2011-11-07T00:00:10 its seems like AI's on the tcp server always trick my massive armies to die 2011-11-07T00:00:19 at the expense of like 3 of their ants 2011-11-07T00:00:20 *** Saulzar has quit IRC (Read error: Connection reset by peer) 2011-11-07T00:00:21 *** ChrisH has joined #aichallenge 2011-11-07T00:00:37 why does reconstruct_path want me to add two nodes 2011-11-07T00:00:50 in the pseudo for a* 2011-11-07T00:03:47 *** ChrisH has quit IRC (Client Quit) 2011-11-07T00:04:52 *** Jak_o_Shadows1 has joined #aichallenge 2011-11-07T00:06:09 *** datachomper1 has joined #aichallenge 2011-11-07T00:07:53 *** Jak_o_Shadows has quit IRC (Ping timeout: 245 seconds) 2011-11-07T00:09:02 x_x o_o -_- o_o 2011-11-07T00:10:00 Is the food spawn rate going to be changed for the real tournament? 2011-11-07T00:10:23 i think its changed every game no? 2011-11-07T00:10:55 does any one know who this "A" bot is on the tcp server? Its killing me always... 2011-11-07T00:12:56 A is FlagCapper/_flag 2011-11-07T00:13:54 did he just let it run to kill newbs? 2011-11-07T00:14:07 it shouldn't match newbs up with him 2011-11-07T00:14:07 its been in 6610 games 2011-11-07T00:14:18 but yeah, might as well leave it running 2011-11-07T00:14:23 but wow that's a lot 2011-11-07T00:14:29 i have never won a game, and been mostly killed mostly by A 2011-11-07T00:14:39 thats on tcp server 2011-11-07T00:14:44 http://ants.fluxid.pl/ranking 2011-11-07T00:14:46 sort by games 2011-11-07T00:15:16 *** Areks has joined #aichallenge 2011-11-07T00:16:07 my combat stuff doesn't seem to actually...work 2011-11-07T00:17:16 well, i don't seem to have been responsible for your death at all yet 2011-11-07T00:17:38 what is your bot? 2011-11-07T00:17:50 http://ants.fluxid.pl/replay.22675 2011-11-07T00:18:17 these maps are ridiculous bloodbaths 2011-11-07T00:20:24 *** marina has quit IRC (Quit: marina) 2011-11-07T00:21:01 your bot seems to be doing quite well 2011-11-07T00:21:40 it's not bad, it's just not as good as xathis or A 2011-11-07T00:22:00 i don't even attempt to defend my hills yet 2011-11-07T00:22:26 A kills you here on 1v1 http://ants.fluxid.pl/replay.23970 2011-11-07T00:23:21 i was just watching that 2011-11-07T00:23:40 i have no idea what my last two ants were thinking 2011-11-07T00:24:05 just that they could take a couple ants down with them i guess 2011-11-07T00:24:54 the last 2 out of the hive before its taken over? 2011-11-07T00:24:58 yeah 2011-11-07T00:25:06 "fuck it, we're doomed" 2011-11-07T00:27:01 lol, that AI i say 2011-11-07T00:27:19 mine out puts a log of what they think 2011-11-07T00:27:28 but that does not matter i loose any ways 2011-11-07T00:28:03 *** kiv has joined #aichallenge 2011-11-07T00:28:35 i wander if any one is using clusters of computers to fight the TCP challange 2011-11-07T00:32:07 for some reason it takes some time after the replay is over before its posted to the server 2011-11-07T00:32:12 a1k0n: can't you just port your lightcycle code to play ants? 2011-11-07T00:32:21 it's the same problem, right? 2011-11-07T00:33:43 *** djr_ has quit IRC (Ping timeout: 245 seconds) 2011-11-07T00:34:21 treeform: I've considered doing GP, or simulated annealing on parameters for my bot 2011-11-07T00:34:59 nickjohnson: yeah me too after i can get at-least some thing working, then tweak parameters 2011-11-07T00:35:06 but my main problem is getting some thing working 2011-11-07T00:35:17 here is my first victory shared with some one http://ants.fluxid.pl/replay.24001 2011-11-07T00:35:22 i am so happy 2011-11-07T00:35:29 but it looks like i just won out of luck though 2011-11-07T00:35:40 *** pairofdice has joined #aichallenge 2011-11-07T00:36:29 *** u_ has quit IRC (Quit: u_) 2011-11-07T00:43:19 any haskellers awake? 2011-11-07T00:43:51 if i know which nearby node i want to goto 2011-11-07T00:43:56 adjacent* 2011-11-07T00:44:11 is there a way to tell which direction it is 2011-11-07T00:44:27 'node'? 2011-11-07T00:44:36 location tuple 2011-11-07T00:44:58 say if im (23, 3) and i want to goto (23, 4) is there a built in function for telling that its north? 2011-11-07T00:45:09 *** cat-alex has joined #aichallenge 2011-11-07T00:45:10 *** u_ has joined #aichallenge 2011-11-07T00:46:29 *** VoidXC has quit IRC (Quit: Page closed) 2011-11-07T00:47:01 *** u_ has quit IRC (Client Quit) 2011-11-07T00:48:08 ecarpenter84: i did not like my starter packadge so i wrote all of my own code, that does not sound like a hard thing to write? 2011-11-07T00:48:30 just subtract them and see? 2011-11-07T00:48:37 (0, -1) means north? 2011-11-07T00:48:46 i just solved it 2011-11-07T00:49:00 it was ants.direction i looked in the starter :P. i thought it was in there 2011-11-07T00:54:51 treeform: hey you won a game 2011-11-07T00:55:07 oh you saw heh 2011-11-07T00:55:26 bmh: i love sentences that start with "can't you just" 2011-11-07T00:55:55 *** kiwi1 has joined #aichallenge 2011-11-07T00:56:01 a1k0n: all I'm saying is $10k in a plain envelope. You know where my desk is. kthx. 2011-11-07T00:56:40 *** Garf has joined #aichallenge 2011-11-07T00:57:38 *** Palmik has joined #aichallenge 2011-11-07T01:03:39 a1k0n: yeah that was exciting, the A bot left some place... 2011-11-07T01:05:42 Hi, is it possible to place the food manually to check my bot first turn ? 2011-11-07T01:05:58 yes. First you must invent the universe. 2011-11-07T01:06:05 [I wish I was joking…] 2011-11-07T01:07:08 cat-alex: create a map file like the tutorial map with pre-placed food and ants then run playgame.py with the --scenario option 2011-11-07T01:07:21 Okay so precalculating the normal distribution values solved that problem. 2011-11-07T01:07:39 http://aichallenge.org/specification.php#Map-Format 2011-11-07T01:07:40 thank 2011-11-07T01:08:08 cat-alex: what i did also is get one of the map replays 2011-11-07T01:08:24 then modify it to have only 1 turn with my ant and food next to it 2011-11-07T01:08:56 and then see what my bot says when i cat the file in 2011-11-07T01:09:00 ahha, --scenario! 2011-11-07T01:09:16 what does a pre-placed ant look like then? 2011-11-07T01:09:34 oh there, hi. thanks. 2011-11-07T01:09:46 :) 2011-11-07T01:09:50 just a 0 0 0 ? 2011-11-07T01:10:13 not in a map file 2011-11-07T01:10:24 you still hvae to use food none or it will ensure symmetry. i see. 2011-11-07T01:10:45 *** dr- has joined #aichallenge 2011-11-07T01:11:10 Haskell -- the only language that gives you meaningful/meaningless choice of arrays. 2011-11-07T01:12:49 and other brain twisters? 2011-11-07T01:13:00 How can the index be out of range... 2011-11-07T01:13:06 How can it be! 2011-11-07T01:13:21 i think that means you didn't specify enough bots on the commandline 2011-11-07T01:14:18 or forgot to put the line escapes \ ? 2011-11-07T01:15:02 If it has the wrong amount of bots, it says so 2011-11-07T01:15:57 *** ecarpenter84 has quit IRC (Quit: Page closed) 2011-11-07T01:23:33 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-11-07T01:25:11 *** TheLinker has quit IRC (Quit: Bye) 2011-11-07T01:41:45 *** alc has quit IRC (Quit: Leaving) 2011-11-07T01:41:49 aichallenge: Janzert epsilon * r4ebbba0 / ants/dist/starter_bots/python/ants.py : 2011-11-07T01:41:49 aichallenge: Merge pull request #355 from jbochi/epsilon 2011-11-07T01:41:49 aichallenge: Python starter bot ants can't see the other side of the toroid - http://git.io/lKQVDw 2011-11-07T01:43:30 (yes they can) 2011-11-07T01:43:43 *** Harpyon has joined #aichallenge 2011-11-07T01:44:59 *** retybok_ has joined #aichallenge 2011-11-07T01:45:31 *** spp has joined #aichallenge 2011-11-07T01:49:37 ahh, I see... someone was trying to get too tricky when they wrote that function :( 2011-11-07T01:51:03 *** kowser has joined #aichallenge 2011-11-07T01:51:41 My ants sure do go for the food, but when I tell them to not be near each other they just look at me funny 2011-11-07T01:52:19 This will drive me crazy 2011-11-07T01:52:31 *** retybok_ has quit IRC (Ping timeout: 248 seconds) 2011-11-07T01:52:34 *** Harpyon has quit IRC (Quit: Textual IRC Client: http://www.textualapp.com/) 2011-11-07T01:52:40 *** kowser has quit IRC (Client Quit) 2011-11-07T01:58:55 *** bmh has quit IRC (Quit: bmh) 2011-11-07T02:00:27 aichallenge: janzert epsilon * rb9bf290 / ants/dist/starter_bots/python/ants.py : Revert unnecessary modulo in python starter vision and add explanatory comment - http://git.io/tzluPw 2011-11-07T02:01:16 *** espes has quit IRC (Remote host closed the connection) 2011-11-07T02:04:36 *** Antimony has quit IRC (Ping timeout: 258 seconds) 2011-11-07T02:05:34 *** Jak_o_Shadows has joined #aichallenge 2011-11-07T02:06:01 *** djr_ has joined #aichallenge 2011-11-07T02:06:30 *** treeform has quit IRC (Remote host closed the connection) 2011-11-07T02:08:36 *** Jak_o_Shadows1 has quit IRC (Ping timeout: 244 seconds) 2011-11-07T02:10:06 *** Antimony has joined #aichallenge 2011-11-07T02:11:06 *** the-mgt has quit IRC (Quit: the-mgt) 2011-11-07T02:11:13 *** master_ninja has joined #aichallenge 2011-11-07T02:11:17 If my costmap's normal level is at 0 and food raises it and the ants go for the food 2011-11-07T02:11:39 And then when I lower the level around my ants, why don't they care about it 2011-11-07T02:15:44 *** g0llum has joined #aichallenge 2011-11-07T02:16:50 *** Antimony has quit IRC (Ping timeout: 258 seconds) 2011-11-07T02:17:05 *** userjjb has joined #aichallenge 2011-11-07T02:17:13 *** delt0r_ has quit IRC (Ping timeout: 258 seconds) 2011-11-07T02:23:16 *** datachomper1 has quit IRC (Quit: Leaving.) 2011-11-07T02:23:43 *** djr_ has quit IRC (Quit: Leaving) 2011-11-07T02:28:54 a1k0n: does your new bot on the server know about enemies? 2011-11-07T02:29:57 *** delt0r_ has joined #aichallenge 2011-11-07T02:34:00 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-11-07T02:42:31 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-07T02:45:13 *** ilinux has joined #aichallenge 2011-11-07T02:47:40 *** GeorgeSebastian has joined #aichallenge 2011-11-07T02:53:25 *** GeorgeSebastian has quit IRC (Ping timeout: 252 seconds) 2011-11-07T02:58:25 *** spp has quit IRC (Quit: Page closed) 2011-11-07T03:05:36 I forget, what's the keyword to designate a new type in C++? 2011-11-07T03:06:27 typedef? 2011-11-07T03:06:38 that only creates a synonym type 2011-11-07T03:06:39 *** GeorgeSebastian has joined #aichallenge 2011-11-07T03:06:54 for instance I want to create two seperate types, row and col 2011-11-07T03:07:27 and use type checking to basically avoid mixing the two up 2011-11-07T03:08:29 use a class or an enum 2011-11-07T03:12:05 *** Areks has quit IRC (Read error: Connection reset by peer) 2011-11-07T03:12:05 *** Areks has joined #aichallenge 2011-11-07T03:13:23 *** epicmonkey has joined #aichallenge 2011-11-07T03:14:49 *** H3racl1tu5 has joined #aichallenge 2011-11-07T03:15:52 userjjb: C++ does not have newtype for arbitrary types. 2011-11-07T03:16:05 An enum would kind of do the trick for integer types. 2011-11-07T03:16:53 *** H3racl1tu5 has quit IRC (Client Quit) 2011-11-07T03:22:01 There's something strange with Sigma calculations. 2011-11-07T03:22:13 No drop from 50 games -> 500 games 2011-11-07T03:22:16 that makes no sense 2011-11-07T03:22:34 It's basically saying those 450 extra games provided no information about strength 2011-11-07T03:23:16 *** ikaros has joined #aichallenge 2011-11-07T03:26:39 on tcp server? 2011-11-07T03:27:00 yes 2011-11-07T03:28:26 I don't think those implement trueskill at all 2011-11-07T03:28:40 they show kill, mu and sigma 2011-11-07T03:28:44 *** tmandry has quit IRC (Ping timeout: 260 seconds) 2011-11-07T03:29:02 but yeah, the calculations there seem to be entirely broken 2011-11-07T03:29:04 Zao: I see what you mean about using a class or an enum. But how would you dynamically populate the enum at execution time? 2011-11-07T03:29:16 dynamically? not in C++ 2011-11-07T03:29:22 C++ has static typing 2011-11-07T03:29:22 heh, thought so 2011-11-07T03:29:43 I guess I want my cake and to eat it to 2011-11-07T03:29:48 too* 2011-11-07T03:31:03 *** olexs has joined #aichallenge 2011-11-07T03:31:19 I don't have a really good reason to want to do so anyways, but these kinds of train of thought are useful for me to fleshing out my understanding of C++ 2011-11-07T03:34:05 *** Blkt has joined #aichallenge 2011-11-07T03:37:17 *** avdg1 has joined #aichallenge 2011-11-07T03:39:17 *** sigh has joined #aichallenge 2011-11-07T03:40:05 *** avdg has quit IRC (Ping timeout: 240 seconds) 2011-11-07T03:44:20 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-07T03:46:55 *** mviel has joined #aichallenge 2011-11-07T03:50:01 Now I understand why the C++ starter package uses the std::vector intsead of dynamic multidimensional arrays 2011-11-07T03:50:13 My god... it's full of pointers! 2011-11-07T03:58:04 *** olexs has quit IRC (Quit: Leaving.) 2011-11-07T03:58:16 *** skunx has joined #aichallenge 2011-11-07T03:58:25 good morning everyone 2011-11-07T03:59:05 Morning 2011-11-07T04:06:04 userjjb: An enum can contain values not corresponding to any of the enumerands. 2011-11-07T04:07:06 userjjb: enum Row { MaxRow = 199; }; // would give you an enum with an underlying integral type that could store at least anything from 0 through 199. 2011-11-07T04:09:32 Row r = (Row)50; Col c = r; // compile-time fault \o/ 2011-11-07T04:11:58 whoo, today's rising star http://aichallenge.org/profile.php?user=589 2011-11-07T04:19:54 after six games it is currently rank 37 and has the second highest mu at 91.26 following xathis at 94.399 2011-11-07T04:20:46 *** ikaros has joined #aichallenge 2011-11-07T04:21:49 *** gcflymoto has quit IRC (Ping timeout: 265 seconds) 2011-11-07T04:30:05 *** epicmonkey has quit IRC (Ping timeout: 240 seconds) 2011-11-07T04:38:22 janzert: C, nice 2011-11-07T04:40:48 *** skunx has quit IRC (Ping timeout: 245 seconds) 2011-11-07T04:47:43 *** liberforce has joined #aichallenge 2011-11-07T04:50:16 *** olexs has joined #aichallenge 2011-11-07T04:52:44 *** replore_ has quit IRC (Remote host closed the connection) 2011-11-07T04:55:18 *** ccc has quit IRC (Quit: Computer has gone to sleep.) 