2012-10-14T00:47:05 *** smjm has quit IRC (Quit: Konversation terminated!) 2012-10-14T00:55:55 *** coeus has joined #aichallenge 2012-10-14T02:15:05 *** coeus_ has joined #aichallenge 2012-10-14T02:17:54 *** coeus has quit IRC (Ping timeout: 260 seconds) 2012-10-14T02:21:30 *** coeus has joined #aichallenge 2012-10-14T02:23:40 *** coeus_ has quit IRC (Ping timeout: 245 seconds) 2012-10-14T02:55:12 *** coeus_ has joined #aichallenge 2012-10-14T02:57:57 *** pairofdice has joined #aichallenge 2012-10-14T02:58:33 *** coeus has quit IRC (Ping timeout: 272 seconds) 2012-10-14T03:44:23 *** epicmonkey has joined #aichallenge 2012-10-14T03:57:16 *** coeus has joined #aichallenge 2012-10-14T03:59:09 *** coeus_ has quit IRC (Ping timeout: 246 seconds) 2012-10-14T04:00:34 *** antimatroidl has joined #aichallenge 2012-10-14T04:07:28 *** coeus_ has joined #aichallenge 2012-10-14T04:10:32 *** coeus has quit IRC (Ping timeout: 255 seconds) 2012-10-14T04:10:51 *** coeus has joined #aichallenge 2012-10-14T04:12:00 *** coeus_ has quit IRC (Ping timeout: 245 seconds) 2012-10-14T04:46:12 *** coeus_ has joined #aichallenge 2012-10-14T04:49:34 *** coeus has quit IRC (Ping timeout: 244 seconds) 2012-10-14T05:00:32 *** mcstar has joined #aichallenge 2012-10-14T05:02:29 *** coeus has joined #aichallenge 2012-10-14T05:04:22 *** iglo has joined #aichallenge 2012-10-14T05:04:33 *** coeus_ has quit IRC (Ping timeout: 244 seconds) 2012-10-14T05:09:58 *** UncleVasya has joined #aichallenge 2012-10-14T05:10:57 http://richardminerich.com/2010/12/an-f-ant-colony-simulation-in-silverlight-4-0-with-dynamic-ai-loading/ 2012-10-14T05:12:09 *** foRei has joined #aichallenge 2012-10-14T05:21:21 no silverlight here 2012-10-14T05:29:35 *** UncleVasya has quit IRC (Read error: Connection reset by peer) 2012-10-14T05:34:55 *** coeus_ has joined #aichallenge 2012-10-14T05:37:07 *** coeus has quit IRC (Ping timeout: 265 seconds) 2012-10-14T05:39:51 *** iglo_ has joined #aichallenge 2012-10-14T05:41:50 *** iglo has quit IRC (Ping timeout: 240 seconds) 2012-10-14T05:46:16 *** coeus has joined #aichallenge 2012-10-14T05:48:32 *** coeus_ has quit IRC (Ping timeout: 260 seconds) 2012-10-14T06:28:05 *** coeus has quit IRC (Ping timeout: 260 seconds) 2012-10-14T06:49:43 *** epicmonkey has quit IRC (Remote host closed the connection) 2012-10-14T06:52:08 *** epicmonkey has joined #aichallenge 2012-10-14T06:55:32 *** epicmonkey has quit IRC (Remote host closed the connection) 2012-10-14T06:55:47 *** epicmonkey has joined #aichallenge 2012-10-14T07:22:41 *** Scooper has joined #aichallenge 2012-10-14T07:26:31 *** xScooper has joined #aichallenge 2012-10-14T07:29:11 *** Scooper has quit IRC (Ping timeout: 260 seconds) 2012-10-14T08:07:18 *** UncleVasya has joined #aichallenge 2012-10-14T08:42:40 *** red1ynx has joined #aichallenge 2012-10-14T08:47:15 Hi all! Maybe you know when is next AI competition planned? 