2010-12-02T00:00:03 it adds some kind of structure to it that simplifies things 2010-12-02T00:00:21 same here anti 2010-12-02T00:00:23 although i'm not quite sure about n-player zero sum games 2010-12-02T00:00:33 n-player zero sum is hard, i know 2010-12-02T00:00:53 i have a game theory library i'm making to solve for different refinements in different game classes 2010-12-02T00:01:08 cool 2010-12-02T00:01:13 i pulled sections of that out for my simultaneous bot in this 2010-12-02T00:01:47 nice, your bot did quite well, considering it's frequency of fail plays 2010-12-02T00:01:47 woah, some intense discussion going on 2010-12-02T00:02:31 always 2010-12-02T00:02:44 everyone is wrong! 2010-12-02T00:02:47 I keep rereading that solved game wiki page... I still don't see how the "outcome can be correctly predicted from any position when each side plays optimally" -- it cannot be predicted in RPS situations... I'm not sure how we can even dispute this, given it's the definition of a "solved game"? 2010-12-02T00:02:55 burny: the tree i uploaded was so broken it's hilarious how well it did :P 2010-12-02T00:03:14 virdo: that's what happens when you read one sentence and treat it as gospel 2010-12-02T00:03:22 virdo.. rps IS the prediction 2010-12-02T00:03:29 the sentence is fine 2010-12-02T00:03:39 sigh: you' 2010-12-02T00:03:46 yes ? 2010-12-02T00:03:59 virdo: consider a game of rock paper scissors, both players playing optimally is both players saying they will choose any optinion with equal probability 2010-12-02T00:04:01 enter got pressed on the quote sign - ;) 2010-12-02T00:04:09 then when the game runs, they use some random process to pick the move they play 2010-12-02T00:04:20 you don't necessarily win or draw, but that is the optimal thing you can do 2010-12-02T00:04:26 that is classified as "solved" 2010-12-02T00:04:32 I'm not disputing the players playing optimally... just if a game is solvable... 2010-12-02T00:04:54 virdo: did you read up on mixed strategies yet? 2010-12-02T00:04:58 my point is that that's the solution 2010-12-02T00:05:05 virdo, take tic-tac-toe for exmaple.. 2010-12-02T00:05:10 that game is clearly solved.. 2010-12-02T00:05:15 the argument is that finding the optimal play *is* solving the game 2010-12-02T00:05:17 but can you tell me what move I'm going to make first? 2010-12-02T00:05:21 virdo: the game is solved when you have found the optimal strategy. it doesn't matter if the optimal strategy always wins 2010-12-02T00:05:26 burny: tic tac toe doesn't have RPS... 2010-12-02T00:06:00 virdo: what distintion do you make between optimal play and solving the game? 2010-12-02T00:06:02 virdo: and "outcome" in the quote you keep citing is "expected outcome over many games" 2010-12-02T00:06:03 burny: bottom left 2010-12-02T00:06:07 jmcarthur: how can you predict the outcome if there is a RPS situation? that's teh whole idea of this right? 2010-12-02T00:06:12 ^^ 2010-12-02T00:06:23 virdo, RPS _IS_ the outcome 2010-12-02T00:06:26 virdo: expected payoffs :) 2010-12-02T00:06:32 jmcarthur: expected outcome of one game, if you are n the bayesian camp 2010-12-02T00:06:35 :P 2010-12-02T00:07:08 sigh: heh 2010-12-02T00:07:11 guys, there is a reason why RPS isn't part of the "solved games".. and no, I didn't remove it... assume two optimal players of RPS... go ahead, predict the outcome... 2010-12-02T00:07:23 am I that dense? :( 2010-12-02T00:07:26 if rock beats everythign except for rock and paper.. and scissors beats everything except scissors and rock.. etc. 2010-12-02T00:07:41 *** davidjliu has quit IRC (Quit: Page closed) 2010-12-02T00:07:41 virdo: yes, I can give you a prediction of the expected outcome 2010-12-02T00:07:42 virdo: the expected outcome of RPS is that you win 50% of the time 2010-12-02T00:07:45 dense in what? 2010-12-02T00:08:11 sigh: what is your prediction? 2010-12-02T00:08:12 then that means, you can choose one of those 3 plays.. and have an at-worst, 33% chance of losing 2010-12-02T00:08:17 you aren't dense, it's just that you are trying to approach this with no background knowledge 2010-12-02T00:08:24 virdo: what jmcarthur said 2010-12-02T00:08:27 and then: if you opponent does NOT play one of those 3 strategys.. 2010-12-02T00:08:30 you punish them, and win 2010-12-02T00:08:44 burny: are you talking about iterated games? 2010-12-02T00:08:59 actually, where did that 33% come from? 2010-12-02T00:09:04 jmcarthur, no.. I'm tlaking about an RPS situation, in a NON RPS game 2010-12-02T00:09:08 oh 2010-12-02T00:09:14 eg. planet wars 2010-12-02T00:09:17 okay 2010-12-02T00:09:38 if a map is RPS between.. attack near planet > waiting > attacing far planet.. 2010-12-02T00:09:46 and your opponent is not aware of this 2010-12-02T00:10:08 then one of those 3 plays, will be more likely to win than others 2010-12-02T00:10:17 virdo: the simple fact is that people have already thought about the issues that you brought up and have *generalised* the definition so that it works for games like rps 2010-12-02T00:10:42 the problem is that you are refusing to consider that the definition can be generalised 2010-12-02T00:11:08 nope, no longer do I refuse it... trying to read up as much on it in the background as I can... and I stand corrected :) 2010-12-02T00:11:26 :) 2010-12-02T00:14:57 virdo: let's break rps down, each player has the action sets {R, P, S}, {r, p, s} a mixed strategy over a players set is just something like (1/3,1/3,1/3) saying how likely you are to play each strategy, the pure strategy space is the Cartesian product of these {(R,r),..., (P, p)}, and the mixed strategy space is the Cartian product of all possible mixed strategies (note that set is infinite) 2010-12-02T00:14:58 you are thinking a solution to the game is an element of the pure strategy space, where u(x,y) is an element of your payoff matrix, but what we're saying is a and b can be from the mixed strategy space and you then define eu(a,b) = sum p(x,y).u(x,y) 2010-12-02T00:15:04 i'm not sure if that'll help or hurt, :P 2010-12-02T00:15:56 ah, i was late anyway :) 2010-12-02T00:21:54 while the game wouldn't be easy, an ai contest with a differential game would be cool 2010-12-02T00:22:14 differential games is like multiplayer optimal control theory 2010-12-02T00:23:18 holy crap, it hit me (while reading wiki page for differential game and getting to the part of "a pursuer and an evader") -- team based capture the flag :P 2010-12-02T00:24:03 yes! 2010-12-02T00:24:22 the primary school version, not quake 2010-12-02T00:24:28 with jails 2010-12-02T00:25:06 tons of variations of tag http://en.wikipedia.org/wiki/Tag_(game) 2010-12-02T00:25:19 1000 cutof 599 games to less 2010-12-02T00:25:19 pacmac! 2010-12-02T00:25:27 it lick winers 2010-12-02T00:25:35 it luck winers 2010-12-02T00:25:52 oh here we go 2010-12-02T00:25:57 it licks whiners? 2010-12-02T00:25:58 newer seen on TCP 2010-12-02T00:26:16 please, the game's over, it was fun 2010-12-02T00:26:21 Accoun: just because people don't test on tcp doesn't mean they don't have good bots 2010-12-02T00:26:28 canadian fail again 2010-12-02T00:26:30 Accoun: you got bocsi's and iouri's code? What were your win / draw / loss? 2010-12-02T00:26:51 Accoun: test on your local computer so it's super fast... on 99999 maps if needed! 2010-12-02T00:27:30 Accoun vs bocsimacko inly 12 games 2010-12-02T00:27:38 Accoun vs bocsimacko inlu 12 games 2010-12-02T00:27:42 only? 2010-12-02T00:27:43 oh noes! 2010-12-02T00:27:47 -_- 2010-12-02T00:27:51 Accoun vs bocsimacko only 12 games 2010-12-02T00:27:53 12 is quite a few for two specific bots 2010-12-02T00:28:00 not 2010-12-02T00:28:00 please don't repeat yourself over and over 2010-12-02T00:28:10 and the results of those 12 games? 2010-12-02T00:28:21 on true win % need 99% games 2010-12-02T00:28:26 u can du it 2010-12-02T00:28:32 byt u not want 2010-12-02T00:28:36 bat u not want 2010-12-02T00:28:39 just stop. your constant whining is annoying and pointless 2010-12-02T00:28:44 and strange map add in process 2010-12-02T00:29:06 Accoun: you can do it too -- bocsi's source code is out! iouri's too! 2010-12-02T00:29:47 if you're going to whine.. atleast use english, or elaborate sentances 2010-12-02T00:29:47 if u will plan 12 games betwin bots whay i add 74 games in process? 2010-12-02T00:29:58 it strange and stuped 2010-12-02T00:30:11 wah wah wah 2010-12-02T00:30:23 Accoun: the contest is over, the results are final. this is getting tedious 2010-12-02T00:30:33 anywey TCP result on 499 maps and 55+ games betwin bots more acurate 2010-12-02T00:30:42 sure, and impractical 2010-12-02T00:30:45 sorry 2010-12-02T00:30:47 get over it 2010-12-02T00:30:54 candead mail again 2010-12-02T00:30:55 this was accurate enough i'm sure 2010-12-02T00:31:02 noone trast to him 2010-12-02T00:31:08 Accoun: post your source on forums -- I'm sure there will be volunteers who will run your code against bocsi's and iouri's... 2010-12-02T00:31:09 i'm like to see the final ratings.txt though, out of curiosity :) 2010-12-02T00:31:22 *i'd 2010-12-02T00:31:30 u can make 99 bots cutof and more games byt u nan tot do it 2010-12-02T00:31:40 u can make 99 bots cutof and more games byt u can not do it 2010-12-02T00:31:42 just stop. last warning 2010-12-02T00:31:49 *** ChanServ sets mode: +o jmcarthur 2010-12-02T00:31:50 he alrady got that 2010-12-02T00:31:56 come on ban me like lase yaer 2010-12-02T00:32:01 come on ban me like last yaer 2010-12-02T00:32:11 ctuped canadian chiters 2010-12-02T00:32:15 Accoun: send me link to your source code -- I'll run you your very own mini contest against all open sourced bots! 2010-12-02T00:32:16 stuped canadian chiters 2010-12-02T00:32:18 i'm not even in canada 2010-12-02T00:32:27 *** jmcarthur sets mode: +q *!*@static-213-88-2-78.chebnet.ru 2010-12-02T00:32:32 thanks 2010-12-02T00:32:33 *** jmcarthur sets mode: -o jmcarthur 2010-12-02T00:32:34 *** Epona has joined #aichallenge 2010-12-02T00:32:36 neither is janzert, but that has been pointed out many times 2010-12-02T00:32:37 oh.. I think I know what chiters means now.. 2010-12-02T00:32:43 any word on bocsimackos code? 2010-12-02T00:32:48 duno why people couldn't have said so when I asked earlier 2010-12-02T00:32:51 Epona: it's out 2010-12-02T00:32:57 link? 2010-12-02T00:33:03 bah, it's in the forums 2010-12-02T00:33:06 or search forums 2010-12-02T00:33:08 sigh 2010-12-02T00:33:08 i'd have to do the same work as you :P 2010-12-02T00:33:19 Epona: his website is here: http://quotenil.com/ 2010-12-02T00:33:45 http://quotenil.com/git/?p=planet-wars.git 2010-12-02T00:35:02 @learn bocsimacko as http://quotenil.com/ 2010-12-02T00:35:02 sigh: Ready to serve, my lord. 2010-12-02T00:35:04 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-12-02T00:35:06 :P 2010-12-02T00:37:11 sigh: http://steve-yegge.blogspot.com/2010/12/haskell-researchers-announce-discovery.html :P 2010-12-02T00:37:20 * jmcarthur sighs 2010-12-02T00:37:23 i can't tell if it's all a joke or not aha 2010-12-02T00:37:28 yah, someone posted that earlier :P 2010-12-02T00:37:43 i think it's a joke. yegge has said before that haskell is in his top ten favorite languages or some crap like that 2010-12-02T00:37:52 according to a comment i read earlier today at least 2010-12-02T00:39:00 a multiplayer cross between dodgeball and tag could be cool 2010-12-02T00:39:19 i was kinda thinking ultimate frisbee could be neat if done right 2010-12-02T00:39:42 jmcarthur: are we still talking about AI? o_O 2010-12-02T00:39:47 which is semi similar to soccer? 