2011-11-07T04:56:24 *** skunx has joined #aichallenge 2011-11-07T05:12:41 *** g0llum has joined #aichallenge 2011-11-07T05:13:04 *** spp has joined #aichallenge 2011-11-07T05:13:51 Its time consuming to get to know the starter packages itself(java). is it that most use the starter template or code everything from scratch? 2011-11-07T05:14:02 Woah, atleast now they avoid enemy ants rather nicely 2011-11-07T05:14:03 *** goffrie has quit IRC (Remote host closed the connection) 2011-11-07T05:16:37 spp: I used the C starter package, and ended up rewriting most of it 2011-11-07T05:18:16 liberforce: Yeah even I feel the same, end up rewriting all. 2011-11-07T05:18:18 spp: Depending on the quality of the starter package, you'd keep it or rewrite it 2011-11-07T05:18:40 the C implementation for aguments parsing was auwful 2011-11-07T05:18:57 yeah I will try to use it the way it is, and if it bothers me a lot I would change 2011-11-07T05:19:03 that's when I understood there was no clever-trick, just poor programming 2011-11-07T05:19:12 I wanted to know like, is it most people reqritimg 2011-11-07T05:19:13 ? 2011-11-07T05:19:14 I wouldn't know any of that so I use the starter package 2011-11-07T05:19:47 Java impl is better in that way 2011-11-07T05:19:50 spp: are you french ? 2011-11-07T05:19:59 liberforce: no\ 2011-11-07T05:20:08 or english on azerty keyboard :) ? 2011-11-07T05:20:23 sorry, lots of typo 2011-11-07T05:20:28 :) 2011-11-07T05:30:27 *** ikaros has quit IRC (Remote host closed the connection) 2011-11-07T05:31:31 *** NoxiaZ^ has joined #aichallenge 2011-11-07T05:40:07 *** marijnfs has joined #aichallenge 2011-11-07T05:53:18 *** AVAVT has quit IRC (Quit: Page closed) 2011-11-07T05:55:01 *** mleise has joined #aichallenge 2011-11-07T05:55:22 hehe, I just made this: http://farm7.static.flickr.com/6050/6322011940_9daa70197e_b.jpg 2011-11-07T05:58:23 Heh, it needs a mouseover! 2011-11-07T05:59:53 When will the challenge end? How will the final rankings be computed? 2011-11-07T06:01:47 The current phase of the contest will end December 18th at 11:59pm EST. At that time submissions will be closed. Shortly thereafter the final tournament will be started. The length of the final tournament has not yet been determined but is expected to last less than one week. Upon completion the contest winner will be announced and all results will be publically available. 2011-11-07T06:02:19 pairofdice: It does have one where I pasted it on Facebook, sort of, since Facebook doesn't seem to like the alt="" part of a href tag I just made it link to my alt text so you'll see it if you mouseover and look below in the status bar 2011-11-07T06:03:19 It says "The only annoying part is that the dimensions have to match the ratio of the first perfect squares..." 2011-11-07T06:03:22 *** kiwi1 has quit IRC (Quit: Page closed) 2011-11-07T06:04:43 *** NoxiaZ^ has quit IRC (Ping timeout: 244 seconds) 2011-11-07T06:08:26 pguillory is rocking 2011-11-07T06:10:29 *** ikaros has joined #aichallenge 2011-11-07T06:11:21 even though he lets his ants go 1vs1 2011-11-07T06:12:30 *** NoxiaZ^ has joined #aichallenge 2011-11-07T06:13:47 pairofdice: thanks for your answer! Is that info on the website somewhere? 2011-11-07T06:14:05 Bottom of the the homepage 2011-11-07T06:14:32 pairofdice: great, thanks. 2011-11-07T06:14:42 the next ai challenge should be stratego! just played against my wife. 2011-11-07T06:14:50 would be cool 2011-11-07T06:15:13 Or Go ;x 2011-11-07T06:15:39 go wouldn't be fun, because there is already a lot of literature on the subject 2011-11-07T06:15:42 meh Go there is lots of Go ai comps out there 2011-11-07T06:15:51 Sure 2011-11-07T06:15:52 a lot of time would simply be spent implementing known strategies 2011-11-07T06:16:04 Yea I guess that would be bad 2011-11-07T06:16:04 deltor_, isn't stratego copyrighted ? 2011-11-07T06:16:27 The game yes... but not the idea... as in like planetwars 2011-11-07T06:16:48 you can only copyright a specific implementation of a idea. 2011-11-07T06:16:55 aka all the monoply clones 2011-11-07T06:17:01 and video game clones 2011-11-07T06:17:10 I don't care what the next challenge is, but it should be something very different 2011-11-07T06:17:12 But yes its something to consider 2011-11-07T06:17:21 ants and planetwars are a bit too similar IMHO 2011-11-07T06:17:25 free and OS is definatly the starndard for here 2011-11-07T06:17:29 even if they are a lot of fun :) 2011-11-07T06:18:19 retybok: what type of game do you would want 2011-11-07T06:19:17 would you want 2011-11-07T06:19:47 I don't know, just something different :) 2011-11-07T06:20:23 Well the basic requirement is that it it should interesting/entertaining to watch 2011-11-07T06:20:23 One thing I really like about ants is that it's multiplayer 2011-11-07T06:20:34 and a basic bot should be possible 2011-11-07T06:20:40 it changes quite a bit from the previous two-palyer games 2011-11-07T06:22:14 hidden information was considered important 2011-11-07T06:22:26 rather than the complete information of planet wars 2011-11-07T06:25:23 I think adding an element of luck would be interesting as well 2011-11-07T06:25:33 dice rolls for example 2011-11-07T06:25:58 I agree fog of war is pretty cool 2011-11-07T06:27:02 How about something with less agents but more things they can do, like mechs 2011-11-07T06:27:46 i wouldn't mind xcom/jagged alliance/frozen synapse turnbased tactics with fog of war 2011-11-07T06:32:54 *** Apophis_ has joined #aichallenge 2011-11-07T06:33:16 *** ikaros has quit IRC (*.net *.split) 2011-11-07T06:33:18 *** nplus has quit IRC (*.net *.split) 2011-11-07T06:33:19 *** Apophis has quit IRC (*.net *.split) 2011-11-07T06:33:20 *** og01 has quit IRC (*.net *.split) 2011-11-07T06:33:20 *** timos has quit IRC (*.net *.split) 2011-11-07T06:33:20 *** chris__0076 has quit IRC (*.net *.split) 2011-11-07T06:33:20 *** Alexer has quit IRC (*.net *.split) 2011-11-07T06:33:21 *** ping-- has quit IRC (*.net *.split) 2011-11-07T06:33:22 *** antonh has quit IRC (*.net *.split) 2011-11-07T06:33:22 *** Zao has quit IRC (*.net *.split) 2011-11-07T06:33:22 *** f4hy has quit IRC (*.net *.split) 2011-11-07T06:33:22 *** Lerc has quit IRC (*.net *.split) 2011-11-07T06:33:22 *** Acy- has quit IRC (*.net *.split) 2011-11-07T06:33:23 *** wonklab has quit IRC (*.net *.split) 2011-11-07T06:33:25 *** GeorgeSebastian has quit IRC (Ping timeout: 265 seconds) 2011-11-07T06:33:27 *** ping-- has joined #aichallenge 2011-11-07T06:34:32 *** nplus has joined #aichallenge 2011-11-07T06:35:45 Woooo, fine tuning! 2011-11-07T06:35:57 Finally something other than broke or not broke 2011-11-07T06:37:10 *** Acy- has joined #aichallenge 2011-11-07T06:37:23 *** Alexer has joined #aichallenge 2011-11-07T06:38:01 *** og01 has joined #aichallenge 2011-11-07T06:38:03 *** timos has joined #aichallenge 2011-11-07T06:39:17 Anyone have working solution for live debugging with C#? :) 2011-11-07T06:40:40 *** Zao has joined #aichallenge 2011-11-07T06:41:15 *** Lerc has joined #aichallenge 2011-11-07T06:42:59 *** tncardoso has joined #aichallenge 2011-11-07T06:47:30 *** Jak_o_Shadows has quit IRC (Remote host closed the connection) 2011-11-07T06:49:01 *** ikaros has joined #aichallenge 2011-11-07T06:49:26 *** wonklab has joined #aichallenge 2011-11-07T06:49:44 *** mleise has quit IRC (Quit: Leaving.) 2011-11-07T06:49:49 *** f4hy has joined #aichallenge 2011-11-07T06:49:51 *** chris__0076 has joined #aichallenge 2011-11-07T06:50:03 *** GeorgeSebastian has joined #aichallenge 2011-11-07T06:51:30 *** olexs has quit IRC (Quit: Leaving.) 2011-11-07T06:52:01 NoxiaZ^, http://aichallenge.org/forums/viewtopic.php?f=25&t=1778 2011-11-07T06:55:28 Garf: yea that would be cool... xcom for the win. That was such a brutal game... works under dosbox btw 2011-11-07T06:55:51 NoxiaZ^: Error stream logging? 2011-11-07T06:58:00 delt0r, yes 2011-11-07T06:58:37 *** GeorgeSebastian has quit IRC (Ping timeout: 258 seconds) 2011-11-07T06:59:48 delt0r, i would like to step into the code for debugging, so im not adding any lines to debug, to be sure the hold down the time for executing my code :) 2011-11-07T07:02:20 Fluxid: ping 2011-11-07T07:02:34 ah 2011-11-07T07:02:36 fuck 2011-11-07T07:02:44 your server died again I think 2011-11-07T07:03:25 FUCK 2011-11-07T07:03:30 'lsof -i' 2011-11-07T07:03:32 :P 2011-11-07T07:03:40 http://fluxid.pl/private/2011-11-07-130312_541x1237_scrot.png I THINK I KNOW WHY FD'S ARE LEAKED 2011-11-07T07:03:56 HOLY SHIT THIS CODE SUCKS 2011-11-07T07:04:14 lol'd 2011-11-07T07:04:45 fd are not closed because threads keep running 2011-11-07T07:04:49 oh boy 2011-11-07T07:05:18 i couldn't count threads 2011-11-07T07:05:21 killed, restarted 2011-11-07T07:05:23 -_-' 2011-11-07T07:05:35 *** skunx has quit IRC (Ping timeout: 248 seconds) 2011-11-07T07:10:59 *** boegel has joined #aichallenge 2011-11-07T07:11:33 *** GeorgeSebastian has joined #aichallenge 2011-11-07T07:11:53 Fluxid: ah, it never even tries to .join() the thread objects xD 2011-11-07T07:12:45 well... nvm 2011-11-07T07:12:47 weird 2011-11-07T07:15:52 I don't understand the code, where does it limit the thread pool? 2011-11-07T07:16:16 it seems like someone could just connect with 10000 tcp clients and it would allow that 2011-11-07T07:16:54 *** skunx has joined #aichallenge 2011-11-07T07:18:10 *** GeorgeSebastian has quit IRC (Ping timeout: 258 seconds) 2011-11-07T07:21:40 *** hilbert- has joined #aichallenge 2011-11-07T07:22:36 Hey, Rule: Use of multiple processes or OS threads are prohibited. - Im not sure that OS threads means, but to be sure, its not allowed to use any form for threads ? 2011-11-07T07:22:49 that = what* 2011-11-07T07:23:00 what language are you using? 2011-11-07T07:23:03 C# 2011-11-07T07:23:27 so don't use threads, they'll use OS threads 2011-11-07T07:23:29 *** QuirionPT has joined #aichallenge 2011-11-07T07:23:39 oki 2011-11-07T07:24:38 Go lets you use lightweight threads (goroutines) which are multiplexed over N os threads (1 by default) and ofc that's what Erlang is all about 2011-11-07T07:25:30 I think using multiple OS threads (pthreads) would steal processing time from the other bots with the way it's set up 2011-11-07T07:25:40 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-07T07:25:42 *** lorill has joined #aichallenge 2011-11-07T07:26:11 ahh ye, that makes sens 2011-11-07T07:27:51 thanks alot thestinger :) 2011-11-07T07:29:57 *** GeorgeSebastian has joined #aichallenge 2011-11-07T07:36:07 lol, my ants are so hopeless against A http://ants.fluxid.pl/replay.24382 2011-11-07T07:36:10 *** GeorgeSebastian has quit IRC (Ping timeout: 252 seconds) 2011-11-07T07:36:19 it takes like a hundred of them to break through his defense of a dozen ants 2011-11-07T07:36:47 yeah A has pretty good micro 2011-11-07T07:39:02 what color are you in this replay ? 2011-11-07T07:39:26 strcat (both of them :P) 2011-11-07T07:40:01 haha 2011-11-07T07:40:13 I see. You're the one that's always rushing me, then :) 2011-11-07T07:40:19 well, strcat_next is basically my current code and strcat_forage is a branch I'm working on 2011-11-07T07:40:28 "strcat" is old crappy code that sucks :P 2011-11-07T07:40:35 do you try to guess where other hills are, or dou you often get lucky when scouting ? 2011-11-07T07:40:42 atm I just get lucky 2011-11-07T07:41:05 I just have a few heuristics to prevent my ants from suiciding, so they are very aggressive 2011-11-07T07:41:09 damned. I would have hoped you could tell me how you guess them :( 2011-11-07T07:41:16 I'd like to keep them pretty aggressive when I get real combat done 2011-11-07T07:41:21 well, symmetry 2011-11-07T07:41:23 look at the maps 2011-11-07T07:41:31 the hills are all laid out the same way 2011-11-07T07:41:36 yeah, but i don't know how to exploit it 2011-11-07T07:41:51 i mean, I can't simply reverse the coords 2011-11-07T07:41:54 so if you find one hill, I guess you can use the distances to other hills you know about and plot a bunch of possible hills on a map 2011-11-07T07:42:00 I dunno xD 2011-11-07T07:42:28 *** the-mgt has joined #aichallenge 2011-11-07T07:42:28 i think this will be my next evolution, but I don't know how to tackle it yet 2011-11-07T07:43:02 yeah, you could take the shortest vertical/horizontal distance to another hill and then stick a bunch of possible hills on a map 2011-11-07T07:43:37 I think that would be better than my exploration towards any unexplored areas (without prioritizing) that I do atm 2011-11-07T07:43:42 *** kilae has joined #aichallenge 2011-11-07T07:44:26 anyway, when I start improving combat I'd like to keep them pretty aggressive 2011-11-07T07:44:34 atm they're just aggressive because they don't know any better :P 2011-11-07T07:44:54 *** grwip has joined #aichallenge 2011-11-07T07:45:42 lol :D 2011-11-07T07:49:49 *** GeorgeSebastian has joined #aichallenge 2011-11-07T07:53:20 What algorithm are you guys using for discover the terrain? 