2012-10-14T08:53:17 it isnt 2012-10-14T08:56:10 ): 2012-10-14T09:18:15 *** mleise has joined #aichallenge 2012-10-14T09:22:14 *** red1ynx has quit IRC (Quit: Page closed) 2012-10-14T09:23:43 *** coeus has joined #aichallenge 2012-10-14T09:36:41 *** coeus has quit IRC (Read error: Connection reset by peer) 2012-10-14T09:41:49 *** coeus has joined #aichallenge 2012-10-14T09:44:12 *** thestinger has joined #aichallenge 2012-10-14T09:52:49 *** sigh has quit IRC (Remote host closed the connection) 2012-10-14T10:42:26 *** thestinger has quit IRC (Ping timeout: 245 seconds) 2012-10-14T10:44:42 *** thestinger has joined #aichallenge 2012-10-14T12:03:36 *** UncleVasya has quit IRC (Quit: UncleVasya) 2012-10-14T13:32:31 mleise: great, i rewrote that fast algorithm in haskell, in an easy/idiomatic way 2012-10-14T13:32:42 now i do understand what it does 2012-10-14T13:32:54 ofc, it is slow, but thats not the point 2012-10-14T13:33:02 like with everything: once you write it down it becomes clear 2012-10-14T13:33:17 so what does the algorithm reveal ? 2012-10-14T13:33:23 the other codes are very cryptic 2012-10-14T13:33:33 have you looked at any of the kuduku versions? 2012-10-14T13:33:41 -.- you have no measurement of cryptic until you see mine 2012-10-14T13:33:43 no 2012-10-14T13:33:55 kudoku* 2012-10-14T13:34:06 mleise: nah, i dont think so 2012-10-14T13:34:13 to be honest, I haven't looked at any implementations other than those we posted here 2012-10-14T13:34:21 let me paste something 2012-10-14T13:35:49 mleise: decypher this http://hpaste.org/76278 2012-10-14T13:36:30 damn, it isnt that crytic anymore... 2012-10-14T13:37:38 so basically, i rewrote that without arrays 2012-10-14T13:37:50 no, it isn't _that_ cryptic. some more line-breaks and readable var names and you are set 2012-10-14T13:37:52 *** capa has joined #aichallenge 2012-10-14T13:38:37 printf("\n%i solutions\n",Node[81-clues]) ??? 2012-10-14T13:39:07 Hi, will web have another challenge this year ? 2012-10-14T13:39:17 mleise: lol, not quite 2012-10-14T13:39:24 you have no idea what that code does 2012-10-14T13:39:33 and what is the abstraction it started out with 2012-10-14T13:39:49 and the loops arent that trivial either, they are written with labels 2012-10-14T13:40:01 web have? 2012-10-14T13:40:08 we have** 2012-10-14T13:40:11 no, web wont have, unfortunately 2012-10-14T13:40:47 maybe in spring next year 2012-10-14T13:40:57 or not 2012-10-14T13:41:08 hmm, ty 2012-10-14T13:50:57 *** smjm has joined #aichallenge 2012-10-14T13:50:57 *** smjm has joined #aichallenge 2012-10-14T14:04:06 *** capa has quit IRC (Quit: Page closed) 2012-10-14T14:05:56 *** Simas_J has joined #aichallenge 2012-10-14T14:14:41 *** Simas_J has quit IRC (Ping timeout: 246 seconds) 2012-10-14T14:15:12 mleise: the 50 easy puzzles done in 1398.91s 2012-10-14T14:15:22 ^^ 2012-10-14T14:15:36 it was time for a coffee 2012-10-14T14:15:40 *** Simas_J has joined #aichallenge 2012-10-14T14:15:45 or two 2012-10-14T14:16:05 im drinking coffee replacement 2012-10-14T14:16:15 tastes horrible 2012-10-14T14:21:52 *** alehorst has quit IRC (Read error: Connection reset by peer) 2012-10-14T14:21:55 Hello, do you know if there is some ants TCP servers still active ? 2012-10-14T14:23:29 i dont think there is, but i dont know 2012-10-14T14:25:18 *** Kingpin13 has quit IRC (Ping timeout: 240 seconds) 2012-10-14T14:29:04 what ever happened to the guy who claimed he would have won if amstan didn't fake the moon landing? 