2011-07-16T00:03:50 *** foRei has quit IRC (Quit: Bye) 2011-07-16T00:24:54 *** AlliedEnvy has quit IRC (Remote host closed the connection) 2011-07-16T00:28:16 *** AlliedEnvy has joined #aichallenge 2011-07-16T00:33:37 evening 2011-07-16T00:33:53 ugh.. 2011-07-16T00:34:08 arguing with ppl in other channel about merits of windows and not using version control 2011-07-16T00:34:16 can't tell if talking to a troll or to a wall 2011-07-16T00:44:04 *** jmreardon has quit IRC (Quit: jmreardon) 2011-07-16T01:57:45 *** mleise has joined #aichallenge 2011-07-16T02:33:47 *** Mathnerd314 has quit IRC (Ping timeout: 264 seconds) 2011-07-16T02:48:11 *** Accoun has quit IRC (Ping timeout: 264 seconds) 2011-07-16T03:01:47 *** Accoun has joined #aichallenge 2011-07-16T03:04:03 *** Eruquen has quit IRC (Ping timeout: 258 seconds) 2011-07-16T03:04:42 *** Eruquen has joined #aichallenge 2011-07-16T03:10:53 *** Accoun has quit IRC (Ping timeout: 258 seconds) 2011-07-16T03:44:00 *** Palmik has joined #aichallenge 2011-07-16T03:58:58 *** nann has joined #aichallenge 2011-07-16T04:02:00 *** Apophis_ has quit IRC (Ping timeout: 255 seconds) 2011-07-16T04:07:01 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-07-16T05:04:21 *** jugde has quit IRC (Ping timeout: 264 seconds) 2011-07-16T05:06:06 *** jugde has joined #aichallenge 2011-07-16T05:22:57 *** jugde has quit IRC (Ping timeout: 264 seconds) 2011-07-16T05:27:46 *** jugde has joined #aichallenge 2011-07-16T05:45:19 *** smellyhippy has quit IRC (Ping timeout: 263 seconds) 2011-07-16T06:11:56 University of SÃ�Â£o Paulo ? there is something wrong with the encoding 2011-07-16T06:21:30 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-07-16T06:22:21 *** smellyhippy has joined #aichallenge 2011-07-16T06:53:33 *** jugde has quit IRC (Ping timeout: 264 seconds) 2011-07-16T07:20:56 *** mleise has quit IRC (Quit: Leaving.) 2011-07-16T07:30:44 *** mcstar has joined #aichallenge 2011-07-16T07:55:50 @later tell mleise did you notice that when the visualizer is in play mode, the fog of war follows the ant with 1 turn(maybe half a turn) delay? but i dont have the latest files, i probably need to check those first :) 2011-07-16T07:55:50 mcstar: Yep. 2011-07-16T07:57:32 antimatroid: i think with a decent map that is pretty unlikely to happen 2011-07-16T07:57:45 antimatroid: and even if it does, what's available to one bot is available to all (assuming symmetry) 2011-07-16T07:58:03 with hills, i think symmetry is more important, relative to with scores instead 2011-07-16T07:58:25 * mcstar thought the maps were 2D 2011-07-16T07:58:47 huh? 2011-07-16T07:58:54 oh hills 2011-07-16T07:59:00 you missed the conversation 2011-07-16T07:59:00 yeah 2011-07-16T07:59:04 i know :) 2011-07-16T07:59:10 we mean ant hills. they are just objects on the map :P 2011-07-16T07:59:27 we were discussing making the game more interesting again 2011-07-16T07:59:49 thats good to hear 2011-07-16T08:09:45 i dont like player_seed 2011-07-16T08:09:55 it should be written as playerseed 2011-07-16T08:10:10 all the other vars. are written without an underscore 2011-07-16T08:17:59 *** chl has quit IRC (*.net *.split) 2011-07-16T08:19:23 *** chl has joined #aichallenge 2011-07-16T08:26:16 *** Kingpin13 has joined #aichallenge 2011-07-16T08:34:51 *** onensora has joined #aichallenge 2011-07-16T08:37:36 *** jmreardon has joined #aichallenge 2011-07-16T09:00:44 i agree 2011-07-16T09:00:58 but that's kind of bikeshedding 2011-07-16T09:05:16 its a delight to learn new words every day 2011-07-16T09:07:06 *** foRei has joined #aichallenge 2011-07-16T09:21:58 *** fbparis has left #aichallenge 2011-07-16T09:24:15 morning 2011-07-16T09:37:46 *** bhasker_ has quit IRC (Quit: bhasker_) 2011-07-16T09:38:50 *** bhasker has joined #aichallenge 2011-07-16T09:43:06 *** AlliedEnvy has quit IRC (Ping timeout: 255 seconds) 2011-07-16T09:45:16 no one around? 