2011-11-07T07:53:39 several different ones 2011-11-07T07:53:47 1. I assign some scout points on the map 2011-11-07T07:53:53 equidistants, but randomized 2011-11-07T07:54:05 and while they're unexplored, a* to them 2011-11-07T07:54:48 2. for unmoved ants after important things, move to the cell that wasn't walked on for the longest time 2011-11-07T07:55:12 works ok, but not perfect in maze 2011-11-07T07:55:21 ahh but ye its a way to do it 2011-11-07T07:56:01 but i thought at A* needed to "know" the terrain to move? Like D* would do better when you dont know the terrain yet? 2011-11-07T07:56:29 *** ikaros has joined #aichallenge 2011-11-07T07:56:37 i consider all unknown cell as land 2011-11-07T07:56:52 And again i was like thinking, where to start moving? - But thats pretty smart to put up some scout points, and then grap the foot on the way 2011-11-07T07:56:52 after all, the aim is to discover it, so that doesn't matter too much 2011-11-07T07:57:01 ahh ye ofc 2011-11-07T07:57:04 *** GeorgeSebastian has quit IRC (Ping timeout: 252 seconds) 2011-11-07T07:57:15 *** hilbert- has quit IRC (Quit: hilbert-) 2011-11-07T07:57:25 My next problem is, it dosnt seems like there are any ID's on the ants? 2011-11-07T07:57:45 all ants are born equal :) 2011-11-07T07:57:58 do you really care which ant does what ? its position should be enough 2011-11-07T07:58:10 So how do you keep track on what ants you are using to discover, find food, protect your hill and what you wanna start killing with? 2011-11-07T07:58:12 or if you care, you can track them by using their expected position 2011-11-07T07:58:24 i don't care 2011-11-07T07:58:32 i always use the closest available one 2011-11-07T07:58:41 ahh oki 2011-11-07T07:59:08 I had a previous version that did the tracking, but it doesn't really help 2011-11-07T08:00:54 ahh oki, but keep tracking for the next position will also cost much more time then just use the closest one for the job . But ideal will ofc be that they had an id :D 2011-11-07T08:01:26 And ofc for the defence, you can just use the new spawned that move the old one :P - Seems as an idea :D like swapping places ! =) 2011-11-07T08:03:19 *** antonh has joined #aichallenge 2011-11-07T08:05:30 *** TheLinker has joined #aichallenge 2011-11-07T08:10:44 *** GeorgeSebastian has joined #aichallenge 2011-11-07T08:15:58 *** bergmark has quit IRC (Ping timeout: 255 seconds) 2011-11-07T08:17:08 NoxiaZ^: I have a 2d array which holds distances to unseen locations for every tile 2011-11-07T08:17:58 so each turn I just reset it and refill it with the new distances (using a BFS from each unseen location, but most stop right away since they're just surrounded by more unseen locations) 2011-11-07T08:18:10 *** GeorgeSebastian has quit IRC (Ping timeout: 260 seconds) 2011-11-07T08:18:35 I'm going to end up weighting it by starting unseen location at 5 distance, so I can start potential hills location at a lower distance 2011-11-07T08:18:50 s/hills location/hill locations/ 2011-11-07T08:19:26 well, 5 is just an arbitrary number - I'll figure out a good weight 2011-11-07T08:20:30 ahh oki :) smart 2011-11-07T08:21:58 *** geekchimp has joined #aichallenge 2011-11-07T08:29:09 *** Nikozc has joined #aichallenge 2011-11-07T08:30:57 *** GeorgeSebastian has joined #aichallenge 2011-11-07T08:32:08 *** smiley1983 has joined #aichallenge 2011-11-07T08:33:52 *** loll has joined #aichallenge 2011-11-07T08:34:07 *** mviel_ has joined #aichallenge 2011-11-07T08:34:08 *** loll has quit IRC (Client Quit) 2011-11-07T08:37:08 *** GeorgeSebastian has quit IRC (Ping timeout: 258 seconds) 2011-11-07T08:37:51 *** mviel has quit IRC (Ping timeout: 260 seconds) 2011-11-07T08:39:40 *** spp has quit IRC (Quit: Page closed) 2011-11-07T08:44:33 *** the-mgt has quit IRC (Ping timeout: 245 seconds) 2011-11-07T08:45:09 *** the-mgt has joined #aichallenge 2011-11-07T08:45:29 *** u_ has joined #aichallenge 2011-11-07T08:46:12 *** geekchimp has quit IRC (Ping timeout: 265 seconds) 2011-11-07T08:47:34 *** lorill has quit IRC (Quit: Page closed) 2011-11-07T08:50:46 *** grandiose has joined #aichallenge 2011-11-07T08:50:50 *** Antimony has joined #aichallenge 2011-11-07T08:51:25 *** Areks has quit IRC (Ping timeout: 240 seconds) 2011-11-07T08:52:26 *** sigh has quit IRC (Remote host closed the connection) 2011-11-07T08:54:14 *** the-mgt has quit IRC (Quit: the-mgt) 2011-11-07T08:54:54 *** Mooloo has quit IRC (Ping timeout: 240 seconds) 2011-11-07T08:56:27 *** tncardoso has quit IRC (Quit: bye) 2011-11-07T08:59:12 *** yoden has joined #aichallenge 2011-11-07T08:59:15 *** bergmark has joined #aichallenge 2011-11-07T09:02:43 *** amstan_ has joined #aichallenge 2011-11-07T09:02:43 *** ChanServ sets mode: +o amstan_ 2011-11-07T09:02:51 *** pedrosorio has joined #aichallenge 2011-11-07T09:07:37 *** Mooloo has joined #aichallenge 2011-11-07T09:09:58 *** grandiose has quit IRC (Quit: Leaving.) 2011-11-07T09:10:18 *** smiley1983 has quit IRC (Ping timeout: 256 seconds) 2011-11-07T09:11:02 *** f4hy has quit IRC (*.net *.split) 2011-11-07T09:11:04 *** ping-- has quit IRC (*.net *.split) 2011-11-07T09:11:06 *** acieroid has quit IRC (*.net *.split) 2011-11-07T09:11:06 *** netantho has quit IRC (*.net *.split) 2011-11-07T09:11:06 *** nikis has quit IRC (*.net *.split) 2011-11-07T09:11:07 *** chiller- has quit IRC (*.net *.split) 2011-11-07T09:11:08 *** pgpaskar_ has quit IRC (*.net *.split) 2011-11-07T09:11:09 *** amstan_ has quit IRC (Write error: Broken pipe) 2011-11-07T09:11:09 *** ybit has quit IRC (Read error: Operation timed out) 2011-11-07T09:11:27 *** staalebk has joined #aichallenge 2011-11-07T09:11:37 *** spacebat has quit IRC (Read error: Operation timed out) 2011-11-07T09:12:18 *** janzert has quit IRC (Read error: Connection reset by peer) 2011-11-07T09:12:19 *** Sir_Ragnarok_ has quit IRC (Read error: Connection reset by peer) 2011-11-07T09:12:19 *** pedrosorio has quit IRC (Ping timeout: 265 seconds) 2011-11-07T09:12:19 *** Nikozc has quit IRC (Ping timeout: 265 seconds) 2011-11-07T09:12:25 *** caution has joined #aichallenge 2011-11-07T09:12:29 *** user_jjb has joined #aichallenge 2011-11-07T09:12:33 *** Sir_Ragnarok has joined #aichallenge 2011-11-07T09:12:50 *** amstan has joined #aichallenge 2011-11-07T09:12:51 *** ChanServ sets mode: +o amstan 2011-11-07T09:12:54 *** yoden has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:12:54 *** g0llum has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:12:54 *** Raimondi has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:12:54 *** Zepp has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:12:59 *** Zepp has joined #aichallenge 2011-11-07T09:13:15 t 2011-11-07T09:13:19 *** luizribeiro has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:13:19 *** magiik has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:13:29 *** user_jjb has quit IRC (Client Quit) 2011-11-07T09:13:36 *** MikeyG has quit IRC (Ping timeout: 248 seconds) 2011-11-07T09:13:42 *** TheLinker has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:13:43 *** marijnfs has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:13:43 *** cyphase has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:13:44 *** jmcarthur has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:14:01 *** moondust has quit IRC (Ping timeout: 255 seconds) 2011-11-07T09:14:03 *** Bluedgis has quit IRC (Ping timeout: 252 seconds) 2011-11-07T09:14:06 *** AlliedEnvy has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:14:06 *** Israfel has quit IRC (Ping timeout: 240 seconds) 2011-11-07T09:14:10 *** Extrarius has quit IRC (Ping timeout: 260 seconds) 2011-11-07T09:14:19 *** spp has joined #aichallenge 2011-11-07T09:14:22 *** TheLinker has joined #aichallenge 2011-11-07T09:14:23 almost top 800, hope I can hit top 500 when I add pathfinding 2011-11-07T09:14:36 *** jstemmer has quit IRC (Ping timeout: 260 seconds) 2011-11-07T09:14:36 *** marijnfs has joined #aichallenge 2011-11-07T09:14:42 *** Ionic_Groove has quit IRC (Ping timeout: 258 seconds) 2011-11-07T09:14:51 *** mayanks43 has quit IRC (Remote host closed the connection) 2011-11-07T09:14:51 *** Bluedgis has joined #aichallenge 2011-11-07T09:15:00 *** mayanks43 has joined #aichallenge 2011-11-07T09:15:04 *** janzert has joined #aichallenge 2011-11-07T09:15:05 *** spacebat has joined #aichallenge 2011-11-07T09:15:05 *** magiik has joined #aichallenge 2011-11-07T09:15:10 *** galdor has quit IRC (Quit: ZNC - http://znc.sourceforge.net) 2011-11-07T09:15:16 *** g0llum has joined #aichallenge 2011-11-07T09:15:20 *** metonymy has joined #aichallenge 2011-11-07T09:15:21 *** moondust has joined #aichallenge 2011-11-07T09:15:41 *** MikeyG has joined #aichallenge 2011-11-07T09:15:42 *** cat-alex has quit IRC (Ping timeout: 265 seconds) 2011-11-07T09:15:46 *** krishna has quit IRC (Ping timeout: 260 seconds) 2011-11-07T09:16:13 *** Darrow has quit IRC (Ping timeout: 245 seconds) 2011-11-07T09:16:14 *** userjjb has quit IRC (Ping timeout: 258 seconds) 2011-11-07T09:16:22 *** ybit has joined #aichallenge 2011-11-07T09:16:33 *** Darrow has joined #aichallenge 2011-11-07T09:17:03 *** a1k0n has quit IRC (Ping timeout: 245 seconds) 2011-11-07T09:17:08 *** galdor has joined #aichallenge 2011-11-07T09:17:18 *** bluemanshoe has joined #aichallenge 2011-11-07T09:17:25 *** master_ninja has left #aichallenge 2011-11-07T09:17:30 *** f4hy has joined #aichallenge 2011-11-07T09:17:30 *** ping-- has joined #aichallenge 2011-11-07T09:17:30 *** acieroid has joined #aichallenge 2011-11-07T09:17:30 *** netantho has joined #aichallenge 2011-11-07T09:17:30 *** nikis has joined #aichallenge 2011-11-07T09:17:30 *** pgpaskar_ has joined #aichallenge 2011-11-07T09:17:37 *** Israfel has joined #aichallenge 2011-11-07T09:17:52 *** jmcarthur has joined #aichallenge 2011-11-07T09:17:53 *** Ionic_Groove has joined #aichallenge 2011-11-07T09:17:57 *** AlliedEnvy has joined #aichallenge 2011-11-07T09:18:21 *** krishna has joined #aichallenge 2011-11-07T09:18:29 *** cyphase has joined #aichallenge 2011-11-07T09:18:48 *** a1k0n has joined #aichallenge 2011-11-07T09:18:53 *** luizribeiro has joined #aichallenge 2011-11-07T09:20:41 *** the-mgt has joined #aichallenge 2011-11-07T09:22:05 *** jstemmer has joined #aichallenge 2011-11-07T09:22:05 *** bol_ has joined #aichallenge 2011-11-07T09:22:06 *** Extrarius has joined #aichallenge 2011-11-07T09:22:24 *** Joszs has joined #aichallenge 2011-11-07T09:22:49 *** yoden has joined #aichallenge 2011-11-07T09:23:26 *** marijnfs has quit IRC (Quit: leaving) 2011-11-07T09:24:05 *** galdor has quit IRC (Ping timeout: 260 seconds) 2011-11-07T09:24:15 Hi, I am having a problem. I downloaded the tools from the site.... python installed. And when I start a game like said in the tutorial, a console appears and disappears., does anyone know what I'm doing wrong? 2011-11-07T09:24:23 *** kire has quit IRC (Ping timeout: 260 seconds) 2011-11-07T09:24:29 *** kire has joined #aichallenge 2011-11-07T09:24:37 *** galdor has joined #aichallenge 2011-11-07T09:25:08 Hi, I am having a problem. I downloaded the tools from the site.... python installed. And when I start a game like said in the tutorial, a console appears and disappears., does anyone know what I'm doing wrong? 2011-11-07T09:25:11 Run it from the command line instead, then you'll see possible error messages 2011-11-07T09:25:28 Ok 2011-11-07T09:25:52 Uhmm... how? 2011-11-07T09:26:54 *** olexs has joined #aichallenge 2011-11-07T09:27:11 *** smiley1983 has joined #aichallenge 2011-11-07T09:27:21 janzert: http://aichallenge.org/visualizer.php?game=69486 that game shouldn't have ended yet 2011-11-07T09:27:53 *** cafaro has quit IRC (Ping timeout: 245 seconds) 2011-11-07T09:29:54 Joszs: this is a programming competition 2011-11-07T09:30:04 *** cafaro has joined #aichallenge 2011-11-07T09:30:04 *** cafaro has joined #aichallenge 2011-11-07T09:30:53 you need to know some basics of how to write and run a program 2011-11-07T09:31:06 *** luizribeiro has quit IRC (Ping timeout: 276 seconds) 2011-11-07T09:31:56 The programm is just not running... 2011-11-07T09:33:11 *** smiley1983 has quit IRC (Ping timeout: 258 seconds) 2011-11-07T09:33:31 *** smiley1983 has joined #aichallenge 2011-11-07T09:33:42 When I run play_one_game.cmd it runs and dissappears in 1 sec. 2011-11-07T09:34:19 *** Raimondi has joined #aichallenge 2011-11-07T09:34:27 *** luizribeiro has joined #aichallenge 2011-11-07T09:34:42 The tutorial clearly explains how to create a test script 2011-11-07T09:34:46 add "pause" on a new line at the end of the file 2011-11-07T09:35:28 The official game servers uses ruby1.9 right? I was shocked to see the performance difference of my BFS between 1.8 and 1.9, it is more than twice as fast in my case. 2011-11-07T09:37:59 thanks.... python wasn't recognized 2011-11-07T09:38:06 *** foRei has joined #aichallenge 2011-11-07T09:40:20 *** Antimony has quit IRC (Ping timeout: 256 seconds) 2011-11-07T09:40:47 *** bluemanshoe has quit IRC (Quit: Leaving) 2011-11-07T09:41:54 *** metonymy has quit IRC (Ping timeout: 265 seconds) 2011-11-07T09:42:12 *** Joszs has quit IRC (Quit: Page closed) 2011-11-07T09:42:45 can someone tell me how I can print debug information? I use python2.7 and win7! thx 2011-11-07T09:44:40 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-07T09:46:14 *** BYDESIGN is now known as modafinil 2011-11-07T09:47:01 I found it! 2011-11-07T09:47:01 kilae: The client uses sysout, but presumably if you print to syserr it'll be visible on your command line. 2011-11-07T09:47:28 Not for me atleast 2011-11-07T09:48:17 Maybe it needs some particular flag(s) 2011-11-07T09:50:58 *** Scooby has joined #aichallenge 2011-11-07T09:51:02 *** Scooby is now known as Sqqby 2011-11-07T09:53:35 *** amstan has quit IRC (Ping timeout: 248 seconds) 2011-11-07T09:56:57 *** the-mgt has quit IRC (Ping timeout: 258 seconds) 2011-11-07T09:57:28 Hi all, I did not find an answer to this in the FAQ: What will be done with all the info gathered by the challenge? What rights does the challenger grant by taking part? Will the winning algorithms be published and/or discussed publicly in any way? Where can I find more info on such issues? 2011-11-07T09:57:30 *** the-mgt_ has joined #aichallenge 2011-11-07T09:57:30 *** smiley1983 has quit IRC (Ping timeout: 252 seconds) 2011-11-07T09:57:30 run with the -e argument and you an print to stderr 2011-11-07T09:57:31 *** the-mgt_ is now known as the-mgt 2011-11-07T10:00:51 Brain is melting trying to figure out combat 2011-11-07T10:02:41 pairofdice, Run some simulations then. 2011-11-07T10:04:28 buy a bigger brain, or try liquid helium cooling 2011-11-07T10:04:41 lol 2011-11-07T10:04:47 Where to buy a bigger brain? 2011-11-07T10:04:51 Would love to know ! 2011-11-07T10:05:04 at body shop :p 2011-11-07T10:05:19 Would love to upgrade my brain with more memory! 