2012-10-14T14:29:11 ok, thank you 2012-10-14T14:29:16 *** ecidforiap has joined #aichallenge 2012-10-14T14:30:38 *** pairofdice has quit IRC (Ping timeout: 260 seconds) 2012-10-14T14:30:44 thestinger: still here and watching you 2012-10-14T14:46:23 *** Simas_J has quit IRC (Quit: Leaving) 2012-10-14T15:13:27 *** Garf has joined #aichallenge 2012-10-14T15:20:31 *** Accoun has quit IRC () 2012-10-14T15:26:04 is 12 levels of indentation norma ? 2012-10-14T15:27:21 l? 2012-10-14T15:27:27 not, i dont think so 2012-10-14T15:27:51 no* 2012-10-14T15:27:57 damn typos are all over the place 2012-10-14T15:28:26 mleise: it is funny, that your code gets larger and larger, mine gets shorter.. 2012-10-14T15:28:37 *sigh* 2012-10-14T15:29:00 actually I'm down to 1700 again 2012-10-14T15:29:40 down to? 2012-10-14T15:30:01 oh, you said you are at 2k once, right? 2012-10-14T15:38:34 *** Accoun has joined #aichallenge 2012-10-14T15:51:15 *** alehorst has joined #aichallenge 2012-10-14T16:04:50 *** xScooper has quit IRC (Quit: Leaving) 2012-10-14T16:22:41 *** ecidforiap has quit IRC (Quit: In girum imus nocte et consumimur igni.) 2012-10-14T16:27:15 *** Accoun has quit IRC (Ping timeout: 260 seconds) 2012-10-14T16:32:01 *** smjm has quit IRC (Quit: Konversation terminated!) 2012-10-14T16:38:40 *** Accoun has joined #aichallenge 2012-10-14T16:46:49 mcstar: right. and look at this: http://www.sudokuoftheday.com/pages/techniques-11.php 2012-10-14T16:47:15 that's exactly what we call brute force ^^ 2012-10-14T16:48:42 i know 2012-10-14T16:48:55 nishio, read about it as the last 'strategy' on another page 2012-10-14T16:50:00 did you understand the X-Wing strategy ? 2012-10-14T16:50:25 sort of 2012-10-14T16:50:34 its a special case of another 2012-10-14T16:50:54 i think i understood when i read it, i dont remember it now 2012-10-14T16:51:16 mleise: i noticed this in my code: where fun b (i,ch) 2012-10-14T17:01:40 change 'ch' to 'go' 2012-10-14T17:02:35 i dont see the joke 2012-10-14T17:02:48 i see 2012-10-14T17:02:55 where fun b i go 2012-10-14T17:08:38 *** amstan has joined #aichallenge 2012-10-14T17:08:38 *** ChanServ sets mode: +o amstan 2012-10-14T17:22:28 *** iglo_ has quit IRC (Remote host closed the connection) 2012-10-14T17:23:18 *** jacob_strauss has joined #aichallenge 2012-10-14T17:26:11 *** thestinger has quit IRC (Ping timeout: 245 seconds) 2012-10-14T17:27:10 *** foRei has quit IRC (Read error: Connection reset by peer) 2012-10-14T17:32:05 mcstar: Forcing Cains... if we were using a breadth first search, we could make use of that 2012-10-14T17:32:40 these strategies are not important when you solve by computer 2012-10-14T17:33:06 they are just mnemonic devices for the human mind 2012-10-14T17:33:16 e.g. just 'peek' into the recursion and if all options for a cell reveal that another cell will be one fixed value 2012-10-14T17:33:34 we can delay the recursion 2012-10-14T17:35:14 *** amstan has quit IRC (Quit: Konversation terminated!) 