2010-12-02T00:39:57 well, you can't move while you're holding the frisbee 2010-12-02T00:40:27 and you can't steal it 2010-12-02T00:40:34 *** MikeTV has quit IRC (Quit: Page closed) 2010-12-02T00:40:49 with team based, i think it might almost be better to do 2 players per match, but each player has multiple communicating bots 2010-12-02T00:40:51 the problem with AI and that.. 2010-12-02T00:41:02 otherwise ranking gets insane 2010-12-02T00:41:08 is the best strategy, is for everyone on the team to stand still, so they can aim he throw well 2010-12-02T00:41:31 i disagree 2010-12-02T00:41:33 I think mixing the bots together is way more interesting 2010-12-02T00:41:39 they'd just go in and block all your paths 2010-12-02T00:41:56 oh, if you can't throw in the air, ya.. 2010-12-02T00:42:09 then it's liek soccer, if it can't go in the air 2010-12-02T00:42:15 hmmm yeah 2010-12-02T00:42:27 wait not really 2010-12-02T00:42:31 soccer can go in the air too 2010-12-02T00:42:35 and frisbee can go low 2010-12-02T00:42:38 I think a simple RTS could be a neat idea though 2010-12-02T00:42:40 it's the goal state 2010-12-02T00:42:52 anyway, i'm out for a while 2010-12-02T00:42:58 like, each player starts with a base.. and each abse makes some unit, after X time 2010-12-02T00:43:09 and there's 3 or so different units 2010-12-02T00:43:39 *** Utkarsh has joined #aichallenge 2010-12-02T00:44:54 but like planetwars, there should be some sort of neutral parties, that give units less often, or some such 2010-12-02T00:45:21 I think it'd be too much processing though, to make the game be interesting 2010-12-02T00:46:07 the whole thing that makes rts's interesting generally(for simialrly skileld players), is the paper/rock/scissors combination with reaction time 2010-12-02T00:47:39 why not just do the starcraft tournament for rta? 2010-12-02T00:47:42 rts* 2010-12-02T00:47:55 hm, what about car racing? have some randomness in it, so you can't mathematically predict the 'perfect' speed/turn rate, etc. 2010-12-02T00:48:25 burny: antimatroid wanted mario kart 2010-12-02T00:48:50 Top 10 players: bocsimacko(3765), _iouri_(3565), Slin-.-(3524), _Astek_(3501), jimrogerz(3500), Accoun(3498), george(3494), GreenTea(3489), asavis(3480), bix0r4ever(3476) 2010-12-02T00:49:27 I think car racing would be too simple(or hard) likely though 2010-12-02T00:51:02 mario cart is cool because of the weapons 2010-12-02T00:51:08 i wouldn't be that interested in bare racing 2010-12-02T00:51:11 and symmetric maps are hard 2010-12-02T00:51:36 *** Utkarsh has quit IRC (Ping timeout: 250 seconds) 2010-12-02T00:52:14 http://torcs.sourceforge.net/index.php 2010-12-02T00:52:17 there's that for bare racing too 2010-12-02T00:55:54 *** Utkarsh has joined #aichallenge 2010-12-02T00:57:43 *** Appleman1234_ has joined #aichallenge 2010-12-02T00:59:06 *** mega1 has joined #aichallenge 2010-12-02T00:59:34 *** Appleman1234 has quit IRC (Disconnected by services) 2010-12-02T00:59:56 *** Appleman1234_ is now known as Appleman1234 2010-12-02T01:05:39 *** amstan has joined #aichallenge 2010-12-02T01:05:39 *** ChanServ sets mode: +o amstan 2010-12-02T01:08:11 mega1: congrats 2010-12-02T01:08:29 thank you 2010-12-02T01:09:03 how does it feel? 2010-12-02T01:09:39 *** McLeopold has joined #aichallenge 2010-12-02T01:10:23 has anyone tried running mega1's bot? 2010-12-02T01:11:37 *** wh1teside_ has joined #aichallenge 2010-12-02T01:12:57 amstan: feels great ... and it's a huge relief that it's over. 2010-12-02T01:13:50 mega1: do you mind if I run your bot on my tcp server? 2010-12-02T01:15:24 *** wh1teside has quit IRC (Ping timeout: 276 seconds) 2010-12-02T01:16:30 ? 2010-12-02T01:17:44 McLeopold: sure, go ahead. Just give it enough cpu. 2010-12-02T01:18:01 cool, so, do you have the patience to help me? 2010-12-02T01:18:09 I can seem to compile it 2010-12-02T01:18:22 sbcl version 1.0.19 2010-12-02T01:18:55 bad runtime option "--lose-on-corruption" 2010-12-02T01:19:03 oh. I have to go now for a while. 2010-12-02T01:19:06 just remove that 2010-12-02T01:19:37 ok 2010-12-02T01:19:43 just have to find where it is... 2010-12-02T01:24:37 *** pccc has joined #aichallenge 2010-12-02T01:25:06 bin/run-sbcl.sh 2010-12-02T01:25:14 k, got it 2010-12-02T01:25:46 so anyone take a look at gabor's solution? 2010-12-02T01:26:24 mega1: it's using 20% cpu 2010-12-02T01:26:34 and Accoun is probably the only one playing it... 2010-12-02T01:27:10 *** nullkuhl has quit IRC (Read error: No route to host) 2010-12-02T01:28:48 Okay, it's running. 2010-12-02T01:29:21 if anyone still wants to beat mega1, practice at tcp.zeroviz.us :) 2010-12-02T01:31:51 *** ChanServ sets mode: +o jmcarthur 2010-12-02T01:32:01 *** jmcarthur sets mode: -q *!*@static-213-88-2-78.chebnet.ru 2010-12-02T01:32:04 *** jmcarthur sets mode: -o jmcarthur 2010-12-02T01:32:21 (shh) 2010-12-02T01:33:11 *** sigh has quit IRC (Ping timeout: 272 seconds) 2010-12-02T01:33:16 Does anyone have a summary of the bocsimacko algorithm? 2010-12-02T01:33:39 pccc: mega1 probably does ;) 2010-12-02T01:33:54 http://ai-contest.com/forum/viewtopic.php?f=18&p=8057#p8057 2010-12-02T01:34:11 *** ChanServ sets mode: +o jmcarthur 2010-12-02T01:34:16 *** jmcarthur sets mode: +q *!*@static-213-88-2-78.chebnet.ru 2010-12-02T01:34:18 *** jmcarthur sets mode: -o jmcarthur 2010-12-02T01:34:19 enough 2010-12-02T01:34:40 jmarthur: is mega1 gabor? 2010-12-02T01:34:45 holy crap, I beat mega1, first time ever 2010-12-02T01:34:52 yes 2010-12-02T01:37:18 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-12-02T01:38:04 mega1: I'm looking at your code and it looks like you are using a minimax algorithm with alpha-beta pruning, correct? 2010-12-02T01:38:11 McLeopold: it's using more cpu in the opening phase 2010-12-02T01:38:25 pccc: yes, for the opening 2010-12-02T01:38:31 yea, i noticed that, it drops to 3-6% 2010-12-02T01:38:48 on tcp there are big delays between moves 2010-12-02T01:39:24 do you think I'm not running your bot to full capacity? 2010-12-02T01:41:01 mega1: Do you consider all possible moves or is that search space too big? 2010-12-02T01:41:26 pccc: I'll publish the description on my blog soon. 2010-12-02T01:41:29 @repo 2010-12-02T01:41:31 amstan: repo = http://code.google.com/p/ai-contest/ 2010-12-02T01:41:32 (it's way too big) 2010-12-02T01:43:06 mega1: okay cool, looking forward to your post 2010-12-02T01:48:29 *** pccc has quit IRC (Quit: Page closed) 2010-12-02T01:48:51 Top 10 players: bocsimacko(3765), _iouri_(3565), Slin-.-(3524), _Astek_(3501), jimrogerz(3500), Accoun(3498), george(3494), GreenTea(3489), asavis(3480), bix0r4ever(3476) 2010-12-02T01:49:01 pushed the post-mortem 2010-12-02T01:49:14 http://quotenil.com/ 2010-12-02T01:49:59 *** pccc has joined #aichallenge 2010-12-02T01:50:25 *** virdo has quit IRC (Ping timeout: 264 seconds) 2010-12-02T01:52:18 pccc: post-mortem is up 2010-12-02T01:52:46 mega1: that was quick! thanks reading it now. 2010-12-02T02:01:16 *** Frontier has quit IRC (Read error: Connection reset by peer) 2010-12-02T02:02:18 *** Frontier has joined #aichallenge 2010-12-02T02:05:13 *** Rubicon-|-Cross has joined #aichallenge 2010-12-02T02:11:01 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-12-02T02:11:52 *** Appleman1234 has joined #aichallenge 2010-12-02T02:13:25 *** McLeopold has quit IRC (Quit: Page closed) 2010-12-02T02:14:05 mega1: so when you calculated your step target, it only required the minimum number of ships to capture/keep the planet? 2010-12-02T02:16:37 hi 2010-12-02T02:17:09 *** nullkuhl has joined #aichallenge 2010-12-02T02:17:34 *** Azrathud has joined #aichallenge 2010-12-02T02:19:12 how do I read bocsimackos code? 2010-12-02T02:19:46 aside from "learn lisp"? 2010-12-02T02:20:03 nevermind I think i see it now 2010-12-02T02:20:19 I was just confused by the massive amount of planet files at first 2010-12-02T02:20:29 Epona: he just posted a retrospective on his blog: http://quotenil.com/Planet-Wars-Post-Mortem.html 2010-12-02T02:20:38 Epona: code is published ?? 2010-12-02T02:20:42 I was looking at his algorithm just now as well 2010-12-02T02:21:24 which one? 2010-12-02T02:21:32 nullkuhl: yes, here's the download link: http://quotenil.com/git/?p=planet-wars.git;a=snapshot;h=9aad7d1d4554824511420e5114acbde48312cda3 2010-12-02T02:21:45 link to repository: http://quotenil.com/git/?p=planet-wars.git 2010-12-02T02:21:54 Epona: his final submission 2010-12-02T02:23:41 got it 2010-12-02T02:24:02 looking at "control flow" but I guess I don't know shit about lisp 2010-12-02T02:24:27 Epona: if your goal is to understand his algorithm I would suggest reading his blog post 2010-12-02T02:25:02 if you want to look at his code, i think most of the interesting stuff is in player.lisp 2010-12-02T02:27:45 when is the next AI Challenge :P ? 2010-12-02T02:28:27 pccc, thankyou 2010-12-02T02:29:45 nullkuhl: i don't know, but I can't wait. I found out about this contest too late 2010-12-02T02:32:35 mega1: you're not very good at keeping promises :D 2010-12-02T02:42:53 *** sigh has joined #aichallenge 2010-12-02T02:43:57 *** Mistmanov has joined #aichallenge 2010-12-02T02:44:15 Error323: I don't know what to say.. 2010-12-02T02:48:51 Top 10 players: bocsimacko(3765), _iouri_(3565), Slin-.-(3524), _Astek_(3501), jimrogerz(3500), Accoun(3498), george(3494), GreenTea(3489), asavis(3480), bix0r4ever(3476) 2010-12-02T02:49:05 @scheduler remove rankings 2010-12-02T02:49:05 amstan: OK 2010-12-02T02:49:54 @topic remove 3 2010-12-02T02:49:54 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Contest is now closed, Results Wednesday || Go play Minecraft!" 2010-12-02T02:49:56 @topic remove 3 2010-12-02T02:49:56 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Go play Minecraft!" 2010-12-02T02:49:57 @topic remove 3 2010-12-02T02:49:57 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/" 2010-12-02T02:50:25 @topic add Congratulations to bocsimacko for first place! 2010-12-02T02:50:25 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Congratulations to bocsimacko for first place!" 2010-12-02T02:54:27 *** bss03-AFK is now known as bss03 2010-12-02T02:55:18 *** wvdschel has joined #aichallenge 2010-12-02T02:55:18 *** wvdschel has joined #aichallenge 2010-12-02T02:55:35 *** wvdschel has left #aichallenge 2010-12-02T02:58:06 *** amstan has quit IRC (Ping timeout: 245 seconds) 2010-12-02T03:03:24 *** narnach has quit IRC (Ping timeout: 240 seconds) 2010-12-02T03:06:41 *** bss03 has left #aichallenge ("Konversation terminated!") 