2011-11-07T10:05:25 :D 2011-11-07T10:05:59 I want a math co-processor 2011-11-07T10:06:55 *** Antimony has joined #aichallenge 2011-11-07T10:07:06 Waiting for upgrades to my implants 2011-11-07T10:17:03 *** delt0r_ has quit IRC (Ping timeout: 245 seconds) 2011-11-07T10:18:25 on small maps my foraging code actually works well http://ants.fluxid.pl/replay.24588 :) 2011-11-07T10:18:58 *** boegel has quit IRC (Remote host closed the connection) 2011-11-07T10:21:45 Can you explain what do you do when you mean foraging. It attacked the enemy really well 2011-11-07T10:22:10 going back to locations I've explored to look for food 2011-11-07T10:22:31 *** Harpyon has joined #aichallenge 2011-11-07T10:22:43 do u keep a separate set of ants to look for food always. it looks like that 2011-11-07T10:22:59 nope, I don't track my ants at all 2011-11-07T10:23:30 only the closest ant will go to an unseen area that's expired (hasn't been seen for N turns) 2011-11-07T10:23:34 *** mcstar has joined #aichallenge 2011-11-07T10:23:48 *** george has joined #aichallenge 2011-11-07T10:24:01 ok so you explore it again and again by doing so? 2011-11-07T10:24:24 ejls: hi 2011-11-07T10:24:32 if I haven't seen an area for N turns, I'll make it a target for re-exploration 2011-11-07T10:24:38 I do it by areas, not tiles 2011-11-07T10:24:39 while exploring, do you also send ants as some group so that they dont get killed 2011-11-07T10:24:45 ok got it 2011-11-07T10:24:54 while exploring they go as group, not while re-exploring 2011-11-07T10:24:59 as groups* 2011-11-07T10:25:06 ook 2011-11-07T10:26:21 I need like a 2:1 or 3:1 ratio of ants to beat BenJackson2 because he seems to have real combat stuff 2011-11-07T10:26:32 *** Raimondi has quit IRC (Read error: Connection reset by peer) 2011-11-07T10:27:27 Yeah I noticed that. Your exploration is good enough to create enough ants in the recent run 2011-11-07T10:28:22 foraging is now 140 lines of code :P 2011-11-07T10:28:25 Do you also maintain any state, when sending army to raze enemy? 2011-11-07T10:28:34 *** tmandry has joined #aichallenge 2011-11-07T10:28:34 *** tmandry has joined #aichallenge 2011-11-07T10:28:51 140 lines, thats little on the higher side 2011-11-07T10:28:55 *** smiley1983 has joined #aichallenge 2011-11-07T10:29:33 i just recently looked up "forage", couple of days ago 2011-11-07T10:29:44 since ive seen you use it twice, was worth it 2011-11-07T10:29:56 *** delt0r_ has joined #aichallenge 2011-11-07T10:30:19 mcstar: can you point me to "foraging" - if any interesting 2011-11-07T10:30:46 hm? i didnt know what that word meant before 2011-11-07T10:30:53 so do I 2011-11-07T10:31:23 forage is something you feed animals 2011-11-07T10:31:30 with 2011-11-07T10:31:58 i think you wanted to say: neither do i 2011-11-07T10:32:09 thestinger: any thoughts? 2011-11-07T10:32:27 searching for and gathering food basically 2011-11-07T10:32:30 https://secure.wikimedia.org/wikipedia/en/wiki/Foraging 2011-11-07T10:33:12 *** amstan has joined #aichallenge 2011-11-07T10:33:12 *** ChanServ sets mode: +o amstan 2011-11-07T10:34:26 *** smiley1983 has quit IRC (Ping timeout: 258 seconds) 2011-11-07T10:35:38 it's a trade off between a higher long term supply of ants and a higher short term number of attackers or explorers 2011-11-07T10:35:54 *** dwins has joined #aichallenge 2011-11-07T10:36:29 lol modafinil 2011-11-07T10:37:18 *** cyphase has quit IRC (Ping timeout: 240 seconds) 2011-11-07T10:38:13 *** cyphase has joined #aichallenge 2011-11-07T10:39:42 *** Akranis has joined #aichallenge 2011-11-07T10:40:52 *** spp has quit IRC (Quit: Page closed) 2011-11-07T10:41:22 *** master_ninja has joined #aichallenge 2011-11-07T10:42:05 *** Sqqby has quit IRC (Quit: Page closed) 2011-11-07T10:42:39 *** amstan has quit IRC (Ping timeout: 252 seconds) 2011-11-07T10:46:34 *** food has joined #aichallenge 2011-11-07T10:46:46 how to generate a replay from a 0.replay file? 2011-11-07T10:50:48 *** ChristianK has joined #aichallenge 2011-11-07T10:50:58 anybody know how to generate a replay from a 0.replay file? 2011-11-07T10:51:25 food: what do you want to achieve 2011-11-07T10:51:40 make a replay html to watch it with 2011-11-07T10:51:45 food: do you want to see your ants moving on the screen? 2011-11-07T10:51:48 rename it to something.html, put it on a web server with /js and /data folders with the appropriate files in them 2011-11-07T10:52:27 food: if you want it locally, compile visualizer and run java -jar visualizer.jar 0.replay 2011-11-07T10:52:56 Minthos: webserver works Fluxid: thanks to both of you! 2011-11-07T10:54:00 *** food has quit IRC (Quit: Page closed) 2011-11-07T10:54:35 *** jef_ has joined #aichallenge 2011-11-07T11:01:33 *** kilae has quit IRC (Quit: ChatZilla 0.9.87 [Firefox 7.0.1/20110928134238]) 2011-11-07T11:01:33 yay for juniper! 2011-11-07T11:03:05 *** Antimony_ has joined #aichallenge 2011-11-07T11:03:33 *** NotABug has joined #aichallenge 2011-11-07T11:04:18 what did they do now? 2011-11-07T11:04:38 *** Antimony has quit IRC (Ping timeout: 276 seconds) 2011-11-07T11:04:45 *** Antimony_ is now known as Antimony 2011-11-07T11:05:42 *** tmandry has quit IRC (Ping timeout: 240 seconds) 2011-11-07T11:07:53 *** NoxiaZ^ has quit IRC (Ping timeout: 245 seconds) 2011-11-07T11:11:01 *** olexs1 has joined #aichallenge 2011-11-07T11:11:43 *** olexs has quit IRC (Ping timeout: 260 seconds) 2011-11-07T11:13:02 *** Blkt has quit IRC (Quit: gogogo) 2011-11-07T11:15:08 mcstar: :D 2011-11-07T11:15:30 do you have prescription? 2011-11-07T11:15:56 *** Larose has joined #aichallenge 2011-11-07T11:16:16 nah just the internet 2011-11-07T11:16:27 :) 2011-11-07T11:20:30 *** GarfTop has joined #aichallenge 2011-11-07T11:24:47 *** Garf has quit IRC (Ping timeout: 276 seconds) 2011-11-07T11:24:58 *** GarfTop has quit IRC (Ping timeout: 244 seconds) 2011-11-07T11:24:59 *** spp has joined #aichallenge 2011-11-07T11:25:37 Minthos: http://twitter.com/#!/pymaunier/status/133559015425843200 2011-11-07T11:27:52 *** QuirionPT has quit IRC (Quit: Leaving) 2011-11-07T11:28:36 *** bol_ has quit IRC (Ping timeout: 265 seconds) 2011-11-07T11:30:43 *** Pepelac has joined #aichallenge 2011-11-07T11:30:55 hi evrebody 2011-11-07T11:35:53 *** JorgeB has joined #aichallenge 2011-11-07T11:36:43 ? 2011-11-07T11:37:12 *** jef_ has left #aichallenge 2011-11-07T11:37:26 Hi, someone can explain how run my own c++ code with bots on my pc& 2011-11-07T11:37:34 ? 2011-11-07T11:39:17 Each turn, do they notify all food tiles or the tiles in area we have explored? 2011-11-07T11:40:04 what is view radius? 2011-11-07T11:40:49 *** kilae has joined #aichallenge 2011-11-07T11:40:50 @settings 2011-11-07T11:40:51 avdg1: settings could be http://aichallenge.org/game_settings.php. 2011-11-07T11:41:07 ^ for view radius 2011-11-07T11:41:57 Sorry, what does view radius mean? 2011-11-07T11:43:40 Ok got it. thats the fog of war and all the information we get is within this fog of war 2011-11-07T11:45:36 *** RasmusL has joined #aichallenge 2011-11-07T11:45:38 ooops, i inadvertantly wrote a snake-bot 2011-11-07T11:45:49 :-) 2011-11-07T11:48:00 ) 2011-11-07T11:48:54 i have a little question :P 2011-11-07T11:48:55 How do you get your ants to spread out? 2011-11-07T11:49:04 isbric: nice, so that's what the netsplits earlier today were about :) 2011-11-07T11:49:20 in python, i have a global variable in one module that i use in another module 2011-11-07T11:49:32 but its value is constant, it doesn't change although i change it in the first module? 2011-11-07T11:50:04 perhaps you need to use the global keyword 2011-11-07T11:50:14 i use global when i change it 2011-11-07T11:50:25 if the global is called 'x' and you're in a function, x = 5 will make a local called x, shadowing the global 2011-11-07T11:50:27 etc. 2011-11-07T11:50:47 yeah i do global x then x=5 2011-11-07T11:50:54 but in the other module it's still 0 2011-11-07T11:51:01 well how do you import the module? 2011-11-07T11:51:07 import module :) 2011-11-07T11:51:16 and of course module.x 2011-11-07T11:51:23 so why do you use global x? 2011-11-07T11:51:43 hmm? 2011-11-07T11:52:00 paste your code 2011-11-07T11:52:15 sure 2011-11-07T11:52:32 *** ikaros has joined #aichallenge 2011-11-07T11:54:16 RasmusL: globals are immutable, when you change it, they become local 2011-11-07T11:54:22 you need to wrap it in a list 2011-11-07T11:54:35 mygloba=[val] 2011-11-07T11:54:41 hmm 2011-11-07T11:54:46 now you can chage myglobal[0] 2011-11-07T11:55:05 so its not possible with a number? 2011-11-07T11:55:10 oh well, minor inconvenience 2011-11-07T11:55:14 only, if you wrap it 2011-11-07T11:55:18 lets see 2011-11-07T11:55:47 im not the one to explain this aspect, but when i used python i run into this 2011-11-07T11:55:55 the best one* 2011-11-07T11:56:31 it's still constantly 0 2011-11-07T11:56:47 show us the code 2011-11-07T11:56:54 globals aren't immutable, that's what the global keyword is for 2011-11-07T11:56:59 or we will kill you! 2011-11-07T11:57:00 write a test case showing your problem 2011-11-07T11:57:04 http://pastebin.com/bKM4h6Zv 2011-11-07T11:57:17 that's how i do 2011-11-07T11:57:41 RasmusL: i told you to make a list out of it 2011-11-07T11:57:48 turnnum=[0] 2011-11-07T11:57:59 infunc: global turnnum 2011-11-07T11:58:04 turnnum[0]+=1 2011-11-07T11:58:18 i did 2011-11-07T11:58:22 same scenario, just a list 2011-11-07T11:58:23 same problem 2011-11-07T11:58:56 i have no pyhton interpreter 2011-11-07T11:59:04 *** NotABug has quit IRC (Ping timeout: 265 seconds) 2011-11-07T11:59:13 wrapping in a list is ugly and not required 2011-11-07T12:00:12 anyway, its working for me 2011-11-07T12:00:24 :/ 2011-11-07T12:00:27 caution: if you have a better solution you should share it 2011-11-07T12:00:30 *** spp has quit IRC (Ping timeout: 265 seconds) 2011-11-07T12:00:40 mcstar: that's what the global keyword does 2011-11-07T12:01:04 *** Antimony has quit IRC (Ping timeout: 258 seconds) 2011-11-07T12:01:07 ? 2011-11-07T12:01:33 mcstar: http://paste.pocoo.org/show/504178/ 2011-11-07T12:02:16 thestinger look at my code 2011-11-07T12:02:23 that's what i do, it doesn't work.. not with a list either 2011-11-07T12:02:27 thestinger: what version? 2011-11-07T12:02:47 you're doing something else wrong 2011-11-07T12:02:50 *** Antimony has joined #aichallenge 2011-11-07T12:03:03 i use 3.2.2 2011-11-07T12:03:20 mcstar: 2.7.2 or 3.2.2 2011-11-07T12:03:32 *** McLeopold has joined #aichallenge 2011-11-07T12:03:43 works the same way 2011-11-07T12:03:48 the variable works perfectly in the module that it's allocated in 2011-11-07T12:03:59 thestinger: ok, well i have to think, but i had a similar problem when this(or something similar) wouldnt work 2011-11-07T12:04:00 but anywhere i use it in the other module 2011-11-07T12:04:00 you're overlooking something else 2011-11-07T12:04:01 it's 0 ALWAYS 2011-11-07T12:04:09 RasmusL: import doesn't work how you think it does :P 2011-11-07T12:04:21 python modules aren't like separate instances of a class 2011-11-07T12:04:53 let me try what you're doing 2011-11-07T12:04:55 sec 2011-11-07T12:04:58 or what :/ 2011-11-07T12:04:58 is it locally redefined as 0 because of import? 2011-11-07T12:05:29 nope 2011-11-07T12:05:31 works for me 2011-11-07T12:05:31 if you pasted the actual code instead of your interpretation it'd be easier to spot 2011-11-07T12:05:48 http://sprunge.us/ZCUS test.py 2011-11-07T12:06:21 http://sprunge.us/eIaR <- REPL session 2011-11-07T12:06:56 *** Garf has joined #aichallenge 2011-11-07T12:07:00 same thing in python 2.7.2 2011-11-07T12:07:15 you're doing something else wrong 2011-11-07T12:08:30 thestinger: i found uses of global in my scripts, llike you did 2011-11-07T12:08:33 mcstar: the list wrapper isn't required, that's the better solution 2011-11-07T12:08:49 caution: why am i remembering that? 2011-11-07T12:09:13 how long ago did you use python? 2011-11-07T12:09:22 I'd like to find out what the cause is now 2011-11-07T12:09:23 maybe, cause without the gloval keyword, it would only work that way? 2011-11-07T12:09:24 scoping used to be a lot different, but I think that was a long time ago 2011-11-07T12:10:15 I think it's been the same way as long as python 2 has existed 2011-11-07T12:10:20 thestinger: not that long ago, i havent used it for 2 years, since i only wrote little scripts for data analysis 2011-11-07T12:10:23 hmm 2011-11-07T12:10:30 python 2.1* 2011-11-07T12:10:35 yeah, so it had proper scoping when you used it 2011-11-07T12:10:43 sure it did 2011-11-07T12:10:51 maybe i just didnt know about the global keyword 2011-11-07T12:11:22 the thing is, the turnnum variable isn't even referenced anywhere other than what i pasted 2011-11-07T12:11:26 i just searched for it 2011-11-07T12:11:39 so there's not much more to see 2011-11-07T12:11:42 btw, you can make things like 'static' local variables in C/C++ 2011-11-07T12:11:48 are you sure the main module is even changing it? 2011-11-07T12:12:09 when you use a dict/list as a default parameter, it's a reference to the same list in each call 2011-11-07T12:12:13 *** ping-- has quit IRC (Ping timeout: 240 seconds) 2011-11-07T12:13:51 *** ping-- has joined #aichallenge 2011-11-07T12:14:40 caution: yes, cause i print it to the log from both modules 2011-11-07T12:14:49 and it's 0 during the whole match in module 2 :P 2011-11-07T12:14:54 are you redefining mybot in func.py? 2011-11-07T12:15:10 silly example: http://paste.pocoo.org/show/504191/ 2011-11-07T12:15:19 that removes the need for almost any use case for a global variable 2011-11-07T12:15:34 almost every* 2011-11-07T12:16:45 thestinger: are trying to emulate a closure? 2011-11-07T12:17:28 a static variable inside a function 2011-11-07T12:17:32 static variables are only useful for testing 2011-11-07T12:17:38 well, you could also write that as a closure 2011-11-07T12:17:47 otherwise they are bad design practice 2011-11-07T12:18:22 so how do you do memoization? a global var? 2011-11-07T12:18:55 a class 2011-11-07T12:18:58 i think these are called accumulators 2011-11-07T12:19:01 or a counter 2011-11-07T12:19:06 yes, use a class 2011-11-07T12:19:22 a member holds the counter, if you call the class, it increments it, and returns it 2011-11-07T12:19:27 derp this won't work at all :( 2011-11-07T12:19:43 you can do that, cant you? i mean call a class, but not "contruct" it 2011-11-07T12:19:49 i remember something like that 2011-11-07T12:20:24 make a simple test case of what you're trying and then when that works, investigate out how your real code is different 2011-11-07T12:20:42 s/out// 2011-11-07T12:20:46 hmm 2011-11-07T12:20:50 so the problem is that he is trying to make this work across modules? 