, etc.)
19:15
@El'endiaStarman hm, this does not seem to provide a rectangle selection
whatever I found another solution
@flawr Well, from what I can tell, it uses the multiple-cursors feature, so you can use shift+left/right to move the right edge and, I presume, the other keyboard shortcut to move the bottom edge.
Yes that is what it seems to do, but if I use the left/right arrow, it does not select character by character, but jumps whole "words"
Ah, oh well.
Maybe they changed that between v2 and v3 'cause that doesn't happen for me.
19:24
or does [alt] interfere in in another way?
That's possible, but usually it's ctrl+left/right that jumps words.
19:37
oh man I love Haskell...
19:56
I think I'll soon finished a backtracking solver for this puzzle, but it seems it is just so little code
20:13
oh no=/
code is kill?
no, bugged
but I found it=)
Bugged.....by the NSA?!
nope, by myself
(and I am not the NSA)
(yeah, a typical thing to say if you are the NSA, but for real, I am not the NSA)
this is amaaaazing=)
I'm literally the NSA
20:24
I <3 Haskell :D
barely 50 lines for solving this puzzle
20:44
@orlp I tried to write a solver for the puzzle: tinyurl.com/yc3x8qpl
^ @PhiNotPi @El'endiaStarman
currently it tries to find all possible solutions, so not optimal just to find one
can probably be heavily optimized, but it was a nice exercise in Haskell=)
Ah right I made it way too inefficient
I'm just always amazed how short those programs get=)
@flawr It doesn't output anything when the whole board is empty
it just takes too long=)
But I just found out that I made it way to inefficient, trying to optimize that=)
Or rather, it prints []
hmm curious
ah I see, another little mistake of mine
I was a little bit too enthusiastic=)
@PhiNotPi Thanks for finding those bugs, I updated it: tinyurl.com/y883sksx
21:14
Sorry for my spam, I recently started getting a little bit deeper into haskell and now finally wanted to try a little project to test my skills, and now I was very happy how it worked out:)
(and I still recommend anyone to learn haskell, especially if you have some math-affinity)
21:28
hm it seems pure backtracking is not quick enough even for 6x6

« first day (156 days earlier)      last day (1068 days later) »