2010-12-02T03:07:32 *** amstan has joined #aichallenge 2010-12-02T03:07:32 *** ChanServ sets mode: +o amstan 2010-12-02T03:08:51 *** narnach has joined #aichallenge 2010-12-02T03:11:29 *** Utkarsh has quit IRC (Ping timeout: 245 seconds) 2010-12-02T03:11:50 *** Utkarsh has joined #aichallenge 2010-12-02T03:15:33 pccc: yes, the bare minimum usually. Except when :SUREP T is passed to maybe-take-over-and-defend 2010-12-02T03:16:04 antimatroid: good at keeping promises? thank you. That's a first. 2010-12-02T03:23:20 mega1: i said you aren't good* and it was a poke at your write up :) 2010-12-02T03:23:42 oh, missed that negation 2010-12-02T03:24:23 negation is a fundamental problem automated sentiment analysis 2010-12-02T03:26:24 you mean text recognition? 2010-12-02T03:26:25 :P 2010-12-02T03:26:44 well, a subset of 2010-12-02T03:27:38 "traditionally been dominated by entries written in compiled languages like C++" 2010-12-02T03:28:16 *** iamluck|2 has joined #aichallenge 2010-12-02T03:28:16 *** iamluck has quit IRC (Read error: Connection reset by peer) 2010-12-02T03:28:16 #lisp will kill me for making the "lisp is not compiled" myth spreading :-) 2010-12-02T03:28:26 *spread 2010-12-02T03:28:32 *** amstan has quit IRC (Remote host closed the connection) 2010-12-02T03:28:41 hehe 2010-12-02T03:28:56 i wouldn't worry too much, there was a comment in there about people who "speak" lisp 2010-12-02T03:29:02 i mean c'mon 2010-12-02T03:29:57 I guess I'm ignorant, I thought lisp was an interpreted language 2010-12-02T03:30:51 Depends on the implementation. Most lisps are compiled but also have an interpreter. 2010-12-02T03:33:05 *** amstan has joined #aichallenge 2010-12-02T03:33:05 *** ChanServ sets mode: +o amstan 2010-12-02T03:34:35 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-12-02T03:39:13 *** Appleman1234 has quit IRC (Ping timeout: 272 seconds) 2010-12-02T03:39:59 *** ewreqqrew has joined #aichallenge 2010-12-02T03:40:24 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&p=8057#p8057 2010-12-02T03:40:36 *** ewreqqrew has quit IRC (Client Quit) 2010-12-02T03:40:45 *** Appleman1234 has joined #aichallenge 2010-12-02T03:41:16 *** Utkarsh has joined #aichallenge 2010-12-02T03:41:34 As with tron, UCT was on my mind. 2010-12-02T03:41:38 what is UCT 2010-12-02T03:42:23 it's a guidd mont-carlo tree search 2010-12-02T03:42:53 stands for "Upper Confidence bounds applied to Trees" 2010-12-02T03:42:57 *guided 2010-12-02T03:43:27 trees are the entire game combinations mapped out, correct? 2010-12-02T03:43:33 basically you evaluate a position by running a large numbr of random simulations, then expand along the paths which gave the best results 2010-12-02T03:43:43 Epona: yes 2010-12-02T03:43:55 ah 2010-12-02T03:44:06 its a pretty brutal way to win 2010-12-02T03:44:19 it's been *very* successful for the game of Go 2010-12-02T03:44:35 and I understand that it is also the basis for mega1's hex bot 2010-12-02T03:44:42 im trying to see if boc ever consi9dered using neural nets 2010-12-02T03:45:24 sigh: it's not. It's the basis of the hex bot that beat mine. 2010-12-02T03:45:35 ah ok, sorry :) 2010-12-02T03:45:54 boc said it was the basis of his one in the tron contest 2010-12-02T03:45:59 although it failed 2010-12-02T03:46:13 he got 11th in tron which is hardly failing, I think :P 2010-12-02T03:47:01 its not 1st 2010-12-02T03:47:07 :) 2010-12-02T03:47:13 mega1: speaking of which, have you ever written a Go bot? 2010-12-02T03:47:50 *** nullkuhl has quit IRC (Remote host closed the connection) 2010-12-02T03:48:29 he didn't submit his uct bot for tron 2010-12-02T03:48:45 sigh: no, I shopped for something with simpler rules and ended up with hex. 2010-12-02T03:48:58 hex does seem fun 2010-12-02T03:49:10 it is a good game for humans 2010-12-02T03:49:14 mega1: fair enough :) 2010-12-02T03:49:43 I wouldn't know where to start writing a go bot even starting with UCT :P 2010-12-02T03:50:37 is UCT superior? 2010-12-02T03:51:38 superior to what? 2010-12-02T03:52:01 in something like Go, it is superior to minimax, because the branching factor is simply too large for minimax 2010-12-02T03:52:14 im just trying to find the best algorithms to learn 2010-12-02T03:52:35 what is your current background? 2010-12-02T03:52:38 what do you know already> 2010-12-02T03:52:45 start with minimax with alpha beta search if you haven't implemented that before 2010-12-02T03:52:54 yes, this ^^ 2010-12-02T03:53:32 alright 2010-12-02T03:53:38 ive been working with python' 2010-12-02T03:53:48 that's fine 2010-12-02T03:54:01 have you studied algorithms at all? 2010-12-02T03:54:21 i don't think you really even need algorithms for minimax, it's that easy 2010-12-02T03:54:51 djkstras algorithm, the napsack problem, neural nets, algebra, trig 2010-12-02T03:55:00 then you should be fine with minimax 2010-12-02T03:55:27 Epona: it would be easy to make one for tron 2010-12-02T03:55:42 and all the starter packs/engines etc. are available online 2010-12-02T03:55:50 don't try to implement minimax for planetwars ;) 2010-12-02T03:56:03 at least, not as your first project, that will scar you :P 2010-12-02T03:56:08 it's possibly, but you need to do a hell of a lot of pruning if you want to make many moves 2010-12-02T03:56:14 possible* 2010-12-02T03:56:33 ive been building my own algorithms 2010-12-02T03:56:40 its pretty fun :) 2010-12-02T03:56:45 :) 2010-12-02T03:56:54 *** p4p4 has joined #aichallenge 2010-12-02T03:57:03 I should probably keep exploring that for now 2010-12-02T03:57:13 after i finish this blog of course 2010-12-02T03:57:14 nah, projects are great for learning 2010-12-02T03:57:30 once you hit a brick wall you're forced to blast you're way through 2010-12-02T03:57:41 your* 2010-12-02T03:57:43 bah 2010-12-02T03:58:01 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-12-02T03:58:13 If I ever run out of ideas 2010-12-02T03:58:15 I will 2010-12-02T03:58:43 you wont :P 2010-12-02T03:58:51 if you run out of ideas, you're doing it wrong 2010-12-02T03:59:24 lol 2010-12-02T03:59:53 I wish it was a little more math literal 2010-12-02T04:00:07 but maybe I could build that 2010-12-02T04:00:15 antimatroid: that's a nice pessimistic feedback loop you are setting up for yourself there, if you ever run out of ideas 2010-12-02T04:00:44 :P 2010-12-02T04:02:06 *** Utkarsh has joined #aichallenge 2010-12-02T04:02:22 *** aerique has joined #aichallenge 2010-12-02T04:02:24 *** eytreytr has joined #aichallenge 2010-12-02T04:02:41 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&p=8057#p8057 2010-12-02T04:02:49 *** eytreytr has quit IRC (Client Quit) 2010-12-02T04:03:06 sigh: i don't plan to :P 2010-12-02T04:03:26 there's more to learn than anyone could every hope to even touch the surface of 2010-12-02T04:03:33 :) 2010-12-02T04:03:52 Epona: what do you mean by math literal? 2010-12-02T04:04:02 literate 2010-12-02T04:04:10 like how its taught in school 2010-12-02T04:04:29 what is taught? sorry i'm lost now 2010-12-02T04:04:54 i wish relations etc. were taught in middle school 2010-12-02T04:05:02 and well, just proper maths in general 2010-12-02T04:05:15 http://www.wolframalpha.com/ is the kind of math organization im talking about 2010-12-02T04:05:58 i'm just going off what i've read on the net, but apparently wolfram is a dick 2010-12-02T04:06:31 that's the general consensus :P 2010-12-02T04:06:46 then again, i'm an arrogant little prick to people sometimes :P 2010-12-02T04:06:51 Im not sure how you came to that 2010-12-02T04:06:58 possibly remove that sometimes part too 2010-12-02T04:09:40 FAIL 2010-12-02T04:09:42 Dissapointingly, there was no learning, adapting to opponent behaviour, etc. All that made it to the todo list, but had to take second seat to more pressing concerns. 2010-12-02T04:10:31 *** medrimonia has quit IRC (Quit: Leaving.) 2010-12-02T04:12:09 *** medrimonia has joined #aichallenge 2010-12-02T04:15:52 Epona: yeah, i started out with imaginative ideas of updating my beliefs about them as a game progressed 2010-12-02T04:15:57 was never going to happen properly 2010-12-02T04:20:57 well 2010-12-02T04:21:13 I found out some proper syntax for a python neural net 2010-12-02T04:21:19 if you wanna see it 2010-12-02T04:24:07 people seem to think neural nets suck 2010-12-02T04:24:18 i've only ever had a brief introduction to them in a class, and a few worked examples by hand updating input thingys 2010-12-02T04:27:08 as is obvious from my blog I'm a big fan of deep learning and boltzmann machines 2010-12-02T04:27:28 that's a trendy kind of neural network with probabilistic roots 2010-12-02T04:27:44 trendy? so whats the true way 2010-12-02T04:29:39 mega1: i assume your work is related to ai as well? 2010-12-02T04:30:22 *** pccc has quit IRC (Ping timeout: 265 seconds) 2010-12-02T04:31:15 aerique: Currently, no. It used to be AI related as the eburnette wrote. But not for the last year or so. 2010-12-02T04:31:32 although I'm also doing a recommender system on the side. 2010-12-02T04:31:43 mega1: what are you doing these days? 2010-12-02T04:32:22 I'm a lisp consultant. 2010-12-02T04:32:42 oh man.. ebrunette totally sneaked in and spied on us to make an article 2010-12-02T04:32:44 kinda cool 2010-12-02T04:33:16 amstan: yah, he was scraping the channel logs for quotes :P 2010-12-02T04:33:39 woot, my logs have proven to be useful! 2010-12-02T04:33:40 hehe 2010-12-02T04:33:43 he also competed 2010-12-02T04:33:57 amstan: Can the "traditionally been dominated by entries written in compiled languages" part be reworded? 2010-12-02T04:34:18 mega1: yeah, not compiled, eh? 2010-12-02T04:34:22 lisp is compiled too 2010-12-02T04:34:24 compiled* 2010-12-02T04:34:26 and also, "this years winner", wasn't a1k0n such a winner :P 2010-12-02T04:34:33 no disrespect mega1 2010-12-02T04:34:40 this term's 2010-12-02T04:34:52 i could do a dirty change.. 2010-12-02T04:35:00 it's been compiled for 20 years and people keeping saying it's interpreted for just as long 2010-12-02T04:35:15 *keep 2010-12-02T04:35:41 just s/compiled//, I guess 2010-12-02T04:36:03 *** Naktibalda has joined #aichallenge 2010-12-02T04:36:40 I also have thoughts on "innovative algorithm", but hey, it's in the eye of the beholder :-) 2010-12-02T04:38:00 yeah, I was surprised to learn your bot was mostly a bunch of heuristics like everyone else :P 2010-12-02T04:38:11 just... more awesome 2010-12-02T04:39:24 k 2010-12-02T04:39:26 changes done 2010-12-02T04:39:57 thank you 2010-12-02T04:41:06 amstan, changes doen to what? 2010-12-02T04:41:08 any time 2010-12-02T04:41:24 Epona: few correction on that text in index.php and rankings.php 2010-12-02T04:43:02 ah ok 2010-12-02T04:57:32 mega1: I stumbled across "innovative algorithm" as well. After all with ~200 ELO points in the lead, you come to believe there MUST be an innovation under the hood ;). 2010-12-02T04:58:49 I even thought your bot is deliberately sending fleets forth and back randomly to confuse the enemy ai. 2010-12-02T05:01:20 *** GreenTea has quit IRC (Quit: Page closed) 2010-12-02T05:01:25 hehe 2010-12-02T05:02:08 I believe that the more advanced stuff would only pay back if the general population was sufficiently good at tactics. 