2011-11-07T12:20:56 it does work across modules 2011-11-07T12:21:07 RasmusL: try my simple test case, build from there 2011-11-07T12:21:28 http://sprunge.us/ZCUS test.py http://sprunge.us/eIaR REPL session 2011-11-07T12:21:42 when you find out what's wrong, share it with me 2011-11-07T12:22:18 im gonna kill python 2011-11-07T12:23:08 function decorators are nicer for memoization anyway 2011-11-07T12:23:17 there's actually @functools.lru_cache in the standard lib 2011-11-07T12:23:47 *** KBentley57 has joined #aichallenge 2011-11-07T12:24:03 hey guys, is anyone running a local "ant server"? 2011-11-07T12:24:29 *** master_ninja has left #aichallenge 2011-11-07T12:24:39 *** dfsdf has joined #aichallenge 2011-11-07T12:25:03 http://pko.ch/2008/08/22/memoization-in-python-easier-than-what-it-should-be/ 2011-11-07T12:25:36 haha, i realized a bosonic ant-condensate 2011-11-07T12:26:08 caution: it's in the standard lib :P 2011-11-07T12:26:14 caution: http://docs.python.org/dev/library/functools.html 2011-11-07T12:26:33 with maxsize set to None, it's just a cache and doesn't do the LRU discarding 2011-11-07T12:26:47 New in version 3.2 2011-11-07T12:27:02 I still use <=2.6 2011-11-07T12:27:17 it's pure python 2011-11-07T12:27:31 it probably works in py2 if you copy the code to a module yourself 2011-11-07T12:27:46 http://sprunge.us/AOcC 2011-11-07T12:28:19 *** KBentley57 has quit IRC (Quit: Leaving) 2011-11-07T12:29:38 *** djr_ has joined #aichallenge 2011-11-07T12:30:19 ah, it uses nonlocal 2011-11-07T12:30:28 *** RasmusL has quit IRC (Ping timeout: 265 seconds) 2011-11-07T12:31:14 the link I posted has a concise memoization decorator 2011-11-07T12:31:35 it uses pickle, that's going to be really slow 2011-11-07T12:32:16 *** amstan has joined #aichallenge 2011-11-07T12:32:17 *** ChanServ sets mode: +o amstan 2011-11-07T12:32:35 nonlocal can be replaced with a dict, sec 2011-11-07T12:32:39 amstan: hello 2011-11-07T12:32:47 hey 2011-11-07T12:33:13 I see the new Language Support forum (I still don't know what to call these? boards? sections? areas?) 2011-11-07T12:33:21 forum 2011-11-07T12:33:24 and plural is forums 2011-11-07T12:33:25 subforum? 2011-11-07T12:33:39 okay, can you create the other forums so we can separate dev & tools 2011-11-07T12:34:11 later today, i'm in a break and i have to go soon 2011-11-07T12:34:27 oh, is it hard? 2011-11-07T12:34:39 no, but i want to make sure it's done right 2011-11-07T12:34:52 also.. i need to move a bunch of posts in the newly made forum 2011-11-07T12:35:12 also, my next big goal is to get the map gen constrainst correct, how much do you have in flux? 2011-11-07T12:35:47 i don't have any constrait or validity checking yet 2011-11-07T12:36:19 okay, I'm going to spruce up map.py, and make symmetric_map inherit and do proper checking 2011-11-07T12:36:32 do you get memory errors if you use too much memory or do you get dropped from the game? 2011-11-07T12:36:40 McLeopold: can you work from my code? 2011-11-07T12:36:49 McLeopold: mapgen branch 2011-11-07T12:36:55 amstan_mapgen i think 2011-11-07T12:37:02 amstan: I'll try, I already have a bunch of changes that haven't been checked it 2011-11-07T12:37:14 they've been there for weeks :( 2011-11-07T12:37:59 amstan: Hi all, I did not find an answer to this in the FAQ: What will be done with all the info gathered by the challenge? What rights does the challenger grant by taking part? Will the winning algorithms be published and/or discussed publicly in any way? Where can I find more info on such issues? 2011-11-07T12:38:00 the thing i like about my mapgen is that you just do map[pos]=LAND or WATER, and it symmetrically fills the other spots 2011-11-07T12:38:11 and it supports negative and positive values too, modules automatically 2011-11-07T12:38:53 caution: we don't have many legal things established 2011-11-07T12:39:08 amstan: okay, that stuff should go in the map.py 2011-11-07T12:39:08 caution: but from the past we didn't use the code for anything but for testing the bots 2011-11-07T12:39:15 and we don't publish them either 2011-11-07T12:39:37 McLeopold: please look at my code 2011-11-07T12:39:38 g2g 2011-11-07T12:39:41 cya 2011-11-07T12:39:46 *** ping-- has quit IRC (Read error: Connection reset by peer) 2011-11-07T12:39:52 *** ping-- has joined #aichallenge 2011-11-07T12:40:35 *** mviel_ has quit IRC (Remote host closed the connection) 2011-11-07T12:41:29 caution: there, got lru_cache to work with python2 2011-11-07T12:42:08 caution: http://sprunge.us/JGQU 2011-11-07T12:42:21 just had to replace misses/hits with a cacheinfo dictionary (since nonlocal is py3 only) 2011-11-07T12:43:08 OrderedDict? 2011-11-07T12:43:41 ? 2011-11-07T12:43:49 that's not python2 2011-11-07T12:44:06 oh 2011-11-07T12:44:44 *** amstan has quit IRC (Ping timeout: 276 seconds) 2011-11-07T12:45:11 it works for me :\ 2011-11-07T12:45:13 python 2.7.2 2011-11-07T12:45:20 @later tell amstan: you just deleted all the mapgen code and started over? :( I think I'll stick with the main branch 2011-11-07T12:45:20 McLeopold: Ready to serve, my lord. 2011-11-07T12:45:33 maybe it was added in 2.7 then 2011-11-07T12:45:53 caution: http://code.activestate.com/recipes/576693/ there's an implementation for earlier than 2.7 2011-11-07T12:46:08 http://pypi.python.org/pypi/ordereddict 2011-11-07T12:46:26 even better :P 2011-11-07T12:46:40 *** Lithosphere has quit IRC (Remote host closed the connection) 2011-11-07T12:46:54 *** Lithosphere has joined #aichallenge 2011-11-07T12:46:54 *** Lithosphere has joined #aichallenge 2011-11-07T12:47:24 you should compare the performance to the pickle version 2011-11-07T12:49:56 with maxsize=None it doesn't actually use the OrderedDict 2011-11-07T12:50:27 this guy just suicides: http://aichallenge.org/profile.php?user=1454 2011-11-07T12:50:32 doing pretty well 2011-11-07T12:50:41 *** skunx has quit IRC (Read error: Connection reset by peer) 2011-11-07T12:51:03 gives up his hill to rush someone else's early on 2011-11-07T12:51:10 *** smiley1983 has joined #aichallenge 2011-11-07T12:51:21 caution: the version of my bot on the site is pre-combat, pre-foraging, pre-defense :P 2011-11-07T12:51:29 *** ping-- has quit IRC (Read error: Connection reset by peer) 2011-11-07T12:51:34 *** ping-- has joined #aichallenge 2011-11-07T12:51:36 I have all that now but... http://aichallenge.org/profile.php?user=2255 2011-11-07T12:51:39 look how I did without it 2011-11-07T12:51:41 xD 2011-11-07T12:52:51 *** ajf|offline is now known as ajf 2011-11-07T12:54:10 oh so you're strcat 2011-11-07T12:54:15 yeah :) 2011-11-07T12:55:11 the name was already registered on freenode (I think it may have even been me... maybe not) 2011-11-07T12:56:16 what's its secret? 2011-11-07T12:56:46 out of curiousity: in the past did people keep their good bots hidden or did the level of play on the main server and tcp servers reflect pretty much where people where 2011-11-07T12:56:47 were 2011-11-07T12:57:28 *** Antimony has quit IRC (Ping timeout: 245 seconds) 2011-11-07T12:57:42 caution: the one on the site is seriously just like what you get from the tutorial + pathfinding + fancy movement and choice of path 2011-11-07T12:58:08 that got it up to the top 100, but I'm not really sure why it made it up even higher 2011-11-07T12:58:22 *** tmandry has joined #aichallenge 2011-11-07T12:58:23 *** tmandry has joined #aichallenge 2011-11-07T12:58:23 http://aichallenge.org/visualizer.php?game=2117&user=2255 2011-11-07T12:58:28 not so good back then 2011-11-07T12:59:06 originally I uploaded the python starter bot like the tutorial suggested xD 2011-11-07T12:59:13 I did my real bot in C++ 2011-11-07T12:59:54 I think a leftybot that collected food would do well 2011-11-07T13:00:03 probably better than my bot 2011-11-07T13:00:19 *** Antimony has joined #aichallenge 2011-11-07T13:00:27 probably not so well on the 'randomwalk' levels 2011-11-07T13:00:34 yeah, my defense code currently makes my bot a lot worse on maze levels 2011-11-07T13:00:40 so it's in a branch for now 2011-11-07T13:00:51 same with foraging code (but I'm almost ready to merge that) 2011-11-07T13:02:56 *** Relax has joined #aichallenge 2011-11-07T13:04:43 caution: do you use the tcp server? 2011-11-07T13:04:50 sometimes, always lose though 2011-11-07T13:05:01 *** RobotCaleb has joined #aichallenge 2011-11-07T13:05:11 yeah, the bots on there are way tougher 2011-11-07T13:05:40 the gameserver uses a fixed selection of maps right? 2011-11-07T13:05:59 I dunno, but I think they are going to regenerate all of them for the finals 2011-11-07T13:06:05 *** retybok_ has joined #aichallenge 2011-11-07T13:06:12 so people can't just identify the map and know where hills are, etc. 2011-11-07T13:06:19 *** RasmusL has joined #aichallenge 2011-11-07T13:06:21 so you can do that at the moment? 2011-11-07T13:06:44 I dunno if it would be within the rules (spirit of the game, etc.) 2011-11-07T13:07:03 you can definitely identify map symmetry and use that to find hills (or food I guess) 2011-11-07T13:08:11 *** dvladim has joined #aichallenge 2011-11-07T13:09:18 *** Antimony has quit IRC (Ping timeout: 258 seconds) 2011-11-07T13:09:47 *** marijnfs has joined #aichallenge 2011-11-07T13:11:36 *** QuirionPT has joined #aichallenge 2011-11-07T13:14:47 i solved the global variable issue.. just moved the global variable to its own module and import it in both modules 2011-11-07T13:14:52 seems to work now 2011-11-07T13:15:29 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-07T13:17:13 *** twymer has joined #aichallenge 2011-11-07T13:18:58 *** cyphase has quit IRC (Ping timeout: 258 seconds) 2011-11-07T13:20:06 *** blubber_ has joined #aichallenge 2011-11-07T13:20:14 hi 2011-11-07T13:20:25 *** tmandry has quit IRC (Ping timeout: 260 seconds) 2011-11-07T13:20:26 *** Cali__ has joined #aichallenge 2011-11-07T13:20:54 *** blubber_ has quit IRC (Client Quit) 2011-11-07T13:21:08 Hello, I had a question. How do I pass the exact instance of MyBot to another class so I can call MyBot methods in another class in Python? 2011-11-07T13:22:18 I will call a method in MyBot, say execute, and pass it in as Ant.execute(some variable, self) so in the class Ant I can call MyBot.someMethod() 2011-11-07T13:22:41 Is it possibe to do this with MyBot? 2011-11-07T13:22:56 *** datachomper has joined #aichallenge 2011-11-07T13:23:01 I want to modify the global variables in MyBot 2011-11-07T13:23:02 Squirrel away a reference to MyBot's self in the global dict? 2011-11-07T13:23:58 away? 2011-11-07T13:24:49 *** VoidXC has joined #aichallenge 2011-11-07T13:25:27 *** pedrosoro has joined #aichallenge 2011-11-07T13:25:50 As in "put in" 2011-11-07T13:26:45 *** pairofdice has quit IRC (Quit: Just keep breathing) 2011-11-07T13:26:48 you can pass the function itself to the library 2011-11-07T13:28:10 eg in a MyBot function call library.execute(blah, blahblah, self.someMethod) 2011-11-07T13:28:29 where execute is defined as def execute(blah, blahblah, func) 2011-11-07T13:28:33 ah okay thanks 2011-11-07T13:28:43 *** spp has joined #aichallenge 2011-11-07T13:28:55 so within execute you call func(a, b, c) and that's the equivalent of self.someMethod(a, b, c) 2011-11-07T13:29:07 (which is the equivalent of MyBot.someMethod(self, a, b, c) 2011-11-07T13:29:09 ) 2011-11-07T13:29:32 Will it automatically call, or can I give it a condition if passed in this way? 2011-11-07T13:30:07 func that is, when passed into execute, I only want to func to be run under a condition in execute 2011-11-07T13:30:52 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-07T13:31:04 *** cyphase has joined #aichallenge 2011-11-07T13:32:43 *** g0llum has joined #aichallenge 2011-11-07T13:35:20 well, you have to put the call in execute 2011-11-07T13:35:36 with your given condition 2011-11-07T13:35:46 *** QuirionPT has quit IRC (Quit: Leaving) 2011-11-07T13:36:12 *** RasmusL has quit IRC (Ping timeout: 265 seconds) 2011-11-07T13:36:33 remember you can edit all the files you got in the starter 2011-11-07T13:37:10 *** Cali__ has quit IRC (Ping timeout: 265 seconds) 2011-11-07T13:37:52 *** Hexren has joined #aichallenge 2011-11-07T13:38:04 *** merk has joined #aichallenge 2011-11-07T13:39:55 *** VoidXC has quit IRC (Quit: Page closed) 2011-11-07T13:41:02 *** ReverendFlibble has quit IRC (Ping timeout: 265 seconds) 2011-11-07T13:42:34 *** ChristianK has quit IRC (Remote host closed the connection) 2011-11-07T13:45:19 *** pedrosoro has quit IRC (Quit: Page closed) 2011-11-07T13:48:02 *** ikaros has joined #aichallenge 2011-11-07T13:48:07 *** liberforce has left #aichallenge 2011-11-07T13:48:22 *** Murashka has joined #aichallenge 2011-11-07T13:50:58 *** Antimony has joined #aichallenge 2011-11-07T13:55:04 *** RasmusL has joined #aichallenge 2011-11-07T13:55:05 hallo guys 2011-11-07T13:55:10 greetings from greece 2011-11-07T13:55:14 so umm 2011-11-07T13:55:24 i have a problem with compilation 2011-11-07T13:55:27 in c++ and windows 2011-11-07T13:55:29 how many nodes is it reasonable to go through with A* in a turn? 2011-11-07T13:55:33 can someone help me? 2011-11-07T13:55:44 i time out at around 40000 nodes in 1s 2011-11-07T13:56:01 but idk if i need to cut down on the pathfinding or fix the algorithm :) 2011-11-07T13:56:25 rasmus: 40000 /s sounds slow 2011-11-07T13:56:25 merk what's the issue? 2011-11-07T13:56:34 language? 