2012-10-14T17:40:46 hm, that doesn't buy us anything, right 2012-10-14T17:52:21 no, wait. when this peeking into the options of 1 cell reveals any consistent reduction in options on any other cell, we can directly apply that difference to our current board 2012-10-14T17:52:57 mleise: i mean that these strategies are unnecessary 2012-10-14T17:53:14 there is a more general and better solution 2012-10-14T17:53:25 mcstar: I don't understand you. do you mean brute forcing ? 2012-10-14T17:53:29 no 2012-10-14T17:53:42 not like you know it 2012-10-14T17:54:00 mleise: then you didnt understand that C code after all :) 2012-10-14T17:54:18 no I didn't look into it much 2012-10-14T17:55:30 but I guess you are saying that all the human approaches are made for humans and a computer program can use simpler, yet effective methods 2012-10-14T17:56:30 yes 2012-10-14T17:57:46 but at some point you have to resort to guessing/brute force, right? 2012-10-14T17:57:55 after all it is NP-complete, etc.. 2012-10-14T17:59:44 NP complete only means, that the problem space grows very fast, super-polinomially 2012-10-14T17:59:51 *** sigh has joined #aichallenge 2012-10-14T18:01:06 so does that C program you ported use brute forcing or not ? 2012-10-14T18:01:23 oh, thats not complete, there does not exist a polynimial time algorithm to solve the problem ofc 2012-10-14T18:01:40 mleise: it brute forces in the least possible way 2012-10-14T18:02:11 it captures all the 'smartness' there is in sudoku 2012-10-14T18:02:49 it is rather short for 'all the smartness' :-( 2012-10-14T18:03:43 yeah :) 2012-10-14T18:04:07 is it only using an exact-coverage algorithm ? 2012-10-14T18:05:26 yeah 2012-10-14T18:05:34 but thats all you need :) 2012-10-14T18:05:39 exact hitting 2012-10-14T18:06:16 the way sudoku is formulated as an exact hitting problem means you can apply maximum reduction at every state 2012-10-14T18:06:40 so, these heuristics are contained in that 2012-10-14T18:23:06 mcstar: which heuristics? 2012-10-14T18:24:14 the ones you mentioned 2012-10-14T18:24:31 hidden , naked pair, x-wing, chaining 2012-10-14T18:24:36 omg, so they are a total waste of time ??!?! 2012-10-14T18:24:39 yeah 2012-10-14T18:24:42 no! 2012-10-14T18:24:50 can't ... BE 2012-10-14T18:24:56 i thought you new 2012-10-14T18:24:58 knew 2012-10-14T18:25:27 i need to learn how this exact coverage stuff works 2012-10-14T18:37:51 *** thestinger has joined #aichallenge 2012-10-14T18:43:54 *** sigh has quit IRC (Remote host closed the connection) 2012-10-14T19:17:11 mcstar: I tried a "speed optimized" version of that cryptic C program I found here: http://www.setbb.com/phpbb/viewtopic.php?t=119&mforum=sudoku 2012-10-14T19:17:53 you know what ? It is quite a bit slower in the 50x20 benchmark than my old solver. 2012-10-14T19:18:56 but what about kudoku? it is based on this sudexco 2012-10-14T19:19:06 and it is faster 2012-10-14T19:19:18 hmm, never tried that 2012-10-14T19:19:36 2.0s on the benchmark 2012-10-14T19:19:46 download that, and pipe the puzzles to it 2012-10-14T19:20:39 where do i find a download ? 