2010-12-02T05:02:33 Mistmanov: yeahh... I don't wanna talka bout it :P 2010-12-02T05:02:34 I found that safety/conservatism was key. 2010-12-02T05:03:26 mega1: where is your post? 2010-12-02T05:03:41 @bocsimacko 2010-12-02T05:03:42 sigh: bocsimacko = http://quotenil.com/ 2010-12-02T05:03:54 ain't there 2010-12-02T05:03:56 youuu, who asked for it every hour, haven't read it yet? 2010-12-02T05:04:08 just woke up :$ 2010-12-02T05:04:09 it's there 2010-12-02T05:04:18 top post on the page 2010-12-02T05:04:31 oh cached here or sth 2010-12-02T05:04:32 hmz 2010-12-02T05:19:57 *** yasith has quit IRC (Ping timeout: 265 seconds) 2010-12-02T05:21:40 *** Migi32 has joined #aichallenge 2010-12-02T05:26:08 *** Ari2 has joined #aichallenge 2010-12-02T05:31:48 *** yasith has joined #aichallenge 2010-12-02T05:32:33 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-12-02T05:34:56 *** phreeza has joined #aichallenge 2010-12-02T05:39:40 *** antimatroid1 has joined #aichallenge 2010-12-02T05:41:17 *** wh1teside has joined #aichallenge 2010-12-02T05:41:26 *** antimatroid has quit IRC (Ping timeout: 245 seconds) 2010-12-02T05:42:08 *** Rubicon-|-Cross has joined #aichallenge 2010-12-02T05:45:09 *** wh1teside_ has quit IRC (Ping timeout: 276 seconds) 2010-12-02T06:57:05 *** perror has joined #aichallenge 2010-12-02T07:10:59 btw, no mention to the winning organization in the home page :P 2010-12-02T07:12:07 "Other" 2010-12-02T07:13:32 are the websites for the contests before tron online? 2010-12-02T07:14:17 * Naktibalda didn't hear anything about contests before tron :) 2010-12-02T07:15:02 *** ffsgsdgsfdg has joined #aichallenge 2010-12-02T07:15:05 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&p=8057#p8057 2010-12-02T07:15:13 *** ffsgsdgsfdg has quit IRC (Client Quit) 2010-12-02T07:18:50 "The first contest was called Minesweeper Flags, followed by Rock Paper Scissors,..." 2010-12-02T07:20:55 I am never playing minecraft again 2010-12-02T07:21:07 OH LOOK SOME REDSTONE 2010-12-02T07:21:18 TK TK TK TK, ahhhhhh (falls in lava) 2010-12-02T07:25:00 *** amstan sets mode: +b *!*@213.88.2.78 2010-12-02T07:25:12 wow that guy sounds bitter... 2010-12-02T07:25:27 he would sound like that anymore... get it? lol 2010-12-02T07:25:48 s/sound/"sound" 2010-12-02T07:26:06 huh? 2010-12-02T07:26:06 :D 2010-12-02T07:30:15 omg 2010-12-02T07:30:33 btw did you guys check the video at his about me? 2010-12-02T07:30:35 its... 2010-12-02T07:30:36 special 2010-12-02T07:30:44 Error323: yes, do you know russian? 2010-12-02T07:30:52 video about you? o_O 2010-12-02T07:30:53 link? 2010-12-02T07:30:54 no 2010-12-02T07:31:02 a friend of mine does though 2010-12-02T07:31:04 i'll ask him =) 2010-12-02T07:31:17 why do you get a video? 2010-12-02T07:31:22 Error323: we had russians in here claiming it was something about wanting 16-18 year old girls to marry him 2010-12-02T07:31:33 amstan: O_O 2010-12-02T07:31:38 ... 2010-12-02T07:31:56 damn, he's seriuosly twisted 2010-12-02T07:32:02 yeah, lol 2010-12-02T07:32:16 what video? 2010-12-02T07:32:27 on his about me 2010-12-02T07:32:49 are you a 16-18 yo girl? :P 2010-12-02T07:32:52 Mistmanov: I hate you 2010-12-02T07:32:55 can we see this anywhere? 2010-12-02T07:33:00 don't we all 2010-12-02T07:33:01 http://www.youtube.com/watch?v=p-5nXqUF29Q 2010-12-02T07:33:03 =) 2010-12-02T07:33:16 nah good job hehe, gratz 2010-12-02T07:33:31 i like his other videos too 2010-12-02T07:33:33 lol 2010-12-02T07:33:44 I wouldn't mind 18 year old girls marrying me.. 2010-12-02T07:33:56 title is "I want to marry" :) 2010-12-02T07:34:06 time for lunch now 2010-12-02T07:34:06 do you think I can get one when I impress them with my awesome FIRST PLACE FINISH FOR THE NETHERLANDS? 2010-12-02T07:34:28 second place would've been a total turn-off ofcourse.. 2010-12-02T07:34:37 I hope she claws your eyes out 2010-12-02T07:34:40 xD 2010-12-02T07:34:41 that's like 2010-12-02T07:34:47 "you tried really hard, but totally failed" 2010-12-02T07:34:52 ;) 2010-12-02T07:35:11 grmbl grmbl 2010-12-02T07:35:20 I seriously had a nightmare that I would wake up to check the rankings, and both of us would've dropped to like 150 :p 2010-12-02T07:35:22 :] 2010-12-02T07:35:29 hahahhahah 2010-12-02T07:35:53 I just needed one more game 2010-12-02T07:35:56 one more win 2010-12-02T07:36:05 it'll haunt me forever 2010-12-02T07:36:17 or well.. till next contest 2010-12-02T07:36:23 k.. so what about this contest.. 2010-12-02T07:36:24 hmhm =) 2010-12-02T07:36:26 what shall it be? 2010-12-02T07:36:40 who's up for 2d minecraft on infinite sized maps 2010-12-02T07:36:45 ahahahah wtf i'm looking at the other videos 2010-12-02T07:36:48 mmorts 2010-12-02T07:37:01 8bit stick figure porn 2010-12-02T07:37:05 well he had some brainstorming on soccer, towerdefense, rts and some kind of liquid game. 2010-12-02T07:37:23 * Error323 doesn't know minecraft 2010-12-02T07:37:37 I think an mmo kind of thing would be awesome, but hard to score, and maybe hard to actually make it an interesting problem for bots 2010-12-02T07:37:38 *** antimatroid has joined #aichallenge 2010-12-02T07:38:02 who had videos asking for 18 year old girls to marry them? 2010-12-02T07:38:07 lol 2010-12-02T07:38:12 hahahha 2010-12-02T07:38:24 were you tailing logs or sth? :P 2010-12-02T07:38:42 it was just before 2010-12-02T07:38:50 antimatroid: accoun 2010-12-02T07:38:59 *** antimatroid1 has quit IRC (Ping timeout: 245 seconds) 2010-12-02T07:39:02 so confused? 2010-12-02T07:39:09 has anyone actually translated that movie? 2010-12-02T07:39:23 it's the same video he had on his tron profile 2010-12-02T07:39:59 title tranlate: http://translate.google.com/#auto|en|%D0%9E%D1%87%D05%D0D%D1%8C%20%D0%A5%D0E%D1%87%D1%83%20%D0%96%D05%D0D%D08%D1%82%D1%81%D1%8F 2010-12-02T07:40:03 darn.. 2010-12-02T07:40:28 http://tinyurl.com/23ptgtb 2010-12-02T07:40:41 lol 2010-12-02T07:40:45 hahah 2010-12-02T07:41:03 "I very much want Marries" awww it even sounds like him 2010-12-02T07:41:09 lol 2010-12-02T07:41:15 except it's not repeated 5 times 2010-12-02T07:41:19 lol 2010-12-02T07:41:31 don't forget the 4999 maps 2010-12-02T07:41:43 :P 2010-12-02T07:42:04 "don't smoke let'd fuck" ahahah 2010-12-02T07:42:06 @chiting 2010-12-02T07:42:07 Error323: chiting = (#1) rank(x) = country(x) == canadian ? +9999 : -9999, or (#2) rank(x) = country(x) == canada ? +9999 : -9999 2010-12-02T07:42:26 Error323: lol.. who added that? 2010-12-02T07:42:36 Error323: oh, it was you 2010-12-02T07:42:41 hahaha 2010-12-02T07:42:47 i think i'd have brought australia up to second place 2010-12-02T07:42:50 @winner 2010-12-02T07:42:51 amstan: winner = bhasker 2010-12-02T07:42:54 xD 2010-12-02T07:42:57 ... 2010-12-02T07:42:59 haha 2010-12-02T07:43:03 @forget winner 2010-12-02T07:43:03 amstan: Yes master! 2010-12-02T07:43:07 I wonder who did that 2010-12-02T07:43:25 @forget chiting 1 2010-12-02T07:43:25 Error323: Ready to serve. 2010-12-02T07:43:32 @chiting 2010-12-02T07:43:33 Error323: chiting = rank(x) = country(x) == canada ? +9999 : -9999 2010-12-02T07:43:33 @forget predict 2010-12-02T07:43:34 amstan: Error: 2 factoids have that key. Please specify which one to remove, or use * to designate all of them. 2010-12-02T07:43:44 @forget predict * 2010-12-02T07:43:44 amstan: Ready to serve, my lord. 2010-12-02T07:43:56 @predict 2010-12-02T07:43:56 I predict that Ari2 will win! 2010-12-02T07:44:00 :p 2010-12-02T07:44:22 predict wasn't a factoid, it was an alias 2010-12-02T07:44:32 for some reason.. 2010-12-02T07:44:36 i tried making it a factoid 2010-12-02T07:44:39 but noo 2010-12-02T07:44:40 =) 2010-12-02T07:45:24 @alias 2010-12-02T07:45:24 amstan: (alias [] []) -- Adds a new key for factoid associated with . is only necessary if there's more than one factoid associated with . The same action can be accomplished by using the 'learn' function with a new key but an existing (verbatim) factoid content. 2010-12-02T07:45:29 @alias list 2010-12-02T07:45:29 amstan: (alias [] []) -- Adds a new key for factoid associated with . is only necessary if there's more than one factoid associated with . The same action can be accomplished by using the 'learn' function with a new key but an existing (verbatim) factoid content. 2010-12-02T07:45:32 .. 2010-12-02T07:45:51 @help predict 2010-12-02T07:45:51 sigh: (predict ) -- Alias for "echo I predict that $randomNick will win!". 2010-12-02T07:47:00 yay, mega1's online 2010-12-02T07:47:22 http://www.codecup.nl/intro.php 2010-12-02T07:47:26 that game looks quite cool actually 2010-12-02T07:47:52 it has hexagons and math.. 2010-12-02T07:47:57 nobody likes math.. 2010-12-02T07:48:20 "The source code must consist of one single file" deal breaker 2010-12-02T07:48:24 i'm not doing that, fuck that 2010-12-02T07:48:34 amstan: i love math :( 2010-12-02T07:48:46 you can always merge several of your source files just before submission 2010-12-02T07:48:56 so much effort 2010-12-02T07:49:05 just use #includes 2010-12-02T07:49:18 use include guards.. 2010-12-02T07:49:27 then run gpp on your source files then cat them together 2010-12-02T07:49:53 that should work, i think 2010-12-02T07:50:03 gpp? 2010-12-02T07:50:14 g++ 2010-12-02T07:50:30 preprocessor 2010-12-02T07:50:31 *** Mistmanov has quit IRC (Quit: Page closed) 2010-12-02T07:50:45 ah yeah, forgot about that 2010-12-02T07:50:57 wait.. that won't work... you have to cat them first.. then gpp it 2010-12-02T07:51:11 the include guards will make sure you don't have duplicate declarations 2010-12-02T07:51:18 :P i would just copy them from .h files i think 2010-12-02T07:51:48 amstan: why would you need to cat them together if you are running it though the preprocessor 2010-12-02T07:52:11 *** Azrathud has quit IRC (Quit: Leaving) 2010-12-02T07:52:16 sigh: because those c files will not merge by themselves 2010-12-02T07:52:34 unless you do evil stuff like #include 2010-12-02T07:52:36 they will if you are #including the .cpp files 2010-12-02T07:52:39 yes :P 2010-12-02T07:52:55 so sketchy.. 