2011-11-07T13:56:38 python 2011-11-07T13:57:11 hallo rasmusl 2011-11-07T13:57:18 i run windows 2011-11-07T13:57:24 yeah 2011-11-07T13:57:24 and i have dev-cpp 2011-11-07T13:57:39 i make a project with all the files in the starter package 2011-11-07T13:57:48 and i compile it 2011-11-07T13:58:48 *** merk1 has joined #aichallenge 2011-11-07T13:58:53 and then 2011-11-07T13:58:55 i run 2011-11-07T13:58:55 it 2011-11-07T13:58:56 with no problem 2011-11-07T13:58:59 the exe 2011-11-07T13:59:06 what commands i have to type 2011-11-07T13:59:10 in the windows 2011-11-07T13:59:12 to see if it works? 2011-11-07T13:59:15 go and ready? 2011-11-07T13:59:30 oh.. so the compilation is fine 2011-11-07T13:59:40 that's all i can help with, i have no idea how to test the exe 2011-11-07T13:59:41 yes i have the exe 2011-11-07T13:59:45 ok 2011-11-07T13:59:45 i use python :D 2011-11-07T13:59:50 but when 2011-11-07T13:59:53 i am going 2011-11-07T13:59:57 to test it in python 2011-11-07T14:00:05 play_one_game.cmd 2011-11-07T14:00:09 the bot crashes 2011-11-07T14:00:28 hmm.. i bet someone here who uses c++ can help with that 2011-11-07T14:00:39 you didnt change any code? 2011-11-07T14:00:56 and also the bot in python starter package 2011-11-07T14:00:56 ok thanks 2011-11-07T14:00:56 no 2011-11-07T14:00:58 i make no change 2011-11-07T14:01:00 in code 2011-11-07T14:01:04 in c++ and python starter 2011-11-07T14:01:06 package 2011-11-07T14:01:17 python starter package crashes too? 2011-11-07T14:01:23 yes 2011-11-07T14:01:37 maybe it is one of the bots 2011-11-07T14:01:57 meh doesnt sound likely 2011-11-07T14:02:05 are you missing some python packages maybe? 2011-11-07T14:02:13 *** merk has quit IRC (Ping timeout: 240 seconds) 2011-11-07T14:02:20 the other bots in sample_bot folder 2011-11-07T14:02:21 does it say anything after the crash? 2011-11-07T14:02:22 works good 2011-11-07T14:02:30 like whty 2011-11-07T14:02:33 -t 2011-11-07T14:02:47 *** Murashka has quit IRC (Ping timeout: 265 seconds) 2011-11-07T14:02:48 wait to check it 2011-11-07T14:03:11 *** Zannick has quit IRC (Ping timeout: 248 seconds) 2011-11-07T14:03:16 *** Pepelac has quit IRC () 2011-11-07T14:04:24 *** Zannick has joined #aichallenge 2011-11-07T14:04:47 *** djr_ has quit IRC (Ping timeout: 248 seconds) 2011-11-07T14:06:24 it writes nothing 2011-11-07T14:06:35 *** boegel has joined #aichallenge 2011-11-07T14:07:10 *** thobjo has joined #aichallenge 2011-11-07T14:07:42 *** TheLinker has quit IRC (Ping timeout: 240 seconds) 2011-11-07T14:07:50 *** kilae has quit IRC (Ping timeout: 252 seconds) 2011-11-07T14:07:52 i dont know what to do 2011-11-07T14:08:43 *** heinrich5991 has quit IRC (Ping timeout: 258 seconds) 2011-11-07T14:09:36 merk1: show us the command you're running to start the game 2011-11-07T14:10:05 play_one_game.cmd 2011-11-07T14:10:08 and it runs 2011-11-07T14:10:29 have you tried testbot too? 2011-11-07T14:11:00 no 2011-11-07T14:11:04 i havent 2011-11-07T14:11:09 how i run it 2011-11-07T14:11:10 ? 2011-11-07T14:13:07 i check it right now 2011-11-07T14:13:07 wait 2011-11-07T14:13:15 is there something called test_bot? 2011-11-07T14:13:21 yes 2011-11-07T14:13:22 you add your bot as an argument 2011-11-07T14:13:31 i do this right now 2011-11-07T14:13:33 and i run it 2011-11-07T14:14:28 *** heinrich5991 has joined #aichallenge 2011-11-07T14:15:21 umm.. do you get 500ms or 1s on the server? 2011-11-07T14:16:43 According to http://aichallenge.org/game_settings.php it's 500ms 2011-11-07T14:16:58 okay thanks 2011-11-07T14:17:18 its quite a lot if you are programming on a low level, unless you do a lot of heavy loops 2011-11-07T14:17:26 marijnfs do you use windows? 2011-11-07T14:18:24 i use python, so it's not a lot ;) 2011-11-07T14:18:49 *** Naktibalda has joined #aichallenge 2011-11-07T14:20:55 *** TheLinker has joined #aichallenge 2011-11-07T14:21:30 oh well this helped a lot 2011-11-07T14:21:49 down to around 2 ms/ant for pathfinding 2011-11-07T14:22:36 merk: no ubuntu 2011-11-07T14:22:41 and go 2011-11-07T14:25:30 RasmusL: What pathfinding algo are you using? 2011-11-07T14:25:54 A* 2011-11-07T14:26:11 rolled your own? 2011-11-07T14:26:40 yea 2011-11-07T14:27:38 probably not very efficient.. but i think other parts of my bot are worse 2011-11-07T14:27:45 so it doesn't hurt too bad :D 2011-11-07T14:28:45 that seems pretty good actually 2011-11-07T14:28:59 I used the A* algo in the networkx library and it was taking up to 10ms in some cases 2011-11-07T14:29:38 i have trouble when i get around 50 ants.. 2011-11-07T14:29:53 think it's unrelated to the pathfinding 2011-11-07T14:30:37 *** epicmonkey has joined #aichallenge 2011-11-07T14:33:02 datachomper: networkx isn't going to use a 2D array to represent the graph - it's too generic 2011-11-07T14:33:13 rolling your own A* is the way to go (or use something better) 2011-11-07T14:33:36 thestinger: I came to that conclusion. I'm rolling my own matrix + bfs thingy now 2011-11-07T14:33:40 *** smiley1983 has quit IRC (Ping timeout: 260 seconds) 2011-11-07T14:33:49 actually i'm using a heapq 2011-11-07T14:34:24 anyway, regular A* is stupid if you want to send 10 ants to 1 location 2011-11-07T14:34:26 and a visited set 2011-11-07T14:34:42 yeah 2011-11-07T14:34:48 *** dvladim has quit IRC (Read error: Operation timed out) 2011-11-07T14:34:53 or 400 ants :P 2011-11-07T14:34:56 RasmusL: Is the visited set included in your 2ms ? 2011-11-07T14:34:57 my bot is stupid though, it just goes for food and it never sends several ants for the same food 2011-11-07T14:35:24 it is 2011-11-07T14:35:28 *** choas has joined #aichallenge 2011-11-07T14:35:28 ah 2011-11-07T14:35:28 2 ms isn't a worst case at all though, im sure it can do much worse 2011-11-07T14:35:32 it won't be as stupid tommorow :) 2011-11-07T14:36:23 well :P 2011-11-07T14:36:49 I keep improving my bot but I still regard it as pretty bad because I know all the flaws 2011-11-07T14:37:07 it's going to be a while before I run out of obvious things I need to improve/implement 2011-11-07T14:37:33 best way to check if path[1] is allocated? 2011-11-07T14:37:34 *** hacklash has joined #aichallenge 2011-11-07T14:37:41 exception? 2011-11-07T14:37:47 len(path) ? 2011-11-07T14:37:57 oh right ^^ 2011-11-07T14:38:47 *** smiley1983 has joined #aichallenge 2011-11-07T14:41:04 Fluxid: I think your server hit the file descriptor limit again :P 2011-11-07T14:41:23 looks alive 2011-11-07T14:41:28 should i kick it with kill? 2011-11-07T14:41:36 hmm wait 2011-11-07T14:41:48 oh wait 2011-11-07T14:41:54 ok, so 2011-11-07T14:42:06 well, it doesn't seem to be playing games anymore 2011-11-07T14:42:09 1. it can crash on too many fd's open 2011-11-07T14:42:21 2. it can crash on sqllite (cannot open db filel) 2011-11-07T14:42:26 3. it can just stop 2011-11-07T14:42:30 lol 2011-11-07T14:42:37 its kinda bitchy isn't it ? 2011-11-07T14:42:46 last log is from 12 minutes ago 2011-11-07T14:43:21 before i kill it, let's count threads 2011-11-07T14:43:50 is it actually running? or just sleeping waiting on I/O 2011-11-07T14:43:52 actually, around 20 threads... 2011-11-07T14:44:06 two threads are running with ~60% CPU 2011-11-07T14:44:08 *** NoxiaZ^ has joined #aichallenge 2011-11-07T14:44:15 weird :\ 2011-11-07T14:44:24 *** Naktibalda has left #aichallenge 2011-11-07T14:44:30 and those threads have like 17 hours of time 2011-11-07T14:44:31 lol 2011-11-07T14:44:41 killing this shit 2011-11-07T14:45:01 oh, interrupt apparently unblocked it 2011-11-07T14:45:08 didn't have to kill 2011-11-07T14:45:10 holy shit 2011-11-07T14:45:29 i don't have enought word to describe this 2011-11-07T14:45:36 that's even more weird - that seems to imply it was stuck on I/O 2011-11-07T14:45:42 but it was using CPU? :\ 2011-11-07T14:46:02 thestinger: main thread was hung on data = client.recv(4096).strip() 2011-11-07T14:46:07 great 2011-11-07T14:46:20 one can open a connection and block whole server from accepting connections 2011-11-07T14:46:25 *** Larose has quit IRC (Quit: Leaving.) 2011-11-07T14:47:12 wow weird effect in my pathfinding 2011-11-07T14:47:52 https://github.com/berak/ants-tcp/blob/master/tcpserver.py#L470 yeah, great 2011-11-07T14:47:54 they allqueue up at a bottleneck and go one at a time ^^ 2011-11-07T14:48:12 this server is sooooo unsafe 2011-11-07T14:48:47 RasmusL: maybe try randomize them a little? 2011-11-07T14:49:02 so if they can't go forward, the go to the sides 2011-11-07T14:49:13 yeah 2011-11-07T14:49:23 so far i don't even consider other ants in the pathfinding except not colliding 2011-11-07T14:49:29 so that's probably why :) 2011-11-07T14:49:43 *** Murashka has joined #aichallenge 2011-11-07T14:49:52 I keep track of ants that were blocked by my idle ants, and I go back and try to move those in a loop until I can't move them anymore (yay for brute force) 2011-11-07T14:50:05 lol 2011-11-07T14:50:08 so that makes sure you get 1 ant through per turn (instead of leaving gaps) 2011-11-07T14:50:22 I still need to deal with chokepoints in a better way though 2011-11-07T14:51:57 *** marijnfs has quit IRC (Quit: leaving) 2011-11-07T14:53:03 *** thobjo has quit IRC (Ping timeout: 265 seconds) 2011-11-07T14:53:12 *** smiley1983 has quit IRC (Remote host closed the connection) 2011-11-07T14:53:27 *** smiley1983 has joined #aichallenge 2011-11-07T14:55:38 *** tmandry has joined #aichallenge 2011-11-07T14:55:38 *** tmandry has joined #aichallenge 2011-11-07T14:56:41 *** Accoun has quit IRC () 2011-11-07T14:56:58 *** dfsdf has quit IRC (Quit: Page closed) 2011-11-07T14:58:32 *** willvarfar has joined #aichallenge 2011-11-07T15:00:45 *** knott has joined #aichallenge 2011-11-07T15:01:58 hey guys. Any body know, is there any time that not counted in load time(for example when other players make their turns) 2011-11-07T15:03:09 *** Garf has quit IRC (Quit: Make a new plan, Stan!) 2011-11-07T15:03:12 *** spp has quit IRC (Quit: Page closed) 2011-11-07T15:03:19 *** bergmark has quit IRC (Ping timeout: 256 seconds) 2011-11-07T15:03:42 *** Garf has joined #aichallenge 2011-11-07T15:03:49 (I just use java with full OOP -> thus big memory pressure, and i want to make GC when my turn time doest countdown) 2011-11-07T15:04:01 *** tmandry has quit IRC (Read error: Connection reset by peer) 2011-11-07T15:04:06 *** tmandry has joined #aichallenge 2011-11-07T15:04:06 *** tmandry has joined #aichallenge 2011-11-07T15:05:05 knott: As far as you're concerned, your turn time is always counting down 2011-11-07T15:05:10 I believe, although I'm not sure if it's within the rules of the contest, that you can do your gc cycle after your turn has ended but before the next one begins 2011-11-07T15:05:58 knott: I believe you receive a SIGSTOP between turns, making any inter-turn calculation impossible 2011-11-07T15:06:08 *** interested has joined #aichallenge 2011-11-07T15:07:41 Minthos: This is time after I send "go" command to the server? 2011-11-07T15:08:27 Relax: unfortunately, there are no way to get sigstop in java 3: 2011-11-07T15:08:42 yeah 2011-11-07T15:09:01 *** Redgis has joined #aichallenge 2011-11-07T15:09:03 there is no way to catch sigstop in any language. feature of the OS 2011-11-07T15:09:04 thnx, i will check it out 2011-11-07T15:09:26 knott: Or any other language. SIGSTOP cannot be trapped: http://en.wikipedia.org/wiki/SIGSTOP 2011-11-07T15:09:37 you don't get sent a sigstop until after all the bots either send "go" or time out. 2011-11-07T15:10:51 so you could send your "go" then start the gc worst case is your garbage collection does not finish before you get a sigstop and it consumes some of the time on the next turn. 2011-11-07T15:13:37 yay my bot survived the first 1000 turn battle without timeout ^^ 2011-11-07T15:14:58 i dont get it guys. Why SIGSTOP raised when i just programmaticly call the GC? 2011-11-07T15:16:39 there should be no causality between the two 2011-11-07T15:17:01 *** Accoun has joined #aichallenge 2011-11-07T15:17:44 Instead of messing with the GC, couldn't you simply avoid creating a lot of garbage? 2011-11-07T15:17:51 Try re-using your objects 2011-11-07T15:18:16 *** bergmark has joined #aichallenge 2011-11-07T15:18:22 sounds like work 2011-11-07T15:20:44 I can't afford that. I have very tight [IRL] time limits. Rewriting code(i mean the base code) required for introduce obj reusing 2011-11-07T15:23:13 *** NotABug has joined #aichallenge 2011-11-07T15:27:45 *** Apophis_ has quit IRC (Read error: Connection reset by peer) 2011-11-07T15:27:51 *** RasmusL has quit IRC (Ping timeout: 265 seconds) 2011-11-07T15:28:02 *** Redgis has quit IRC (Read error: Connection reset by peer) 2011-11-07T15:28:45 *** Redgis has joined #aichallenge 2011-11-07T15:31:06 *** Apophis has joined #aichallenge 2011-11-07T15:33:01 knott: sigstop is raised by the engine when all bot have submitted their turn and while the engine is calculating the results. 2011-11-07T15:33:20 *** retybok_ has quit IRC (Ping timeout: 260 seconds) 2011-11-07T15:33:38 once the engine has the next turn to send it sends all bots a SIGCONT to resume (which is when the turn timer starts) 2011-11-07T15:34:08 oh, i thought SIGSTOP is fatal like SIGSEGV 2011-11-07T15:34:30 *** Antimony has quit IRC (Ping timeout: 240 seconds) 2011-11-07T15:34:34 no it pauses the process until it gets a SIGCONT 2011-11-07T15:35:19 ok thanks 2011-11-07T15:38:30 SIGKILL is fatal (obviously) 2011-11-07T15:42:52 *** ztfw has joined #aichallenge 2011-11-07T15:46:16 *** ProperChaos has joined #aichallenge 2011-11-07T15:47:31 hey guys, I'm having some trouble using the tcp-server 2011-11-07T15:48:19 I use the following command 2011-11-07T15:48:27 *** maaz139 has joined #aichallenge 2011-11-07T15:48:32 python "tcpclient.py" ants.fluxid.pl 2081 "path\to\exe" username pass -1 2011-11-07T15:48:43 i can see my exe starting in the task manager, but cpu usage remains 0 2011-11-07T15:49:04 the only output I'm getting is 'succesfully connected', but my bot is not moving at all 2011-11-07T15:49:04 whats the -1 for? 2011-11-07T15:49:09 no actions are being done 2011-11-07T15:49:10 unlimited games 2011-11-07T15:49:17 i dont use that parameter 2011-11-07T15:49:19 I've also tried 1 or 10 or whatever 2011-11-07T15:49:26 I'll try without 2011-11-07T15:49:49 btw, there are some superfluous quotes in your command 2011-11-07T15:50:14 *** tobym has joined #aichallenge 2011-11-07T15:50:21 the actual path has some spaces in it 2011-11-07T15:50:39 im sure you dont have to quote tcpclient 2011-11-07T15:50:52 so, is it working? 