2011-07-16T09:46:01 lurkers 2011-07-16T09:46:05 *** jmreardon has left #aichallenge 2011-07-16T09:46:22 *** jmreardon has joined #aichallenge 2011-07-16T09:47:08 jmreadon your bot and mine seem to be pretty close 2011-07-16T09:50:50 ya 2011-07-16T09:55:40 *** AlliedEnvy has joined #aichallenge 2011-07-16T10:05:03 have any of you hit the 1sec limit yet? 2011-07-16T10:05:28 (did you need to rewrite something, so that you dont time out?) 2011-07-16T10:06:42 ya, one version of my bot starting hitting it on large maps 2011-07-16T10:06:51 *** jmreardon has left #aichallenge 2011-07-16T10:07:20 *** jmreardon has joined #aichallenge 2011-07-16T10:07:59 ugh, got to stop closing this window 2011-07-16T10:08:33 that red X must be very tempting to you :) 2011-07-16T10:08:55 yup i did 2011-07-16T10:09:07 most of my query functions i rewrote to be more efficient 2011-07-16T10:09:11 im glad 2011-07-16T10:09:17 especially all the find in radius kind of functions 2011-07-16T10:09:33 haven't had a single timeout since i fixed that 2011-07-16T10:09:53 i had trouble with a recursive logic, it took too much time when my bot had many ants 2011-07-16T10:10:17 im not sure if it was a bug, or a design-failure 2011-07-16T10:10:17 ah k 2011-07-16T10:11:04 I've avoided doing much logic on a per-ant basis; My runtime is mostly bound by map size 2011-07-16T10:11:48 jmreardon: is your bot "board" oriented? 2011-07-16T10:11:53 board=map 2011-07-16T10:11:55 oriented? 2011-07-16T10:12:09 well, if it uses a representation of the map/board often 2011-07-16T10:12:29 umm, I guess, I'll explain the basics if you like 2011-07-16T10:12:31 for example mine is rather ant based, it doesnt use the board 2011-07-16T10:12:43 sure 2011-07-16T10:12:52 not in detail, just roughly :) 2011-07-16T10:14:28 so, when I start my turn, I construct a series of 2d int arrays; then for each ant, they inspect the nearby values on these arrays in a set order, whenever the ant find an adjacent value on the array that is valid and lower than the value of it's current location, it moves to that lower location 2011-07-16T10:14:52 so my ants' behaviour is entirely determined by how I construct the arrays 2011-07-16T10:15:23 ah mine uses the board for most things 2011-07-16T10:15:44 i just use my ant array to iterate over the ants thats about it 2011-07-16T10:15:49 almost every thing else uses the board 2011-07-16T10:16:06 so my queries like find ants in radius X etc use the board exclusively 2011-07-16T10:16:28 for example, one array is for collecting food, in that array, I run a floodfill (avoiding water), starting from a food tile (which has value 0), each tile further from the food has a value one greater than the next tile 2011-07-16T10:16:39 jmreardon: how do you initialize those arrays? do they get the values from a master map? 2011-07-16T10:17:26 i see, squares has distance values 2011-07-16T10:17:30 have 2011-07-16T10:17:48 the arrays are initialized to 2^32-1; currently the flood fill code avoids water, and assumes anything else is land 2011-07-16T10:18:11 jmreardon: im just curious, why do you need many arrays? 2011-07-16T10:18:21 more than 1 actually 2011-07-16T10:18:47 is it because you use haskell? 2011-07-16T10:18:49 because I want my ants to do many things; and it is easier to represent in multiple arrays than get the logic right to fill one array 2011-07-16T10:19:13 jmreardon, so you flood fill the array to get distance to closest food from any square is that it? 2011-07-16T10:19:18 no, my code (well the array stuff at least) isn't terribly functional anyways 2011-07-16T10:19:30 bhasker: no, I have an array for everything my ants do 2011-07-16T10:19:36 ah k 2011-07-16T10:19:45 (ok, except staying still) 2011-07-16T10:20:15 one for food, one for attacking, running away, searching, and spreading out (currently at least) 2011-07-16T10:20:20 ah k 2011-07-16T10:20:32 multiplied this by the number of ants, right? 