2010-12-02T07:52:58 I actually did that in this contest so that I could have src files in directories :P 2010-12-02T07:53:54 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-12-02T07:53:55 ideally for next contest we'll sandbox the compilation and allow makefiles and stuff 2010-12-02T07:54:00 yes 2010-12-02T07:54:17 *** Azrathud has joined #aichallenge 2010-12-02T07:55:16 *** ebrahim has joined #aichallenge 2010-12-02T07:56:20 i liked the idea of people submitting binaries 2010-12-02T07:56:36 that too 2010-12-02T07:56:55 one thing i don't like about it is that people will be able to submit closed source bots 2010-12-02T07:57:27 and that might be a problem if we want to analyze a bot for security reasons 2010-12-02T07:57:39 though, i don't think we had to do this yet 2010-12-02T07:58:43 hmmm, i do think it's important the code can be accessed 2010-12-02T07:58:51 and shared afterwards 2010-12-02T07:59:01 imo its the most valuable aspect of the contest 2010-12-02T07:59:06 maybe require people submit source as well, and state that if it's later tested and not the same as the binary, they get kicked 2010-12-02T07:59:40 although then how do you know if it's the same if it's not determnistic 2010-12-02T07:59:41 hmmmm 2010-12-02T07:59:59 *** fgump has joined #aichallenge 2010-12-02T08:00:03 antimatroid: that can easily be solved 2010-12-02T08:00:04 makes no sense, just upload code and let the server compile 2010-12-02T08:00:06 yeah i don't like binaries :P 2010-12-02T08:00:09 by using a preset seed 2010-12-02T08:00:20 given by the contest 2010-12-02T08:00:32 I would much more prefer some I/O space per user 2010-12-02T08:00:52 then Accoun will complain that you chose a chiting seed 2010-12-02T08:01:02 amstan: one thing that is very important next time is very clear upper bounds on parameters 2010-12-02T08:01:11 i really like the general rules, but i think they need to be a bit more clear 2010-12-02T08:01:19 antimatroid: it would be nice to have a json communications protocol 2010-12-02T08:01:35 and just include everything in there.. 2010-12-02T08:02:00 i had this awesome idea for the mmorts... let bots communicate 2010-12-02T08:02:05 what is the plan at this point with fixing things up anyway? 2010-12-02T08:02:14 antimatroid: fixing? define fixing 2010-12-02T08:02:21 i've said before, but i'm happy to help so far as i can and i don't become ineligible to compete 2010-12-02T08:02:29 improving* 2010-12-02T08:02:32 might be a better work 2010-12-02T08:02:45 antimatroid: as i said before, unless you get root, it won't matter 2010-12-02T08:03:04 we'll probably do a rewrite 2010-12-02T08:03:07 clear upper bounds and important changes should always be communicated directly via mail or sth. I stil don't understand why this didn't happen. 2010-12-02T08:03:07 yeah, so how long do you plan to break for before getting to work on the framework stuff for next contest? 2010-12-02T08:03:20 i'll probably start in jan 2010-12-02T08:03:27 since i have exams till then 2010-12-02T08:03:38 cool 2010-12-02T08:03:39 yep 2010-12-02T08:03:45 you bloody set this to end after mine! :P 2010-12-02T08:03:58 ? 2010-12-02T08:04:01 that was a very painful few weeks 2010-12-02T08:04:08 i went through exams with like 3 weeks left in the contest 2010-12-02T08:04:15 i was doing an overload too 2010-12-02T08:04:57 i ended up studying for each exam the day before, some of the marks weren't particularly crash hot :P 2010-12-02T08:05:25 *** Epona has quit IRC (Ping timeout: 264 seconds) 2010-12-02T08:05:31 antimatroid: i know the feeling, i'll probably end up doing the same, lol 2010-12-02T08:07:41 I postponed the talk with my supervisor on intermediate results for my thesis 3 times... 2010-12-02T08:08:12 xD 2010-12-02T08:08:16 Error323: hehe, bet they loved that 2010-12-02T08:08:25 http://www.facebook.com/felon <- i want to punch that guy in the face 2010-12-02T08:08:44 he's the guy that tipped the fbi off about manning, and he's just such a douche 2010-12-02T08:08:54 uhh? 2010-12-02T08:09:16 Manning is the american dude that released us government stuff to wikileaks and thown behind bars 2010-12-02T08:10:44 ahh right 2010-12-02T08:11:03 "Wrong Adrian Lamo?" Yes 2010-12-02T08:12:25 mega1:? 2010-12-02T08:12:38 that's what facebook asks 2010-12-02T08:13:38 http://www.wired.com/threatlevel/2010/06/wikileaks-chat/ 2010-12-02T08:13:46 i got through to the profile page from a picture linked on that page 2010-12-02T08:14:14 *** fdssdgfsgfd has joined #aichallenge 2010-12-02T08:14:17 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&p=8057#p8057 2010-12-02T08:14:22 *** fdssdgfsgfd has quit IRC (Client Quit) 2010-12-02T08:14:43 uhhhh 2010-12-02T08:14:49 *** nullkuhl has joined #aichallenge 2010-12-02T08:16:09 can people run his bot? 2010-12-02T08:16:27 he didn't even give the source 2010-12-02T08:16:31 what the hell is he hiding 2010-12-02T08:17:14 I think hes one of those guys thats all about conspiracies and security =) 2010-12-02T08:17:41 did he just uploaded the exe? 2010-12-02T08:18:31 hence the reason for the appearance of maps was another, Janzert lying 2010-12-02T08:18:33 maps have been added when I was in 2nd position 2010-12-02T08:18:38 XD 2010-12-02T08:18:54 yep, i was going to take a look at his source 2010-12-02T08:19:00 see what he did, but nope :P 2010-12-02T08:19:09 although, it's probably in russian so 2010-12-02T08:19:33 Either russian, or completely scrambled stuff 2010-12-02T08:19:51 I do hope it were russian, and rather pull it through a translator myself =0 2010-12-02T08:20:09 no offense 2010-12-02T08:20:21 but i just didn't understand a lot he said 2010-12-02T08:20:35 *** smellyhippy has quit IRC (Ping timeout: 260 seconds) 2010-12-02T08:21:23 *** amstan sets mode: +b *!*@*213.88.2.78 2010-12-02T08:21:48 someone ping me if he goes around that 2010-12-02T08:22:47 *** smellyhippy has joined #aichallenge 2010-12-02T08:23:14 Error323: it wasn't that hard to work out what he was saying most of the time 2010-12-02T08:23:26 and frankly, i'm willing to let people slip on poor english if it's not their native tongue 2010-12-02T08:23:37 he as a person I got sick of though 2010-12-02T08:23:39 well perhaps for native speakers it was more easy 2010-12-02T08:23:43 :P 2010-12-02T08:23:47 that probably helped 2010-12-02T08:23:50 but some stuff I just couldn't parse 2010-12-02T08:24:03 i can't talk, i know no other languages 2010-12-02T08:24:08 yeah, that was the problem, i couldn't parse his posts 2010-12-02T08:24:37 and if i do, they sometimes don't make sense 2010-12-02T08:24:45 like the sentences do, but the thoughts don't 2010-12-02T08:25:14 also, what's the obsesion with 100k-1 map count 2010-12-02T08:25:21 where k in {4,9} 2010-12-02T08:25:29 no idea 2010-12-02T08:25:55 whilest i do agree on the more maps being better, i don't know why he wanted that much 2010-12-02T08:26:28 amstan: It's because 500 maps are less accurate than 499 =P 2010-12-02T08:26:35 he also suggested that pairs should play on all maps 2010-12-02T08:27:03 he must have known that that's unfeasable 2010-12-02T08:27:04 medrimonia: oh man... us chitting organizers introducing that one map.. 2010-12-02T08:27:26 he was stupid and not stupid at the same time 2010-12-02T08:27:29 ^_^ 2010-12-02T08:28:53 to be fair, people in the channel were fanning the flames quite a bit 2010-12-02T08:29:52 * antimatroid raises his hand 2010-12-02T08:29:53 me being one of them i admit 2010-12-02T08:30:06 i can't help myself, it 2010-12-02T08:30:13 s fun getting under peoples skin 2010-12-02T08:30:17 but after trying normal conversations 2010-12-02T08:30:36 i did try to be patient with him, i even had a pm convo with him once trying to get him to chill 2010-12-02T08:30:44 some stuff was so absurd i seriously thought he was joking 2010-12-02T08:31:13 yeah, me too 2010-12-02T08:31:27 then I thought he was just trolling 2010-12-02T08:31:38 then he just kept going D: 2010-12-02T08:35:28 *** delt0r_ has quit IRC (Quit: Leaving) 2010-12-02T08:35:59 *** AlliedEnvy has quit IRC (Read error: Connection reset by peer) 2010-12-02T08:36:59 *** AlliedEnvy has joined #aichallenge 2010-12-02T08:48:36 *** sigh has quit IRC (Remote host closed the connection) 2010-12-02T08:50:12 *** Appleman1234 has quit IRC (Ping timeout: 265 seconds) 2010-12-02T08:51:15 *** Palmik has joined #aichallenge 2010-12-02T08:51:31 *** Appleman1234 has joined #aichallenge 2010-12-02T08:58:21 *** Appleman1234 has quit IRC (Excess Flood) 2010-12-02T08:58:50 *** Appleman1234 has joined #aichallenge 2010-12-02T09:03:13 *** Appleman1234 has quit IRC (Ping timeout: 255 seconds) 2010-12-02T09:29:42 *** delt0r has joined #aichallenge 2010-12-02T09:30:57 *** ebrahim has quit IRC (Remote host closed the connection) 2010-12-02T09:34:15 *** Azrathud has quit IRC (Ping timeout: 240 seconds) 2010-12-02T09:57:46 *** perror has quit IRC (Quit: Bye all !) 2010-12-02T10:03:20 *** Azrathud has joined #aichallenge 2010-12-02T10:05:27 *** bhasker has joined #aichallenge 2010-12-02T10:12:59 *** JamesMG has joined #aichallenge 2010-12-02T10:27:21 *** mega1 has quit IRC (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.) 2010-12-02T10:27:29 *** mega1 has joined #aichallenge 2010-12-02T10:32:37 *** p4p4 has quit IRC (Read error: Connection timed out) 2010-12-02T10:33:13 *** fgump has quit IRC (Ping timeout: 255 seconds) 2010-12-02T10:34:07 *** p4p4 has joined #aichallenge 2010-12-02T10:37:37 *** aerique has quit IRC (Quit: ...) 2010-12-02T10:45:36 *** Meatkat has joined #aichallenge 2010-12-02T10:45:50 Party? 2010-12-02T10:45:57 @dance 2010-12-02T10:45:58 Meatkat: You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2010-12-02T10:46:41 @rainbow 2010-12-02T10:46:41 phreeza: (rainbow ) -- Returns colorized like a rainbow. 2010-12-02T10:46:56 @rainbow disco dance! 2010-12-02T10:46:56 phreeza: disco dance! 2010-12-02T10:47:58 welp, see you later 2010-12-02T10:48:19 *** Meatkat has quit IRC (Quit: Leaving) 2010-12-02T10:48:51 *** amstan has quit IRC (Ping timeout: 265 seconds) 2010-12-02T10:52:41 *** Utkarsh has quit IRC (Ping timeout: 245 seconds) 2010-12-02T10:54:07 *** aaa_ has joined #aichallenge 2010-12-02T10:59:01 *** Utkarsh has joined #aichallenge 2010-12-02T11:02:20 *** fgump has joined #aichallenge 2010-12-02T11:03:54 *** fgump has quit IRC (Client Quit) 2010-12-02T11:15:16 *** narnach has quit IRC (Ping timeout: 240 seconds) 2010-12-02T11:18:43 *** superflit has quit IRC (Read error: Connection reset by peer) 2010-12-02T11:19:27 *** superflit has joined #aichallenge 2010-12-02T11:19:51 *** Azrathud has quit IRC (Ping timeout: 240 seconds) 2010-12-02T11:20:42 *** narnach has joined #aichallenge 2010-12-02T11:33:19 the chemists method for numerical integration.Plot curve on paperPrecisely cut out shapeWeigh paper shape with highly accurate scales 2010-12-02T11:33:22 ehehe 2010-12-02T11:35:33 *** Migi32 has quit IRC (Quit: www.ai-contest.com) 2010-12-02T11:37:12 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]) 2010-12-02T11:52:51 haha. mega1 copied me :( 2010-12-02T11:52:53 *** superflit has quit IRC (Quit: superflit) 2010-12-02T11:53:29 yeah, looks like you started a tradition? :) 2010-12-02T11:53:36 evidently 2010-12-02T11:54:00 his lead was much more decisive. 2010-12-02T11:54:19 mine was pretty last-minute 2010-12-02T11:54:44 that's why I added the third sentence to explain. 2010-12-02T11:54:49 I am pretty amazed how long he held such a strong lead 2010-12-02T11:54:52 yeah i know. i'm amused. 2010-12-02T11:55:21 Hi mega1 :) 2010-12-02T11:55:45 I found that what you wrote summed up pretty well how I feel. 2010-12-02T11:55:48 janzert: hello 2010-12-02T11:55:50 mega1: i don't understand your bit vector or ship counts thing 2010-12-02T11:55:58 i also haven't finished reading your post 2010-12-02T11:56:14 I'm happy to explain. 2010-12-02T11:57:20 *** rebelxt has joined #aichallenge 2010-12-02T12:02:45 hah, and now we know what bocsimacko is. 2010-12-02T12:03:31 it's hard to part with such grave secrets 2010-12-02T12:03:47 I used this name first in the netflix contest. 