2011-11-07T15:50:52 I had a problem that port 2081 was blocked by my router you might just check it's open for you 2011-11-07T15:51:15 nope not working 2011-11-07T15:51:23 outgoing connections aren't blocked by my router 2011-11-07T15:51:26 change -1 to 1 for now 2011-11-07T15:51:31 also I can succesfully connect 2011-11-07T15:51:57 I don't think the new version treats -1 as meaning "infinite games" 2011-11-07T15:52:05 oh 2011-11-07T15:52:09 but I've edited the tcpclient.py script myself so I dunno 2011-11-07T15:52:19 well, even without that parameter, or just a 1 I'm not doing any moves 2011-11-07T15:52:21 thestinger: it is working wo that parameter 2011-11-07T15:52:27 and cpu usage stays at 0% for my bot process 2011-11-07T15:52:40 ProperChaos: does your bot work otherwise locally? 2011-11-07T15:52:56 yes 2011-11-07T15:53:03 and on the ranked server as well 2011-11-07T15:53:17 ProperChaos: try turning it off, than back on again 2011-11-07T15:53:32 haha which that would work 2011-11-07T15:53:33 The IT Crowd 2011-11-07T15:53:35 wish* 2011-11-07T15:53:38 lol 2011-11-07T15:53:38 _) 2011-11-07T15:54:18 I'm gonna go see with wireshark if I'm actually receiving some data 2011-11-07T15:55:24 it that port were blocked i dont think you could connect 2011-11-07T15:55:33 buti dont know much about networking 2011-11-07T15:55:37 if* 2011-11-07T15:57:10 yea I guess the problem is that somehow the data isn't inputed in my bot 2011-11-07T15:57:12 *** tmandry has quit IRC (Ping timeout: 258 seconds) 2011-11-07T15:57:35 ProperChaos: the tcpclient echos everything to the screen 2011-11-07T15:58:07 oh well it doesn't, the only thing it says is connected to ants.fluxid.pl:2081 2011-11-07T15:58:27 i mean it does, you are just not part of a game 2011-11-07T15:58:41 my statistics show that I've played games already 2011-11-07T15:58:47 replays show that my ants aren't moving at all 2011-11-07T15:59:37 *** amstan has joined #aichallenge 2011-11-07T15:59:37 *** ChanServ sets mode: +o amstan 2011-11-07T15:59:50 i had to wait quite much to be scheduled 2011-11-07T16:00:11 I've played 9 games already, but no action whatsoever 2011-11-07T16:00:18 I'm using python3 by the way 2011-11-07T16:00:24 and a C# bot 2011-11-07T16:02:12 *** ltriant has joined #aichallenge 2011-11-07T16:02:13 I'll try with Python2 just to be sure 2011-11-07T16:02:59 ProperChaos: what's your username there ? 2011-11-07T16:03:19 guys 2011-11-07T16:03:21 *** tmandry has joined #aichallenge 2011-11-07T16:03:21 *** tmandry has joined #aichallenge 2011-11-07T16:03:22 i installed 2011-11-07T16:03:29 http://ants.fluxid.pl/player/Devicated 2011-11-07T16:03:30 cygwin in order 2011-11-07T16:03:38 to compile in gcc 2011-11-07T16:03:41 the bot in c++ 2011-11-07T16:03:41 merk1: you are breaking the lines 2011-11-07T16:03:46 sorry 2011-11-07T16:03:53 *** Antimony has joined #aichallenge 2011-11-07T16:03:54 but when i type make i get error 2011-11-07T16:04:22 aha 2011-11-07T16:04:35 *** maaz139 has quit IRC (Ping timeout: 265 seconds) 2011-11-07T16:05:04 whoa, ive never had that many ants on tcp 2011-11-07T16:05:12 Python2 no different from python3 2011-11-07T16:06:35 merk1: we wont be able to help you if you neve tell us the error 2011-11-07T16:06:37 *** bergmark has quit IRC (Ping timeout: 240 seconds) 2011-11-07T16:06:39 r 2011-11-07T16:07:03 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-07T16:07:35 http://pastie.org/2827418 2011-11-07T16:07:41 look at this error mcstar 2011-11-07T16:08:12 *** boegel has quit IRC (Quit: Leaving) 2011-11-07T16:08:17 *** g0llum has joined #aichallenge 2011-11-07T16:08:43 *** tmandry has quit IRC (Ping timeout: 245 seconds) 2011-11-07T16:08:49 is this the c++ starter kit? 2011-11-07T16:08:53 kit->bot 2011-11-07T16:09:05 are you using it's makefile? 2011-11-07T16:09:34 240ms for a bfs search is a bit much right? 2011-11-07T16:09:34 yes 2011-11-07T16:09:39 the c++ starter 2011-11-07T16:09:44 and i type make 2011-11-07T16:09:46 in the directoey 2011-11-07T16:10:49 datachomper: Depends how much you're searching :) 2011-11-07T16:10:54 weird error 2011-11-07T16:11:07 merk1: i think theres something wrong with your gcc install 2011-11-07T16:11:10 But for the size of board ants deals with, yes 2011-11-07T16:11:17 what to do? 2011-11-07T16:11:21 test it in ubuntu> 2011-11-07T16:11:23 ? 2011-11-07T16:11:25 Yea it's just the 39x43 tutorial board 2011-11-07T16:11:28 it will work there 2011-11-07T16:11:34 *beats head against desk* 2011-11-07T16:11:46 merk1: ive never used gcc on windows 2011-11-07T16:11:55 i dont really know how is that possible 2011-11-07T16:12:05 what do you use 2011-11-07T16:12:05 ? 2011-11-07T16:12:07 i know there is something called mingw? 2011-11-07T16:12:23 yes 2011-11-07T16:12:30 is this it? 2011-11-07T16:12:54 ok 2011-11-07T16:12:58 yes it is 2011-11-07T16:13:08 i am going to test in ubuntu and i will see 2011-11-07T16:13:17 i think some environmental variables are not set correctly 2011-11-07T16:13:25 or the linker misses something 2011-11-07T16:17:47 *** treeform has joined #aichallenge 2011-11-07T16:18:36 doh, even if I use a python sample bot (hunterbot in this case) it's not working :/ 2011-11-07T16:19:32 *** bergmark has joined #aichallenge 2011-11-07T16:20:54 ProperChaos: take comfort in that im playing games on over tcp now 2011-11-07T16:21:02 -on 2011-11-07T16:21:09 wish i could help 2011-11-07T16:21:41 ah well 2011-11-07T16:21:46 I'll just keep trying :D 2011-11-07T16:23:52 *** Murashka has quit IRC (Quit: Page closed) 2011-11-07T16:24:34 Does anyone know the param to playgame.py so it follows the 500ms turn time used on the offical server. 2011-11-07T16:24:49 --turntime (use -h to see the various switches it takes) 2011-11-07T16:25:44 thestinger: ./playgame.py --help would have been a good place for me to start. Thanks! 2011-11-07T16:25:48 your computer is probably faster than the EC2 core you get though 2011-11-07T16:25:56 I'm using --turntime=100 atm 2011-11-07T16:26:08 *** MrEmile has joined #aichallenge 2011-11-07T16:26:27 Wow, how the heck did BenJackson2 get so good? 2011-11-07T16:27:56 *** Hexren has quit IRC (Quit: Leaving.) 2011-11-07T16:28:18 it seems like he deals with combat really well 2011-11-07T16:28:38 and he has defenders - but since he deals with combat well, they aren't useless like the normal defenders people use 2011-11-07T16:30:00 *** MrEmile has quit IRC (Client Quit) 2011-11-07T16:30:25 like they don't just stay in a static position 2011-11-07T16:30:28 *** MrEmile has joined #aichallenge 2011-11-07T16:30:35 my bot has a lot of trouble getting past his defense :( 2011-11-07T16:30:41 *** NotABug has quit IRC (Ping timeout: 265 seconds) 2011-11-07T16:30:59 *** merk1 has quit IRC () 2011-11-07T16:31:05 yeah the combat is impressive 2011-11-07T16:31:10 I can't figure out how to do combat 2011-11-07T16:31:27 neither can I, I just added some heuristics to avoid certain death :) 2011-11-07T16:31:52 well, I haven't really tried yet actually xD 2011-11-07T16:32:01 too many other things to do 2011-11-07T16:33:09 * MrEmile made a stupid bot that just goes forward and changes direction 5% of the time, then a few more clever ones and is annoyed that the stupid bot still wins against the others 2011-11-07T16:34:01 yeah, I'm very annoyed that all the foraging/re-exploration code I've written isn't a clear improvement over what I had before 2011-11-07T16:34:10 *** goffrie has joined #aichallenge 2011-11-07T16:34:26 thestinger: same here :O( 2011-11-07T16:35:40 http://aichallenge.org/profile.php?user=2255 <- my bot before I had anything dealing with enemy ants/combat/re-exploration/foraging/defense 2011-11-07T16:36:03 I have no idea what's so good about it :( 2011-11-07T16:36:33 *** ztfw has quit IRC (Remote host closed the connection) 2011-11-07T16:36:39 my new code definitely makes it "better" at fighting and getting lots and lots of ants 2011-11-07T16:36:50 *** conor_f has joined #aichallenge 2011-11-07T16:36:53 but my old code is better at just rushing at a hill and getting points 2011-11-07T16:38:13 *** null_rad1x has joined #aichallenge 2011-11-07T16:38:38 well, my combat code is an 100% positive impact for now b/c I let them be really aggressive/stupid 2011-11-07T16:38:56 foraging/defense was a lot of work and isn't really an improvement though yet 2011-11-07T16:39:33 *** rtward has quit IRC (Ping timeout: 260 seconds) 2011-11-07T16:39:33 *** Relax has quit IRC (Quit: Leaving) 2011-11-07T16:39:44 what's good about it is that it defeats enemies' attempts at clever micro by just rushing in with superior numbers in a concentrated squad 2011-11-07T16:40:19 while not (always) neglecting foraging and defense entirely 2011-11-07T16:40:21 *** djr_ has joined #aichallenge 2011-11-07T16:40:45 the version on the official site doesn't defend :P 2011-11-07T16:40:57 we talkin bout Ants? 2011-11-07T16:40:57 and it doesn't re-explore to find food, so it can't survive for very long 2011-11-07T16:41:32 http://ants.fluxid.pl/replay.24957 there's an example of the new code 2011-11-07T16:41:39 ok but when new ants spawn they don't just follow the other ants in an infinite line to die at some entrenched enemy position 2011-11-07T16:41:44 you must be the guy who posted on LW wondering how he's in the top 200 without even paying paying attention to enemy ants? 2011-11-07T16:42:03 how long does it take for your ranking/skill to stabilize? 2011-11-07T16:42:07 LW? 2011-11-07T16:42:27 wait a minute 2011-11-07T16:42:31 i'm in the wrong window 2011-11-07T16:42:32 of course 2011-11-07T16:42:33 :P 2011-11-07T16:42:47 of course you people are talking about ants. nevermind 2011-11-07T16:42:54 hehe 2011-11-07T16:43:21 I think I should just stop defense/foraging from starting for 50 turns or something xD 2011-11-07T16:46:51 u_: as it happens, I did post at LW :) 2011-11-07T16:46:55 how long to stabilize rank? 2011-11-07T16:47:00 (though I don't think you were talking to me) 2011-11-07T16:47:12 oh god this is embarrassing :P 2011-11-07T16:47:13 I just remembered a short while that I'm not working tomorrow so I had a beer and a glass of whiskey. Time to write AI! :) 2011-11-07T16:48:00 im going insane 2011-11-07T16:48:30 kdevelop handles c++11 well, now im used to it, but it keeps forgetting my shortcut 2011-11-07T16:48:40 i set tab to emacs like align 2011-11-07T16:48:51 in every 10 minutes ive got to set it againm 2011-11-07T16:48:58 for people using, say, c/c++, if your bot crashes how can you work out where? when I run mine piping through the visualiser I don't get any console output or anything in debug.txt, it just exits 2011-11-07T16:49:20 conor_f: you'll generally be pretty close within 10 games 2011-11-07T16:49:29 valgrind and edit playgame.py to print stderr 2011-11-07T16:49:36 thx 2011-11-07T16:49:52 what should the sigma thing be then? 2011-11-07T16:50:40 janzert: ^^ 2011-11-07T16:51:07 because I'm ranked >1300 but I'm playing people ranked <600 2011-11-07T16:51:17 and my skill is 49 but my mu is 59 2011-11-07T16:51:37 *** hacklash has quit IRC (Quit: hacklash) 2011-11-07T16:51:57 conor_f: I think skill is mu - sigma*3 2011-11-07T16:52:05 *** gadhafi has joined #aichallenge 2011-11-07T16:52:08 kwl, ty :) 2011-11-07T16:52:13 your sigma will drop as you play more games (assuming you play consistently) 2011-11-07T16:52:56 yeah 2011-11-07T16:53:17 consistency isn't that good for me :P although everyone I'm playing is over 1000 ranks ahead of me 2011-11-07T16:57:25 random question here, anyone here using linux and using an iPod with it? 2011-11-07T16:58:41 switched from [] to set() and reduced my bfs from 250ms to 25ms, still slow, bet getting there 2011-11-07T16:59:21 why do you need a set()? 2011-11-07T16:59:35 you don't need to keep track of visited nodes if you're sneaky with it 2011-11-07T16:59:43 thestinger: please elaborate 2011-11-07T16:59:52 you're planning on reusing the BFS for multiple ants right? 2011-11-07T16:59:59 like BFS from a target across the map, marking off distances 2011-11-07T17:00:09 sort of 2011-11-07T17:00:10 in a 2d list/array.array/numpy array or whatever 2011-11-07T17:00:25 well, if you mark off distances in a 2d array (for sending 100 ants to 1 target, etc.) 2011-11-07T17:00:32 thestinger: I'm just using the built-in python list-of-lists 2011-11-07T17:00:35 *** peyton has joined #aichallenge 2011-11-07T17:00:58 yeah, a python list is a dynamic array of pointers, so that's fine 2011-11-07T17:01:20 thestinger: would I see an improvement if I switched to numpy multi-dimensional array? 2011-11-07T17:01:34 anyway, if you're marking off all the distances in a 2d array/whatever, you can just make sure the new node is a lower value than what you would put there 2011-11-07T17:01:42 so you don't need the visited set() 2011-11-07T17:02:16 datachomper: well, assuming it's a 2d array of integers (distances), you'll see an improvement with array.array (since it's a dynamic array of values, not pointers to them) 2011-11-07T17:02:31 numpy arrays are like array.array, but you can allocate them _much_ faster 2011-11-07T17:02:59 you should also reuse the same list/array rather than reallocating a new one 2011-11-07T17:03:15 Huh, I didn't think about marking my 2d array as I bfs'ed. 2011-11-07T17:03:43 Currently I do the entire bfs, then return a list of lists for each "distance" a list of nodes is from the target loc 2011-11-07T17:04:01 ah, I just mark off the distances in a 2d array of ints 2011-11-07T17:04:08 so my "paths" are just implicitly stored in that 2011-11-07T17:04:23 I just move an ant to a neighbour with a lower distance for pathfinding 2011-11-07T17:04:36 thestinger: So, you mark your target with some value, subtract something, then mark it's neighbors with a lower value, rinse repeat? 