2011-07-16T10:20:34 i just have one array and i use ancillary structures and query functions to decide per ant what to do 2011-07-16T10:20:49 no i am guessing his array is a 2d array of the map 2011-07-16T10:21:05 some of them (like running away) don't have distance values, since I want them going in the opposite direction 2011-07-16T10:21:28 mcstar: no, one array of each sort is constructed per turn, and it covers the entire map 2011-07-16T10:22:00 I construct a series of 2d int arrays; then for each ant, 2011-07-16T10:22:04 so, I floodfill from each food tile all at the same time, on the same array, any every ant uses that array 2011-07-16T10:22:07 i misunderstood this line :D 2011-07-16T10:22:39 ah, no, each ant inspects the same arrays after they are constructed 2011-07-16T10:23:16 i am guessing when you floodfill you overwrite any squares till they have the lowest number? 2011-07-16T10:23:23 makes it hard to do some clever things like have only one ant go for a food tile, but it scales well 2011-07-16T10:24:09 in other news the amazon micro instances are pretty snappy cpu wise 2011-07-16T10:24:13 just limited to 612 mb of ram 2011-07-16T10:25:12 no, for the floodfill, when the "boundaries" come against each other, I stop floodfilling. the algorithm does the fill breadth-first, so it generally gets the right answer anyways 2011-07-16T10:25:31 ah k 2011-07-16T10:25:48 and I only floodfill for a certain distance, no point having an ant half the map away running for some food 2011-07-16T10:26:46 ah k 2011-07-16T10:29:31 i a bit envy the C guys for their fast arrays 2011-07-16T10:29:42 and fast floating point math 2011-07-16T10:33:28 mcstar: surely lisp has access to some sort of reasonably fast arrays? (I've only used Scheme, and not much at that) 2011-07-16T10:34:10 well, in the sbcl manual they say, that 2d arrays are generally slow 2011-07-16T10:34:18 so i use 1d arrays 2011-07-16T10:34:38 and have functions which translate between 1 and 2 index accesses 2011-07-16T10:35:49 *** chris__0076 is now known as chris_0076 2011-07-16T10:36:09 i had to get my coffee 2011-07-16T10:36:27 and in lisp numbers are boxed 2011-07-16T10:36:52 so you have to explicitely tell the compiler when you want to use unboxed floats/ints 2011-07-16T10:37:05 ic 2011-07-16T10:38:15 is haskell's array access times comparable to c's? 2011-07-16T10:39:03 my understanding is that you can get somewhat close if you want to 2011-07-16T10:39:33 but I don't know much about optimizing array usage in ghc 2011-07-16T10:39:37 heh main reason i picked c++ rather than python for this contest 2011-07-16T10:40:02 last contest i had used python and ran into multiple issues with performance of objects etc 2011-07-16T10:40:11 had to rewrite a bunch of crap to use tuples etc 2011-07-16T10:40:19 makes for highly unreadable code 2011-07-16T10:45:26 http://shootout.alioth.debian.org/ i think this is a good site to get some performance tips 2011-07-16T10:46:41 mcstar: are you hitting performance issues that much? 2011-07-16T10:47:19 no :D, but its a good place to learn 2011-07-16T10:53:24 this shows clearly my ignorance, but what im surprised at, is how well java does on those test, with so little obfuscation 2011-07-16T10:53:33 tests 2011-07-16T10:54:17 java has improved quite a bit over the last decade 2011-07-16T10:56:01 unfortunately Clojure has a factor of 2 disadvantage 2011-07-16T10:59:57 unfortunately clojure doesn't fit into the mold the jvm works well with so much 2011-07-16T11:47:26 *** sigh has quit IRC (Remote host closed the connection) 2011-07-16T11:58:21 *** genericbob has joined #aichallenge 