2010-12-02T12:03:58 i've seen it many times. 2010-12-02T12:04:28 mega1: thanks for the nice write up 2010-12-02T12:04:32 so, you have a bit vector of ship counts in your code, which i haven't fully understood 2010-12-02T12:05:01 janzert: It's a pleasant obligation. 2010-12-02T12:05:18 i had a lot of trouble getting future moves scheduled and not generating incompatible orders subsequently 2010-12-02T12:05:25 a1k0n_: it's not a bit vector. It's an integer vector. 2010-12-02T12:05:43 oh, right. 2010-12-02T12:05:53 (16 bit wide) 2010-12-02T12:06:03 ohhh. 2010-12-02T12:06:15 i thought your comment was about the size of the array, not the width. 2010-12-02T12:06:37 I'm confused now. Which comment? 2010-12-02T12:06:39 in that case i had the exact same thing. 2010-12-02T12:06:46 ok 2010-12-02T12:07:01 :s 2010-12-02T12:07:03 ;;; There are 23 planets, max growth is 5, 200 turns, 100 initial 2010-12-02T12:07:03 ;;; ships. (+ 100 (* 23 5 200)) is 23100 that fits into SHIP-COUNT. 2010-12-02T12:07:34 it's been a while since i'd written lisp; i thought you were implying the arrays were 23100 in size 2010-12-02T12:07:36 yes, except the planet limit was raised and I forgot to update the code. 2010-12-02T12:07:40 yeah 2010-12-02T12:07:44 i had wondered. 2010-12-02T12:07:49 but i'm sure 16 bits sufficed nonetheless 2010-12-02T12:07:50 a1k0n_: no it's calculating the width 2010-12-02T12:07:54 yes 2010-12-02T12:10:02 blog link? 2010-12-02T12:10:09 http://quotenil.com/Planet-Wars-Post-Mortem.html 2010-12-02T12:10:09 by the way, this was hilarious: http://a1k0n.net/lisp/anon.html 2010-12-02T12:10:32 haha, how did you find that? 2010-12-02T12:11:24 #lisp is probably the second-worst support group next to #unix 2010-12-02T12:11:45 *** amstan has joined #aichallenge 2010-12-02T12:11:45 *** ChanServ sets mode: +o amstan 2010-12-02T12:13:03 hmm. also Giraffes? Giraffes! is a band, which i didn't know at the time 2010-12-02T12:13:46 anyway, congrats mega1, you decisively won. :) 2010-12-02T12:13:59 (4:03:51 AM) a1k0n_: hah, and now we know what bocsimacko is. 2010-12-02T12:14:01 what? 2010-12-02T12:14:12 antimatroid: read the bottom of his post 2010-12-02T12:15:01 ah yes, i am always bad at making it to the end of things :) 2010-12-02T12:15:41 is that hungarian or czech? 2010-12-02T12:16:01 the author is czech, I think 2010-12-02T12:16:11 (I'm feeding the kids, bbl) 2010-12-02T12:16:27 later 2010-12-02T12:16:42 nice article mega1 2010-12-02T12:22:17 hopefully later today I will get the range to 90% likelihood of superiority added to the final rankings 2010-12-02T12:22:27 http://accountabuse1.mail.corp.sp1.yahoo.com/mtastats/index.php?primary_domain=yahoo.com.tw&domain=yahoo.com.tw&granularity=300&show_mxes=1&show_stats=420,421,connectFailed,recvFailed,connectTimedOut,recvTimedOut 2010-12-02T12:22:32 http://yw0.ops.tp2.yahoo.com:9999/ywatch.ops/?host=rmip100.mail.tp2.yahoo.com&startDate=-1d 2010-12-02T12:22:40 the dips line up with the spikes in disk i/o 2010-12-02T12:22:45 bhasker: wrong channel 2010-12-02T12:22:57 it's actually already calculated of course I just have to modify the database to hold the information 2010-12-02T12:23:11 doh 2010-12-02T12:23:12 :-\ 2010-12-02T12:23:31 was meant to paste it in messenger 2010-12-02T12:30:45 *** Azrathud has joined #aichallenge 2010-12-02T12:34:39 man people don't give it a rest in the forums do they 2010-12-02T12:35:00 no matter when you "start" the contest, some people are going to start before others 2010-12-02T12:35:19 and the map generator was open source, how is there anything shady about that? 2010-12-02T12:36:47 *** lavalamp has joined #aichallenge 2010-12-02T12:37:43 Did mega1 post his write up? 2010-12-02T12:38:46 yes 2010-12-02T12:38:49 @mega1 2010-12-02T12:38:50 antimatroid: User error, it's not my fault. 2010-12-02T12:38:55 @bocsimacko 2010-12-02T12:38:56 antimatroid: bocsimacko = http://quotenil.com/ 2010-12-02T12:41:48 thanks 2010-12-02T12:45:52 a1k0n_: I'm back if you have questions 2010-12-02T12:46:28 *** bhasker has quit IRC (Quit: bhasker) 2010-12-02T12:49:14 http://ai-contest.com/profile.php?user_id=10752 2010-12-02T12:49:15 ehehe 2010-12-02T12:50:23 "worst sponsorship deal ever" according to him 2010-12-02T12:50:36 couldn't have put it better myself 2010-12-02T12:50:46 although not sure he and i would be talking about the same side of things 2010-12-02T12:54:01 *** mceier has joined #aichallenge 2010-12-02T12:54:56 mega1: did you see the top ranked comment on reddit :P 2010-12-02T12:55:01 http://www.reddit.com/r/programming/comments/eesxh/google_ai_challenge_final_rankings_released/ 2010-12-02T12:57:59 @contestbot learn mega1 as http://quotenil.com/ 2010-12-02T12:58:00 Frontier: An error occured while trying to show the previous error. 2010-12-02T12:58:29 Frontier: it already has it for bocsimacko 2010-12-02T12:58:44 antimatroid: how can people live without reddit ... 2010-12-02T12:59:13 mega1: i did a pretty good job the last 3 months 2010-12-02T12:59:27 I know. The comments on Reddit are as expected: "Do I need to continue reading it after he says Lisp isn't a compiled language?", :p 2010-12-02T12:59:46 hey, that was an informed comment :-) 2010-12-02T12:59:53 *** superflit has joined #aichallenge 2010-12-02T13:00:01 it was exactly as mega1 pretty much predicted 2010-12-02T13:00:17 too many people, too few resources to go around? 2010-12-02T13:00:29 ? 2010-12-02T13:00:40 obscure fallout 2 reference 2010-12-02T13:00:42 i was referring to people bitching about the comment on the homepage 2010-12-02T13:01:09 i can't break the habit of checking my profile page 2010-12-02T13:01:25 i've loaded it at least 10 times today 2010-12-02T13:04:35 antimatroid: same for me... and each time I say to myself: "Nothing will change now, stop it" 2010-12-02T13:07:35 now i'm regretting not working hard enough to get into the top 100 at least 2010-12-02T13:07:57 I'm regretting starting a rewrite around Nov 20 and never finishing 2010-12-02T13:08:52 at least one of my 10 branches was pretty close in approach; they all just suffered from some muddy thinking. i had problems getting coordinated attacks working without constantly invalidating earlier coordinated attacks 2010-12-02T13:09:24 in the end i was a one-trick pony: i data mined an objective function and then greedily followed it. 2010-12-02T13:09:44 and how did that objective function look like? 2010-12-02T13:09:46 (...on the early maps) 2010-12-02T13:10:00 what was it a function of? 2010-12-02T13:10:02 it was a potential field sort of thing. 2010-12-02T13:10:44 the ships on each planet affected the 'potential' of the other planets by dist^-lambda falloff; lambda was optimized by observing a lot of games 2010-12-02T13:11:32 *** bhasker has joined #aichallenge 2010-12-02T13:11:40 so basically any time i moved ships from a to b i would 'lose' the potential on a and gain it on b 2010-12-02T13:11:58 i of course plotted out the entire timeline of the game, etc, and used a fixed horizon to evaluate these things in the future 2010-12-02T13:12:18 *** Migi32 has joined #aichallenge 2010-12-02T13:12:20 the dynamic horizon approach hadn't occurred to me at all.. nor anyone else i think 2010-12-02T13:13:09 anyway, that was just one of my bots. it did better than some others. i don't really like the approach overall. and changing the maps didn't help cuz i didn't retrain it 2010-12-02T13:13:29 (that was 0.5.1 i think) 2010-12-02T13:13:58 were there many fundamentally different ones? 2010-12-02T13:14:01 *** Migi32 has left #aichallenge 2010-12-02T13:14:14 i tried using simulated annealing over the future action space 2010-12-02T13:14:29 but that's where it would get completely destroyed by conflicting actions with no real way to resolve the conflict 2010-12-02T13:14:39 so it was useless in the end 2010-12-02T13:15:41 i also wrote a complete alpha-beta solver using all actions. it was very, very, very slow. i just wanted to know how slow, and what the optimal actions were on small games. 2010-12-02T13:15:54 *** Cold-Phoenix has joined #aichallenge 2010-12-02T13:17:59 i need to look into your code some more to figure out how you narrow the possible move space 2010-12-02T13:18:23 all actions? that's not very slow, that's beyond the computational power of the universe? 2010-12-02T13:18:37 on very small maps with a low number of ships it isn't 2010-12-02T13:19:14 did it handle simultaneous moves as well? 2010-12-02T13:19:18 yeah 2010-12-02T13:19:26 i had to truncate the search depth though 2010-12-02T13:19:34 come to think of it. 2010-12-02T13:19:40 cool, that's a useful tool to test faster algos 2010-12-02T13:19:56 but yeah it's a lot of moves. 2010-12-02T13:20:27 do you actually separately decide how many ships to send, and later where to get them from? 2010-12-02T13:20:33 yes 2010-12-02T13:20:49 'target step' is the vector of needed arrivals 2010-12-02T13:20:53 yeah 2010-12-02T13:21:11 and I experimented with different ways of satisfying those needs 2010-12-02T13:21:17 but... how many potential sources do you generate for each target step? 2010-12-02T13:21:31 in the final code only one 2010-12-02T13:21:44 oh, so it's a greedy fixup kind of thing more than an actual move generator 2010-12-02T13:21:51 I mean one set of orders (= one 'step') 2010-12-02T13:22:11 but how many target steps do you generate and how? 2010-12-02T13:23:14 or step targets rather 2010-12-02T13:24:21 a1k0n_: up to two in the final code 2010-12-02T13:24:25 no, 3 2010-12-02T13:24:43 one to defend against fleets in space 2010-12-02T13:25:20 one to snipe 2010-12-02T13:26:21 and one more certain variant, see SUREP 2010-12-02T13:42:16 *** lavalamp has quit IRC (Quit: Page closed) 2010-12-02T13:48:36 *** Palmik has quit IRC (Ping timeout: 240 seconds) 2010-12-02T13:51:55 looks like the contest site is dead 2010-12-02T13:57:52 *** p4p4 has quit IRC (Read error: Connection timed out) 2010-12-02T14:00:11 *** p4p4 has joined #aichallenge 2010-12-02T14:05:01 gah, sorry had a database backup in progress but there wasn't enough room left on the drive for it 2010-12-02T14:12:21 *** smloh1 has joined #aichallenge 2010-12-02T14:13:44 mega1: thanks for the great blog post 2010-12-02T14:15:51 mega1: i have a question about how your bot managed to do those draw-breaking moves near the end of the game where it lured the opponent to expend ships on neutrals 2010-12-02T14:16:09 *** ulope has joined #aichallenge 2010-12-02T14:16:19 mega1: was it hard-coded or a natural result of optimizing for turn 200? 