2012-10-14T19:21:01 click on the link... 2012-10-14T19:21:11 it is githubbed 2012-10-14T19:21:14 all the codes are 2012-10-14T19:21:53 http://attractivechaos.wordpress.com/2011/06/19/an-incomplete-review-of-sudoku-solver-implementations/ 2012-10-14T19:22:01 the blog you showed me 2012-10-14T19:22:51 mleise: what the fuck am I reading 2012-10-14T19:23:11 who writes code like that? 2012-10-14T19:24:48 thestinger: like this? http://hpaste.org/76278 2012-10-14T19:24:52 thestinger: I've seen that before with the paq8 compression tool 2012-10-14T19:25:01 everything in one 'compresse' C file 2012-10-14T19:25:11 coma operators and no whitepsace 2012-10-14T19:25:42 looks like the output of a weird compiler 2012-10-14T19:25:53 all the variables are global (really strange) with external linkage 2012-10-14T19:26:37 what is strange? 2012-10-14T19:26:45 that code 2012-10-14T19:26:58 that some people dont conform to your standards? 2012-10-14T19:27:08 a person didn't write that, no way 2012-10-14T19:27:15 unless they wrote it and then obfuscated it 2012-10-14T19:27:16 yes ofc he did 2012-10-14T19:27:23 obviously 2012-10-14T19:27:31 that went through several iterations 2012-10-14T19:27:39 and he thought it would look nice that way 2012-10-14T19:28:01 it is not 'generated' code or anything 2012-10-14T19:28:31 * thestinger has doubts 2012-10-14T19:28:36 mcstar: ok, that one is over twice as fast :-) 2012-10-14T19:28:48 finally a worthy opponent 2012-10-14T19:28:50 the C one? 2012-10-14T19:29:09 thestinger: the one from attractivechaos, not the cryptic shit 2012-10-14T19:29:22 it is not less cryptic 2012-10-14T19:29:27 just formatted more sanely 2012-10-14T19:29:53 if I leave out all the whitespace from a technical article it isn't more cryptic ? 2012-10-14T19:29:56 and has some comments... but you need other stuff to understand what the comments mean 2012-10-14T19:30:18 not the whitespace is the problem!!!! 2012-10-14T19:30:24 the code is dense and hard to follow 2012-10-14T19:30:28 regards less whitespace 2012-10-14T19:30:33 regardless* 2012-10-14T19:31:40 mcstar: it was still compiled by hand 2012-10-14T19:31:52 all functions but main removed, all variables made global 2012-10-14T19:32:00 I really doubt it was coded that way 2012-10-14T19:32:19 i dont 2012-10-14T19:32:58 i cant really imagine how a low-level guy might think 2012-10-14T19:33:07 he might found that perfectly ok or nice 2012-10-14T19:33:30 in whatever way people were writing basic or asm, i would find that horrible and hard to read 2012-10-14T19:35:54 from the author after someone fixed the formatting: "I don't think your formatting is more readable but I'm aware that many (most?) people prefer it your way and that's reason enough to post it." 2012-10-14T19:37:11 *** thestinger has quit IRC (Quit: WeeChat 0.3.9) 2012-10-14T20:22:59 *** jacob_strauss has quit IRC (Remote host closed the connection) 2012-10-14T20:48:46 *** Garf has quit IRC (Quit: Make a new plan, Stan!) 2012-10-14T20:52:26 *** antimatroidl has quit IRC (Quit: Leaving.) 2012-10-14T20:59:18 *** amstan has joined #aichallenge 2012-10-14T20:59:19 *** ChanServ sets mode: +o amstan 2012-10-14T21:00:18 *** mcstar has quit IRC (Ping timeout: 240 seconds) 2012-10-14T21:58:54 *** thestinger has joined #aichallenge 2012-10-14T22:57:57 *** sevikkk has quit IRC (Read error: Operation timed out) 2012-10-14T23:02:18 *** sevikkk has joined #aichallenge 2012-10-14T23:49:40 *** antimatroidl has joined #aichallenge