2011-07-16T11:59:13 *** Accoun has joined #aichallenge 2011-07-16T12:05:22 *** genericbob has quit IRC (Quit: leaving) 2011-07-16T13:24:40 *** amstan has joined #aichallenge 2011-07-16T13:24:41 *** ChanServ sets mode: +o amstan 2011-07-16T13:46:31 *** gobang has joined #aichallenge 2011-07-16T14:05:31 *** Eruonen has joined #aichallenge 2011-07-16T14:07:30 *** eashoka has joined #aichallenge 2011-07-16T14:33:59 *** jbrechtel has joined #aichallenge 2011-07-16T14:34:34 Is anyone running a tcp server? I saw someone mention one the other day but don't have the IP anymore 2011-07-16T14:40:30 *** eashoka has quit IRC (Ping timeout: 252 seconds) 2011-07-16T14:42:21 *** amstan has quit IRC (Ping timeout: 255 seconds) 2011-07-16T14:52:30 *** jbrechtel has quit IRC (Ping timeout: 260 seconds) 2011-07-16T14:54:13 *** jbrechtel has joined #aichallenge 2011-07-16T15:09:22 *** jbrechtel has quit IRC (Read error: Connection reset by peer) 2011-07-16T15:09:39 *** jbrechtel has joined #aichallenge 2011-07-16T15:35:41 @later tell mleise i could stand that firefox kept opening up(i think the webbrowser module is not working when no desktop environment is loaded) so i added a subprocess.call to visualize_locally 2011-07-16T15:35:41 mcstar: I think that worked... 2011-07-16T15:36:02 @later tell mleise couldn't 2011-07-16T15:36:02 mcstar: Aye, aye, sir 2011-07-16T15:48:55 *** Accoun has quit IRC () 2011-07-16T15:50:40 *** Crivens has joined #aichallenge 2011-07-16T15:52:42 *** mleise has joined #aichallenge 2011-07-16T15:59:59 *** Accoun has joined #aichallenge 2011-07-16T16:21:54 mcstar: the fog of war is an expensive overlay that I only update once a turn. It is the same that the players see at the start of that turn. 2011-07-16T16:23:17 mleise: it seem as if it was lagging behind the ants 2011-07-16T16:23:25 it was kind of annyoing for me 2011-07-16T16:23:50 its not a big deal, but id like that the fog moves in synchron with the ants 2011-07-16T16:25:12 that's not going to happen, sorry. it is too computationally expensive :) 2011-07-16T16:26:27 there must be some misunderstanding, maybe on my part, but i dont get why would it be "expensive" 2011-07-16T16:29:37 mcstar: btw visualize_locally is McLeopold's work. I don't know Python. I haven't even used visualize_locally yet. I only know the flag for playgame.py to suppress opening a browser window 2011-07-16T16:30:30 well, he then, imports a module called webbrowser, i suppose the problem lies there 2011-07-16T16:30:53 mleise: pls lets figure out if we were talking about the same thing or not 2011-07-16T16:31:46 For the fog of war you have to calculate the visible squares for every ant that takes some time 2011-07-16T16:33:09 i guess what id like, is that you move the fog before the ant 2011-07-16T16:33:10 thats all 2011-07-16T16:33:29 i dont want you to animate the fog 2011-07-16T16:34:54 ok i see. well in that case I'll talk with amstan, antimatroid or McLeopold to get a broader set of opinions on that. Until now I liked the fog to change at the start of the turn as it did when the game was run. 2011-07-16T16:36:36 anyway, its not a big problem(maybe not a problem at all), i usually dont "Play" the replay, but i use the upper slider 2011-07-16T16:41:26 this is horrible, the village i live in must have some kind of "Village Day", and they play some horrible "music" loudly on the football-field 2011-07-16T16:42:10 *** gobang has quit IRC (Read error: Connection reset by peer) 2011-07-16T16:45:21 i think i wrote an intelligent snake 2011-07-16T16:45:47 lets define "intelligence" this way: spontaneous, autonomous behaviour 2011-07-16T16:46:22 when theres no food around, this test version should stay still, but somehow,it reorganizes itself 2011-07-16T16:46:24 hehe 2011-07-16T16:49:40 *** gobang has joined #aichallenge 2011-07-16T16:52:52 *** jbrechtel has joined #aichallenge 2011-07-16T16:53:12 Anyone got the link to that tcp server on github for ants? 2011-07-16T16:54:52 as far as i know it isnt in a usable condition 2011-07-16T16:58:30 really? flag was mentioning running games on it last week 2011-07-16T16:59:29 fwiw, this is what I was looking for: https://github.com/berak/ants-tcp 2011-07-16T17:00:14 yes 2011-07-16T17:00:21 but its quite bad yet 2011-07-16T17:00:34 for example doesnt report water, when it should and such 2011-07-16T17:00:48 i havent used it, but this is what i heard 2011-07-16T17:01:43 ah 2011-07-16T17:01:50 that would indeed be unfortunate 2011-07-16T17:08:42 *** jbrechtel_ has joined #aichallenge 2011-07-16T17:11:58 *** jbrechtel has quit IRC (Ping timeout: 246 seconds) 2011-07-16T17:14:05 http://aichallengebeta.hypertriangle.com/visualizer.php?game=89787&user=13 scoring is broken 2011-07-16T17:14:06 http://aichallengebeta.hypertriangle.com/visualizer.php?game=89787&user=13 2011-07-16T17:14:16 someone with 70% of the ants shouldn't lose the game 2011-07-16T17:14:20 when the cutoff happens 2011-07-16T17:15:13 or the cutoff needs to be amended where its a consistent difference in both score and ants 2011-07-16T17:19:51 *** jbrechtel_ has quit IRC (Ping timeout: 255 seconds) 2011-07-16T17:29:08 *** jbrechtel has joined #aichallenge 2011-07-16T17:29:17 *** mcstar has quit IRC (Quit: WeeChat 0.3.5) 2011-07-16T17:29:18 *** jbrechtel_ has joined #aichallenge 2011-07-16T17:51:11 *** FireFly has joined #aichallenge 2011-07-16T17:55:50 *** gobang has quit IRC (Read error: Connection reset by peer) 2011-07-16T18:05:12 *** amstan has joined #aichallenge 2011-07-16T18:05:12 *** ChanServ sets mode: +o amstan 2011-07-16T18:16:34 *** foRei has quit IRC (Read error: Connection reset by peer) 2011-07-16T18:19:02 *** foRei has joined #aichallenge 2011-07-16T18:30:39 *** Kingpin13 has quit IRC (Quit: this used to be the life, but I don't need another one) 2011-07-16T18:31:12 hi 2011-07-16T19:01:47 *** FireFly has quit IRC (Quit: swatted to death) 2011-07-16T19:08:56 *** jbrechtel has quit IRC (Ping timeout: 252 seconds) 2011-07-16T19:10:58 *** Palmik has quit IRC (Remote host closed the connection) 2011-07-16T19:20:40 *** mleise has quit IRC (Quit: Leaving.) 2011-07-16T19:35:48 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-07-16T19:45:12 *** nann has quit IRC (Read error: Connection reset by peer) 2011-07-16T20:01:07 *** jjmhnct has joined #aichallenge 2011-07-16T20:02:24 *** jjmhnct has quit IRC (Client Quit) 2011-07-16T20:32:26 *** rabidus has quit IRC (Ping timeout: 240 seconds) 2011-07-16T20:34:14 *** Crivens has quit IRC (Read error: Connection reset by peer) 2011-07-16T20:35:06 *** sigh has joined #aichallenge 2011-07-16T21:04:49 *** amstan has quit IRC (Ping timeout: 252 seconds) 2011-07-16T21:28:45 *** amstan has joined #aichallenge 2011-07-16T21:28:45 *** ChanServ sets mode: +o amstan 2011-07-16T21:56:45 *** filter_ has quit IRC (Ping timeout: 276 seconds) 2011-07-16T22:01:00 *** pvarga has joined #aichallenge 2011-07-16T22:23:26 *** fbparis has joined #aichallenge 2011-07-16T22:31:43 *** amstan_ has joined #aichallenge 2011-07-16T22:31:43 *** ChanServ sets mode: +o amstan_ 2011-07-16T22:35:45 *** jmreardon has quit IRC (Ping timeout: 252 seconds) 2011-07-16T22:38:13 *** jmreardon has joined #aichallenge 2011-07-16T22:45:06 *** onensora has quit IRC () 2011-07-16T22:49:53 *** filter has joined #aichallenge 2011-07-16T23:07:04 *** DCaminator has joined #aichallenge 2011-07-16T23:19:30 *** jbrechtel has joined #aichallenge 2011-07-16T23:26:36 *** amstan_ has quit IRC (Ping timeout: 255 seconds) 2011-07-16T23:48:09 *** jmreardon has quit IRC (Quit: jmreardon) 2011-07-16T23:52:27 *** jbrechtel has quit IRC (Ping timeout: 276 seconds)