2010-12-02T14:17:36 hmm, seems like Jeff linking to the wikipedia ELO page is probably going to mislead a lot of people :( 2010-12-02T14:19:02 mega1: like here: http://ai-contest.com/visualizer.php?game_id=9456606 2010-12-02T14:19:06 not that I have a better place to send them 2010-12-02T14:19:39 mega1: " evil, evil move by bocsi" 2010-12-02T14:20:08 smloh1: yeah, it seems he picked up several wins with that 2010-12-02T14:21:08 *** choas has joined #aichallenge 2010-12-02T14:21:11 janzert: yeah it was pretty brilliant, so i wondered if it had to be manually identified as a strategy or if the bot managed to figure it out 2010-12-02T14:21:55 hahaha brilliant 2010-12-02T14:22:22 Zannick: sorry, i meant evil ;) 2010-12-02T14:22:31 no, i think it's brilliant 2010-12-02T14:22:33 *** Frontier has quit IRC (Ping timeout: 276 seconds) 2010-12-02T14:26:25 *** Mistmanov has joined #aichallenge 2010-12-02T14:26:48 *** smloh1 has quit IRC (Ping timeout: 265 seconds) 2010-12-02T14:31:49 *** Epsilon012 has left #aichallenge 2010-12-02T14:33:12 *** Smloh1 has joined #aichallenge 2010-12-02T14:33:28 *** p4p4 has quit IRC (Read error: Connection timed out) 2010-12-02T14:34:11 *** p4p4 has joined #aichallenge 2010-12-02T14:34:46 wow, just watched the bocsimacko tiebreaker move, awesome 2010-12-02T14:35:53 *** Migi32 has joined #aichallenge 2010-12-02T14:35:57 *** Migi32 has left #aichallenge 2010-12-02T14:36:36 how would a top-level bot that bocsimacko plays against evaluate that neutral as a win? that's weird 2010-12-02T14:36:52 you don't eval past the end of the game, pretty simple 2010-12-02T14:37:33 I guess they weren't doing eval that way 2010-12-02T14:37:52 just figuring out which planets they could capture and keep 2010-12-02T14:37:55 *** Smloh1 has quit IRC (Ping timeout: 265 seconds) 2010-12-02T14:45:44 *** Accoun has quit IRC () 2010-12-02T14:51:45 *** lavalamp has joined #aichallenge 2010-12-02T14:52:54 that is hilarious 2010-12-02T14:54:35 *** a1k0n_ is now known as a1k0n 2010-12-02T14:56:11 isnt that like a bug, if the other bot had just sent all ships to the planet and not acquired more neutrals mega1s bot would have lost? 2010-12-02T14:59:05 Smloh1: I hadn't seen the meaning of this moves by bocsimacko because I never expand if the ships spend can't be refund before turn 200, but it seems pretty effective yes 2010-12-02T15:04:33 i think my bot woud just go after his planet 2010-12-02T15:04:39 and not attack neutrals 2010-12-02T15:05:11 should try it out on that map against mega1s bot 2010-12-02T15:06:02 bhasker: mega1 wouldn't have lost 2010-12-02T15:06:14 ah his ships can't reach 2010-12-02T15:06:18 it would still be a draw 2010-12-02T15:06:19 I think he's sending his ships when num_turn+mindist=200 2010-12-02T15:06:20 at turn 200 2010-12-02T15:07:47 so basically launch and hope that the other bot does dumb stuff 2010-12-02T15:07:54 since there is nothing to lose? 2010-12-02T15:11:00 I suspect that all moves evaluated the same and he just happened to pick that one 2010-12-02T15:14:50 heh, just watched that game. nice move ;-) my bot would probably get screwed by that move as well 2010-12-02T15:15:40 my bot starts using "total ship count at turn 200" as its eval function when it gets near the end 2010-12-02T15:16:01 so it shouldn't do that particular stupid thing 2010-12-02T15:19:55 I'd be interested in seeing what yours does do lavalamp 2010-12-02T15:20:22 *** p4p4 has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2010-12-02T15:20:25 *** yasith has quit IRC (Ping timeout: 264 seconds) 2010-12-02T15:20:37 anyone from estonia in here? :) 2010-12-02T15:20:40 I was planning on releasing the source but I scrambled it up so badly near the end that now I'm a little embarrassed to do so... :/ 2010-12-02T15:21:08 lavalamp: meh mine is terrible but i just stuck it on github... 2010-12-02T15:21:33 are you the one coding Go? 2010-12-02T15:24:53 *** RainCT has joined #aichallenge 2010-12-02T15:27:56 Yeah that's me :) 2010-12-02T15:28:03 OK, I made it public: https://bitbucket.org/lavalamp/gromble 2010-12-02T15:28:08 *cringes* 2010-12-02T15:29:10 *** rebelxt has left #aichallenge 2010-12-02T15:29:51 cool 2010-12-02T15:30:11 i've never looked at Go beyond some example snippets 2010-12-02T15:30:41 have fun then :) 2010-12-02T15:30:57 But try not to learn anything from that, it got pretty messy 2010-12-02T15:32:40 *** fgump has joined #aichallenge 2010-12-02T15:33:30 _ is whatever got assigned last= 2010-12-02T15:33:32 ? 2010-12-02T15:33:56 *** yasith has joined #aichallenge 2010-12-02T15:35:09 _ means "ignore this value" 2010-12-02T15:35:36 primarily you use it to ignore one or more values from functions that have multiple returns 2010-12-02T15:37:52 oh ok 2010-12-02T15:38:18 and thats mandatory? 2010-12-02T15:38:23 _ = f 2010-12-02T15:38:31 No 2010-12-02T15:38:32 you can't just do f 2010-12-02T15:38:33 ? 2010-12-02T15:38:38 f() is fine 2010-12-02T15:38:44 it ignores all returns 2010-12-02T15:38:58 https://bitbucket.org/lavalamp/gromble/src/80c1abab0a0c/engine.go#cl-479 2010-12-02T15:39:18 whats the point of that then? 2010-12-02T15:39:24 ah, that's to avoid the "you didn't use this variable" compilation error 2010-12-02T15:39:34 ah :D 2010-12-02T15:39:55 only there because I was commenting in/out blocks of code there 2010-12-02T15:40:05 like I said, it's really messy :/ 2010-12-02T15:40:13 *** sigh has joined #aichallenge 2010-12-02T15:41:15 no worries... i'm just figuring out the syntay, its interesting 2010-12-02T15:44:14 was c++ your language before go? 2010-12-02T15:44:28 yeah, C++ is my day job 2010-12-02T15:44:31 ok 2010-12-02T15:44:49 because to me it looks pretty much like an updated C++ 2010-12-02T15:45:04 with some ideas from python or ruby 2010-12-02T15:45:46 Yeah, it is similar but different in the right ways :) 2010-12-02T15:45:52 hehe 2010-12-02T15:46:09 compiling is supposed to be really fast, right? 2010-12-02T15:46:29 *** superflit has quit IRC (Quit: superflit) 2010-12-02T15:47:57 yeah 2010-12-02T15:48:02 it's super fast 2010-12-02T15:48:12 like < 2 seconds for my bot 2010-12-02T15:50:23 sorry if i'm getting on your nerves just tell me :D func (ob *Oblina) means its a member function of Oblina? 2010-12-02T15:50:39 no worries :) correct, that's what it means 2010-12-02T15:50:57 and "ob" serves the purpose of the "this" pointer in C++ 2010-12-02T15:51:34 aha is that just convention? doesn't get highlighting... 2010-12-02T15:52:02 nah, you can name it whatever you want 2010-12-02T15:52:22 something short is good since you use it to access member variables, etc 2010-12-02T15:52:28 yea 2010-12-02T15:54:53 geez go is a stupid name for a language, googling things keeps turning up other languages :D 2010-12-02T15:55:06 yeah I know XP 2010-12-02T15:55:07 *** sigh has quit IRC (Remote host closed the connection) 2010-12-02T15:55:10 google of all companies should have thought of that 2010-12-02T15:55:11 use "golang" instead 2010-12-02T15:55:18 yeah it's very ironic 2010-12-02T15:59:17 *** Palmik has joined #aichallenge 2010-12-02T15:59:43 *** phreeza_ has joined #aichallenge 2010-12-02T15:59:57 lol "golang indentation" 2010-12-02T16:00:09 4chan is one of the top hits 2010-12-02T16:01:07 *** medrimonia1 has joined #aichallenge 2010-12-02T16:01:40 haha 2010-12-02T16:02:00 *** medrimonia has quit IRC (Ping timeout: 260 seconds) 2010-12-02T16:03:18 *** phreeza has quit IRC (Ping timeout: 276 seconds) 2010-12-02T16:09:28 hey is the website down? 2010-12-02T16:12:39 *** Meatkat has joined #aichallenge 2010-12-02T16:16:10 looks like it 2010-12-02T16:16:24 the wrath of accoun? :D 2010-12-02T16:18:05 bah, I want to make some database changes but before doing that I want a backup. It seems mysql will keep responding for only so long while the backup is in progress then decides to stop responding to queries. 2010-12-02T16:19:23 anyway the backup is through 28G of 57G, maybe I'll just kill it and do it in the wee hours of the morning 2010-12-02T16:22:02 janzert: well the wee hours in your morning are our afternoon :P 2010-12-02T16:22:05 doesn't realy matter hehe 2010-12-02T16:22:59 yeah, but overall traffic is still less in the wee hours of north america :} 2010-12-02T16:25:38 janzert: switching to InnoDB tables would probably help with that 2010-12-02T16:26:57 politik: yeah, there are many reasons mysql w/ myisam aren't a good choice but that was all done before my time 2010-12-02T16:27:55 *** vazhev_alex has joined #aichallenge 2010-12-02T16:28:35 having pretty much only worked with postgresql before, this had been a learning experience on why everyone complains about mysql :P 2010-12-02T16:28:41 *** RainCT has quit IRC (Remote host closed the connection) 2010-12-02T16:31:35 *** Frontier has joined #aichallenge 2010-12-02T16:31:46 although some of tables it looks like someone was having performance problems with so just started piling random indexes on :( I've been afraid to rip them back off though since I *don't* know the real history 2010-12-02T16:35:11 hey janzert did aigamedev sponsort this aswell? 2010-12-02T16:35:35 all I know about it is what they've written in the forums 2010-12-02T16:35:53 I hope Jeff will comment on it 2010-12-02T16:36:59 I just found out the other night, it seems Google does supply at least some money as well as the name 2010-12-02T16:37:10 but I don't know to what extent 2010-12-02T16:39:21 *** vazhev_alex has quit IRC (Quit: Page closed) 2010-12-02T16:39:27 well lets hope they'll kick in a huge amount of servers next time 2010-12-02T16:39:30 :) 2010-12-02T16:40:06 *** superflit has joined #aichallenge 2010-12-02T16:40:36 to be honest i hope jeff will have something to say about it 2010-12-02T16:41:19 well if there were more sponsors and they aren't mentioned its kind of weird 2010-12-02T16:41:32 cause otherwise i won't believe they're actually sponsors 2010-12-02T16:50:02 *** tapwater has joined #aichallenge 2010-12-02T16:50:28 *** Naktibalda has joined #aichallenge 2010-12-02T16:59:49 *** ebrahim has joined #aichallenge 2010-12-02T16:59:49 *** ebrahim has joined #aichallenge 2010-12-02T17:00:59 *** Palmik has quit IRC (Remote host closed the connection) 2010-12-02T17:05:13 *** davidjliu has joined #aichallenge 2010-12-02T17:08:50 *** amstan has quit IRC (Remote host closed the connection) 2010-12-02T17:14:30 *** sigh has joined #aichallenge 2010-12-02T17:26:02 *** choas has quit IRC (Quit: leaving) 2010-12-02T17:45:48 *** Epona has joined #aichallenge 2010-12-02T17:48:51 *** phreeza_ has quit IRC (Quit: Computer has gone to sleep) 2010-12-02T17:53:08 *** Epona has quit IRC (Quit: Leaving) 2010-12-02T18:01:08 *** ebrahim has quit IRC (Remote host closed the connection) 2010-12-02T18:02:41 janzert: can Accoun be unbanned? 2010-12-02T18:04:09 on irc? any reason? 2010-12-02T18:05:18 *** Appleman1234 has joined #aichallenge 2010-12-02T18:05:55 well.. he was in top10 but very annoying 2010-12-02T18:06:05 I didn't have anything to do with banning him in any location, so I'm the wrong one to talk to about it :( 2010-12-02T18:06:47 I'm pretty good now at not taking him to heart so it's fine with me 2010-12-02T18:06:56 oh, hm. 2010-12-02T18:07:12 amstan banned him about 10 hours ago because he was spamming the link to his forum post 2010-12-02T18:07:45 *** ChanServ sets mode: +o Zannick 2010-12-02T18:07:56 *** Zannick sets mode: -q *!*@static-213-88-2-78.chebnet.ru 2010-12-02T18:08:06 *** Zannick sets mode: -o Zannick 2010-12-02T18:08:27 i'm going to leave the hard bans in place for now 2010-12-02T18:10:02 heh, the other problem I would have actually is that it's been so long since I've had ops in any irc channel that I don't remember how anymore :P 2010-12-02T18:10:54 so... any idea of next contest? 2010-12-02T18:14:14 *** Epona has joined #aichallenge 2010-12-02T18:16:26 next contest will involve some sort of game 2010-12-02T18:16:37 players will enter by writing bots to play said game 2010-12-02T18:18:03 or maybe itll be reverse, bots will program people to play games 2010-12-02T18:18:18 you will program a bot that will program a bot 2010-12-02T18:18:42 or the games will play bots that will program people 2010-12-02T18:18:44 what? 2010-12-02T18:19:21 o.o 2010-12-02T18:19:23 we give players a bot and they have to program a game the bot wins at? 2010-12-02T18:19:25 scariest game ever 2010-12-02T18:20:14 it'd be interesting to do a related field of games... like at the beginning the server tells your bot what the rules are 2010-12-02T18:20:14 Zannick: that could actually be an interesting exercise :P 2010-12-02T18:20:27 like the dual problem of this contest 2010-12-02T18:20:52 better yet 2010-12-02T18:21:31 take an arbitrary top10 bot and generate a set of maps such that that bot wins the final 2010-12-02T18:21:38 Zannick: thank you, I asked him to be polite. 2010-12-02T18:21:38 we could call it the Accoun AI contest 2010-12-02T18:21:45 haha 2010-12-02T18:21:52 mega1: cool 2010-12-02T18:22:05 they need to make an ai contest exclusive to ai 2010-12-02T18:22:13 at least 2010-12-02T18:22:27 Epona: what do you mean? 2010-12-02T18:23:12 idk about all of them, but bocsimacko, the winning bot had no neural nets or evolutionary algorithms 2010-12-02T18:23:53 Epona: you have an overly narrow definition of AI compared to everyone else 2010-12-02T18:24:50 *** JamesMG has quit IRC (Quit: home) 2010-12-02T18:24:50 I know 2010-12-02T18:24:55 its a curse 2010-12-02T18:25:33 not really a problem, people have been saying similar things since computers were invented :P 2010-12-02T18:26:06 emphasise the A in AI 2010-12-02T18:26:16 rather the I 2010-12-02T18:26:22 thing is, AI is just whatever works 2010-12-02T18:27:23 *** lavalamp has quit IRC (Quit: Page closed) 2010-12-02T18:27:26 idk. doesn't seem as productive to me 2010-12-02T18:27:35 why? 2010-12-02T18:27:35 *** bhasker_ has joined #aichallenge 2010-12-02T18:27:53 the start of most introductory AI courses is often advanced search algorithms 2010-12-02T18:28:23 because they work, or at least provide a very good base for many problem in the domain 2010-12-02T18:28:32 *** bhasker__ has joined #aichallenge 2010-12-02T18:28:47 but of course, you will argue that those algorithms aren't real AI :P 2010-12-02T18:28:58 no no no I see your point 2010-12-02T18:29:13 algorithms are real AI. I am just favoring ceretain types I guess 2010-12-02T18:29:19 is 'google.com' ai? :P 2010-12-02T18:29:37 idk, ssh into it and find out 2010-12-02T18:29:51 *** bhasker has quit IRC (Ping timeout: 255 seconds) 2010-12-02T18:30:05 I like the definition that wikipedia sites: AI textbooks define the field as "the study and design of intelligent agents" where an intelligent agent is a system that perceives its environment and takes actions that maximize its chances of success. 2010-12-02T18:30:29 takes actions that TRY to maximize it 2010-12-02T18:30:36 it says nothing about how AI is implemented, just specifies the type of problem being addressed 2010-12-02T18:30:44 burny: true :) 2010-12-02T18:32:07 *** bhasker_ has quit IRC (Ping timeout: 255 seconds) 2010-12-02T18:32:53 *** Mistmanov has quit IRC (Quit: Page closed) 2010-12-02T18:36:36 *** jaspervdj has quit IRC (Quit: NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN) 2010-12-02T19:02:13 hmm, yes, the next contest should involve perception of some kind 2010-12-02T19:04:08 *** McLeopold has joined #aichallenge 2010-12-02T19:06:03 perception of what kind? 2010-12-02T19:06:12 would the state be fuzzy? 2010-12-02T19:06:17 fog of war? 2010-12-02T19:06:32 *** bhasker has joined #aichallenge 2010-12-02T19:07:13 maybe a game like planet wars, where you don't know the cost of a planet until you attack it sufficiently 2010-12-02T19:08:27 *** bhasker__ has quit IRC (Ping timeout: 240 seconds) 2010-12-02T19:10:51 *** superflit has quit IRC (Ping timeout: 272 seconds) 2010-12-02T19:16:46 *** jaspervdj has joined #aichallenge 2010-12-02T19:17:08 *** jaspervdj has quit IRC (Client Quit) 2010-12-02T19:17:33 *** narnach_mb has quit IRC (Quit: Leaving.) 2010-12-02T19:29:47 *** mceier has quit IRC (Quit: leaving) 2010-12-02T19:30:50 *** McLeopold has quit IRC (Quit: Leaving) 2010-12-02T19:38:30 *** Mathnerd314 has quit IRC (Ping timeout: 240 seconds) 2010-12-02T20:26:37 *** rebelxt_ has joined #aichallenge 2010-12-02T20:28:49 *** rebelxt_ has quit IRC (Client Quit) 2010-12-02T20:30:33 *** Epona has quit IRC (Ping timeout: 265 seconds) 2010-12-02T20:34:44 *** dmj111 has joined #aichallenge 2010-12-02T20:37:06 *** fgump has quit IRC (Quit: Ex-Chat) 2010-12-02T20:37:52 What would people think if the site were to go back to an open tournament, the current final rankings would remain frozen but people could continue with an informal tournament if they wanted to? 2010-12-02T20:38:24 Not saying there is any certain possibility for this just wondering what people think of the idea even 2010-12-02T20:38:42 I think people would like it: but I think it wouldn't be used as much as many people would think 2010-12-02T20:39:21 *** Epona has joined #aichallenge 2010-12-02T20:52:52 *** bhasker has quit IRC (Quit: bhasker) 2010-12-02T20:53:16 *** Sylph has quit IRC (Ping timeout: 265 seconds) 2010-12-02T20:57:49 *** Sylph has joined #aichallenge 2010-12-02T21:03:20 *** Appleman1234 has quit IRC (Read error: Connection reset by peer) 2010-12-02T21:06:04 *** smloh1 has joined #aichallenge 2010-12-02T21:13:23 *** Mjothvitnir has joined #aichallenge 2010-12-02T21:14:30 *** Epona has quit IRC (Ping timeout: 240 seconds) 2010-12-02T21:26:16 *** Naktibalda has quit IRC (Remote host closed the connection) 2010-12-02T21:34:41 *** McLeopold has joined #aichallenge 2010-12-02T21:34:53 is mega1 around? 2010-12-02T21:36:12 *** Rubicon-|-Cross has joined #aichallenge 2010-12-02T21:54:16 *** iNtERrUpT has joined #aichallenge 2010-12-02T22:02:02 *** yasith_ has joined #aichallenge 2010-12-02T22:04:09 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-12-02T22:04:33 *** Mathnerd314 has joined #aichallenge 2010-12-02T22:06:55 *** McLeopold has quit IRC (Quit: Page closed) 2010-12-02T22:10:36 *** cYmen has quit IRC (Ping timeout: 265 seconds) 2010-12-02T22:10:50 *** cYmen has joined #aichallenge 2010-12-02T22:13:23 *** yasith__ has joined #aichallenge 2010-12-02T22:15:13 *** yasith_ has quit IRC (Ping timeout: 255 seconds) 2010-12-02T22:20:23 janzert: I would like that 2010-12-02T22:25:39 *** yasith__ has quit IRC (Ping timeout: 245 seconds) 2010-12-02T22:29:48 *** Appleman1234 has joined #aichallenge 2010-12-02T22:29:55 but you shouldn't play with my emotions like that :) 2010-12-02T22:45:24 *** yasith has joined #aichallenge 2010-12-02T22:45:37 *** Epona has joined #aichallenge 2010-12-02T22:51:03 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-12-02T22:52:58 anyoen for some 1on1 or 2on2 galcon flash? 2010-12-02T22:53:14 *** Epona has left #aichallenge ("Leaving") 2010-12-02T22:57:16 *** Tobu has quit IRC (Quit: Leaving.) 2010-12-02T23:07:41 *** yasith has joined #aichallenge 2010-12-02T23:19:50 *** wh1teside_ has joined #aichallenge 2010-12-02T23:23:25 *** wh1teside has quit IRC (Ping timeout: 264 seconds) 2010-12-02T23:26:15 *** Frontier has quit IRC (Ping timeout: 240 seconds) 2010-12-02T23:26:19 *** Mjothvitnir has quit IRC (Quit: Page closed) 2010-12-02T23:28:15 *** yasith has quit IRC (Ping timeout: 260 seconds) 2010-12-02T23:32:30 *** yasith has joined #aichallenge 2010-12-02T23:34:51 *** dsfdsfd has joined #aichallenge 2010-12-02T23:35:24 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&start=170#p8057 2010-12-02T23:35:48 LOL 2010-12-02T23:35:50 hidden canadian reasons 2010-12-02T23:37:05 Accoun's posts are written by a small team of clown comedians, i swear. I refuse to believe one person is behind this 2010-12-02T23:38:22 let janzert publish his hidden canadian resons 2010-12-02T23:38:33 it so mirakle 2010-12-02T23:38:44 lol 2010-12-02T23:38:57 look at the follow-up post 2010-12-02T23:38:58 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&sid=f72c43cb8b9da27dd6bbb02a79dfcf2c&start=180 2010-12-02T23:39:06 "why?" 2010-12-02T23:39:20 cnadians chiting again! 2010-12-02T23:39:26 yes 2010-12-02T23:39:32 hate all canadian! 2010-12-02T23:39:47 blame canada! 2010-12-02T23:40:00 Is janzert even canadian? 2010-12-02T23:40:04 i don't think so 2010-12-02T23:40:06 He's not 2010-12-02T23:40:08 he's using charter 2010-12-02T23:40:12 ~lol~ 2010-12-02T23:40:13 kill all man rape all child and woman! 2010-12-02T23:40:17 @ban dsfdsfd 2010-12-02T23:40:18 pgpaskar_: You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2010-12-02T23:40:26 I hate you contestbot 2010-12-02T23:40:29 1 2010-12-02T23:40:58 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&start=170#p8057 2010-12-02T23:41:16 dsfdsfd: please, can you stop now? the contest is over 2010-12-02T23:41:30 di.fm hardstyle 2010-12-02T23:41:31 not 2010-12-02T23:41:35 jeeeeeeeeeeeeeesus this music owns 2010-12-02T23:41:40 im fanatic candaian hater 2010-12-02T23:41:50 dsfdsfd: draw me a picture of a huge erect dick 2010-12-02T23:41:57 and I'll tajke a look at your claims 2010-12-02T23:42:11 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&start=170#p8057 2010-12-02T23:42:27 seriously, stop spamming the channel 2010-12-02T23:42:45 it is somehow entertaining 2010-12-02T23:43:33 Accoun should have a show on comedy 2010-12-02T23:43:34 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-12-02T23:43:56 i think i am becoming a fan of the russian sense of humour 2010-12-02T23:44:40 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&start=170#p8057 2010-12-02T23:45:02 *** ChanServ sets mode: +o sigh 2010-12-02T23:47:12 *** Azrathud has joined #aichallenge 2010-12-02T23:47:19 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&start=170#p8057 2010-12-02T23:48:30 *** epona has joined #aichallenge 2010-12-02T23:48:35 http://ai-contest.com/forum/viewtopic.php?f=18&t=424&start=170#p8057 2010-12-02T23:48:38 *** sigh sets mode: +q *!*@*85.31.186.70 2010-12-02T23:48:57 did that do what it was supposed to> 2010-12-02T23:49:30 i think hes mad ._. 2010-12-02T23:49:37 i hope he isnt 2010-12-02T23:49:45 *** sdfasfasdsfda has joined #aichallenge 2010-12-02T23:49:55 it's quite ridiculous 2010-12-02T23:50:02 here he is again 2010-12-02T23:52:14 *** sdfsdf has joined #aichallenge 2010-12-02T23:53:41 *** ChanServ sets mode: -o sigh 2010-12-02T23:53:45 *** dsfdsfd has quit IRC (Ping timeout: 265 seconds) 2010-12-02T23:54:43 *** sdfasfasdsfda has quit IRC (Ping timeout: 265 seconds) 2010-12-02T23:59:33 *** sdfsdf has quit IRC (Ping timeout: 265 seconds)