2011-11-07T17:04:49 well, I mark the target with 0 2011-11-07T17:04:59 then the 4 neighbouring nodes are 1 2011-11-07T17:05:01 then 2, etc. 2011-11-07T17:05:22 and instead of a visited set, I just make sure the new value I would be setting a cell to is lower than the old 2011-11-07T17:05:35 if it's lower, I set it and put the new neighbouring nodes onto the queue 2011-11-07T17:05:42 *** twymer has quit IRC (Ping timeout: 240 seconds) 2011-11-07T17:05:51 heapq? 2011-11-07T17:06:01 it's not python :P 2011-11-07T17:06:06 oh, ok 2011-11-07T17:06:15 I know python well though (better than C++, which I'm using for this) 2011-11-07T17:06:25 but yea, that's muuuch simpler, thanks for the protip 2011-11-07T17:06:41 *** Palmik has quit IRC (Remote host closed the connection) 2011-11-07T17:07:46 you can use numpy if you're using python2, and that should give you a decent speed boost for allocation or setting an entire 2d array back to a default value like -1 2011-11-07T17:08:04 iirc they don't have the python3 version of numpy on the server 2011-11-07T17:08:19 Oh cool, so I won't have to include it in my source? 2011-11-07T17:08:31 yeah, just 'import numpy' 2011-11-07T17:08:41 it's way to big to fit in the 2MB max size .zip 2011-11-07T17:09:03 numpy also needs native code, so it's rather hard to include manually 2011-11-07T17:09:13 basically, numpy lets you avoid slow python for loops 2011-11-07T17:09:52 array.array is pretty much the same memory layout, but without nice ways to iterate/manipulate it 2011-11-07T17:09:58 *** DeGi has joined #aichallenge 2011-11-07T17:14:42 *** retybok_ has joined #aichallenge 2011-11-07T17:23:28 *** ProperChaos has quit IRC (Quit: Page closed) 2011-11-07T17:24:37 *** twymer has joined #aichallenge 2011-11-07T17:28:02 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-07T17:28:18 *** g0llum has joined #aichallenge 2011-11-07T17:29:38 Is it safe to assume the attack radius won't change? 2011-11-07T17:30:02 probably but why bother? 2011-11-07T17:30:28 Minthos: because implementing generic algorithms is harder? 2011-11-07T17:30:58 shouldn't be that much harder 2011-11-07T17:31:12 depends what you do ;-) 2011-11-07T17:31:15 *** ZaN_ has joined #aichallenge 2011-11-07T17:31:49 you need to calculate the distance anyway, might as well let your program do it when it's running than do it yourelf when you're writing it 2011-11-07T17:32:08 +rather 2011-11-07T17:32:14 *** the-mgt has quit IRC (Quit: the-mgt) 2011-11-07T17:32:17 nevermind 2011-11-07T17:32:17 retybok_: it's been said it's very unlikely for it to change 2011-11-07T17:32:26 *** NoxiaZ^ has quit IRC (Read error: Connection reset by peer) 2011-11-07T17:32:28 caution: thanks for the info 2011-11-07T17:33:57 *** iam-lu4 has joined #aichallenge 2011-11-07T17:34:24 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-07T17:34:35 hello 2011-11-07T17:34:39 there 2011-11-07T17:36:21 *** iam-lu4 has quit IRC (Client Quit) 2011-11-07T17:36:30 hey, thestinger 2011-11-07T17:36:38 hi 2011-11-07T17:36:49 strcat_next is failing miserably to _forage :P 2011-11-07T17:36:53 what rank are you at today conor? 2011-11-07T17:37:18 1300 ish I think 2011-11-07T17:37:20 conor_f: I just merged the foragelite branch :) 2011-11-07T17:37:33 but I should be up ~600-650 I think 2011-11-07T17:37:40 it's not as good as the forage branch at foraging, but not as bad at rushing 2011-11-07T17:37:51 :P I just made defence, tweaking it now 2011-11-07T17:39:11 http://ants.fluxid.pl/replay.25042 2011-11-07T17:39:11 *** Antimony has quit IRC (Ping timeout: 276 seconds) 2011-11-07T17:40:13 nice man :) 2011-11-07T17:40:27 I keep trying to move bots I don't have control over lol :P 2011-11-07T17:40:30 *** Akranis has quit IRC (Ping timeout: 240 seconds) 2011-11-07T17:40:34 *** Scryer has joined #aichallenge 2011-11-07T17:41:04 good luck with that conor_f ;-) 2011-11-07T17:41:14 *** ikaros has joined #aichallenge 2011-11-07T17:41:25 thestinger: your ants have given up their biggest advantage in combat, but gained better foraging 2011-11-07T17:41:42 *** MrEmile has quit IRC (Ping timeout: 240 seconds) 2011-11-07T17:42:00 well, I don't forage for the first 30 turns 2011-11-07T17:42:07 I might increase it to ~50 2011-11-07T17:42:25 the combat/movement is the same, but 25% of my ants are used for foraging 2011-11-07T17:42:56 what do you mean by foraging? 2011-11-07T17:43:08 looking for food in explored areas 2011-11-07T17:43:21 thx :) 2011-11-07T17:43:31 I don't forage until turn 150 2011-11-07T17:44:11 but all my ants eat food of opportunity when they move to their exploration targets 2011-11-07T17:46:53 *** Antimony has joined #aichallenge 2011-11-07T17:48:50 *** marina has joined #aichallenge 2011-11-07T17:48:59 bleh 2011-11-07T17:49:06 my ant tracking isn't as good as I though 2011-11-07T17:49:08 *thought 2011-11-07T17:49:12 *** willvarfar has quit IRC (Quit: Leaving) 2011-11-07T17:49:21 based on me trying to move ants I don't own 2011-11-07T17:50:55 lol 2011-11-07T17:50:59 remember to account for the ants you didn't move and the ants that die 2011-11-07T17:51:39 yeah, I have 2011-11-07T17:51:53 I say dest = current for non-movers 2011-11-07T17:51:58 and the dead ones are erased 2011-11-07T17:52:10 INFO: invalid o 116 80 s # duplicate order 2011-11-07T17:52:14 what does that mean? 2011-11-07T17:52:46 you sent the same ant twice 2011-11-07T17:54:11 *** chris__0076 has quit IRC (Quit: Leaving) 2011-11-07T17:54:32 not only do you need to make sure you don't send two ants to the same spot, you also need to make sure you don't send the same ant to two spots 2011-11-07T17:55:47 *** dwins has quit IRC (Quit: Lost terminal) 2011-11-07T17:56:13 *** eccentric has quit IRC (Ping timeout: 245 seconds) 2011-11-07T17:56:21 *** marina has left #aichallenge 2011-11-07T17:57:48 *** bergmark has quit IRC (Ping timeout: 260 seconds) 2011-11-07T17:58:32 *** The_Digital_Ninj has joined #aichallenge 2011-11-07T17:59:00 hmm 2011-11-07T17:59:01 true 2011-11-07T17:59:48 ok 2011-11-07T18:00:18 *** The_Digital_Ninj has quit IRC (Read error: Connection reset by peer) 2011-11-07T18:00:20 did good, now have the basis of defense :) just need to sort out some logic errors tomorrow and should be all done with the defending 2011-11-07T18:00:29 then onto hill pwning :P 2011-11-07T18:01:02 the tricky part is to alert an ant that will try to step to another ant's present location, that can no longer step to it's desired location, and hence must stay there 2011-11-07T18:02:03 I have them move to a different location 2011-11-07T18:02:19 *** the-mgt has joined #aichallenge 2011-11-07T18:02:53 *** Redgis has quit IRC (Ping timeout: 244 seconds) 2011-11-07T18:03:14 *** rew_ has joined #aichallenge 2011-11-07T18:03:38 *** Antimony has quit IRC (Ping timeout: 260 seconds) 2011-11-07T18:04:11 elo 2011-11-07T18:04:22 *** conor_f has quit IRC (Quit: leaving) 2011-11-07T18:06:38 *** rew_ has quit IRC (Client Quit) 2011-11-07T18:10:15 *** marina_ has joined #aichallenge 2011-11-07T18:10:49 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-07T18:13:09 *** choas has quit IRC (Ping timeout: 256 seconds) 2011-11-07T18:14:43 *** delt0r_ has quit IRC (Read error: Operation timed out) 2011-11-07T18:15:55 *** marina has joined #aichallenge 2011-11-07T18:15:55 *** marina_ has quit IRC (Read error: Connection reset by peer) 2011-11-07T18:16:04 *** marina has left #aichallenge 2011-11-07T18:16:59 *** redblue_ has joined #aichallenge 2011-11-07T18:21:18 *** epicmonkey has quit IRC (Ping timeout: 240 seconds) 2011-11-07T18:22:22 *** ejnahc has joined #aichallenge 2011-11-07T18:24:36 *** marina has joined #aichallenge 2011-11-07T18:25:59 *** marina has left #aichallenge 2011-11-07T18:30:28 *** delt0r_ has joined #aichallenge 2011-11-07T18:30:41 Any chance that the time between games will be reduced? Seems like a long time to wait for feedback. 2011-11-07T18:31:03 use the tcp servers 2011-11-07T18:31:06 Scryer: http://ants.fluxid.pl/howto 2011-11-07T18:31:12 Thanks. 2011-11-07T18:31:18 *** tmandry has joined #aichallenge 2011-11-07T18:31:18 *** tmandry has joined #aichallenge 2011-11-07T18:31:31 *** interested has quit IRC (Ping timeout: 265 seconds) 2011-11-07T18:36:55 *** Jak_o_Shadows has joined #aichallenge 2011-11-07T18:37:51 *** retybok_ has quit IRC (Ping timeout: 248 seconds) 2011-11-07T18:42:48 *** Fandekasp has quit IRC (Ping timeout: 258 seconds) 2011-11-07T18:43:28 i can detect enemies now 2011-11-07T18:43:46 one step away from a 1-ply prediction 2011-11-07T18:44:33 *** datachomper has left #aichallenge 2011-11-07T18:46:31 *** Antimony has joined #aichallenge 2011-11-07T18:48:07 *** ZaN_ has quit IRC (Quit: Page closed) 2011-11-07T18:49:18 *** jl1990 has joined #aichallenge 2011-11-07T18:49:49 *** knott has quit IRC (Quit: Page closed) 2011-11-07T18:51:19 does that mean 2 steps away from 2-ply? 2011-11-07T18:52:45 first ive got to see how ill do time-wise 2011-11-07T18:52:58 almost there 2011-11-07T18:53:49 what is 1-ply? 2011-11-07T18:54:04 any one use full game prediction and pruning? 2011-11-07T18:54:26 as in simulate couple steps forward and pick the outcomes you like? 2011-11-07T18:55:01 "couple" 2011-11-07T18:55:08 literally 2011-11-07T18:56:41 lol, compilation takes more than 5 secs, its annoying 2011-11-07T19:00:26 *** capa has joined #aichallenge 2011-11-07T19:01:00 oh 2011-11-07T19:01:14 *** ccc_ has joined #aichallenge 2011-11-07T19:01:31 4 ants max, 3 ply, couple of that, and my bot hangs for ~300ms 2011-11-07T19:01:39 2ply 2011-11-07T19:01:39 hi, my bot is getting time out, but i have no idea why. any idea? 2011-11-07T19:01:48 not 3, thats waay too much 2011-11-07T19:02:00 *** bergmark has joined #aichallenge 2011-11-07T19:02:08 *** foRei has quit IRC (Read error: Connection reset by peer) 2011-11-07T19:02:09 *** tmandry has quit IRC (Ping timeout: 252 seconds) 2011-11-07T19:02:27 *** Scryer has quit IRC (Ping timeout: 265 seconds) 2011-11-07T19:03:00 hello, how can i avoid timing outs ? 2011-11-07T19:03:23 capa: disable timeouts 2011-11-07T19:03:36 just change the engine, its 1 line 2011-11-07T19:05:20 eh.. wrong answer? 2011-11-07T19:05:48 mcstar: sry, but i cant find 2011-11-07T19:05:53 Minthos: what did you expect? 2011-11-07T19:05:57 capa: check the remaining time in your algorithm and end your turn if it gets too low 2011-11-07T19:06:10 *** Chris_0076 has joined #aichallenge 2011-11-07T19:06:37 Minthos: do you know any functions to use in C? 2011-11-07T19:06:45 to count the time? 2011-11-07T19:06:51 no but refer to your starting kit 2011-11-07T19:06:59 capa: gettimeofday 2011-11-07T19:07:11 if it doesn't have functions for it, use a better starting kit 2011-11-07T19:07:17 yeah, you can use any of the POSIX stuff 2011-11-07T19:07:26 gettimeofday() or clock_gettime() 2011-11-07T19:07:26 Minthos: the c starting kit doesnt have any :-( 2011-11-07T19:07:56 it is 1000 miliseconds ? 2011-11-07T19:08:16 i dont get why people dont read the main page 2011-11-07T19:08:17 your bot is informed at the start of each game 2011-11-07T19:08:32 maybe they don't like to read 2011-11-07T19:08:55 giving them false info is not sportsmanlike 2011-11-07T19:09:17 ? 2011-11-07T19:09:31 who is sportsman? and who gave false info? 2011-11-07T19:09:35 01:03 < mcstar> capa: disable timeouts 2011-11-07T19:09:35 01:03 < mcstar> just change the engine, its 1 line 2011-11-07T19:09:45 Minthos: best advice ever 2011-11-07T19:09:55 i always do that 2011-11-07T19:10:15 the description is still in the topic 2011-11-07T19:10:34 well, not anymore, but when i was debugging in common lisp, the engine was still running 2011-11-07T19:10:38 as well as my bot 2011-11-07T19:10:41 worst advice I've seen today (and yesterday, since today is only 70 minutes old) 2011-11-07T19:10:50 that wouldnt be possible with timeouts enables 2011-11-07T19:10:55 d 2011-11-07T19:11:14 *** jl1990 has quit IRC () 2011-11-07T19:11:32 Minthos: this is not a tutorial competition 2011-11-07T19:11:57 if somebody helps you with your language, or just with programming in general thats ok 2011-11-07T19:12:00 *** pedrosorio has joined #aichallenge 2011-11-07T19:12:01 but dont expect it 2011-11-07T19:12:19 I agree, I'm not saying anyone are obligated to help 2011-11-07T19:12:35 but it looked like you were intentionally giving him false info, which is bad 2011-11-07T19:12:43 Minthos: now really, when you first read his question, werent you smiling? 2011-11-07T19:12:48 if your intentions were good, I misinterpreted you 2011-11-07T19:12:59 not everyone's first language is english 2011-11-07T19:13:04 Minthos: the info wasnt bad, or misleading 2011-11-07T19:13:11 so while i, for example, could understand the sarcasm in the sentence 2011-11-07T19:13:12 i DID disable timeouts 2011-11-07T19:13:20 *** Harpyon has quit IRC (Quit: Textual IRC Client: http://www.textualapp.com/) 2011-11-07T19:13:25 not everyone will 2011-11-07T19:13:57 I thought it was, disabling timeouts probably wouldn't help him - only if he was timing out /only/ when he was pausing the bot in a debugger 2011-11-07T19:14:13 which he most likely would have mentioned if he was 2011-11-07T19:14:17 yep, not the answer he seekd 2011-11-07T19:14:28 but it motivates him to ask again with greater detail 2011-11-07T19:14:58 or it motivates him to just give up and do something else 2011-11-07T19:15:10 or that 2011-11-07T19:15:27 he isnt "contest material" if he gives up after this 2011-11-07T19:15:39 you know, friendliness and helfulness tends to motivate better than deception and bitterness 2011-11-07T19:15:55 so what, he still deserves good treatment? 2011-11-07T19:16:00 Minthos: if i am ying you are yang 2011-11-07T19:16:04 *** grwip has quit IRC (Quit: Leaving) 2011-11-07T19:16:44 probably not but it's clear we disagree on this 2011-11-07T19:16:52 thats ok 2011-11-07T19:16:56 *** Jak_o_Shadows has quit IRC (Read error: Connection reset by peer) 2011-11-07T19:17:06 *** grwip has joined #aichallenge 2011-11-07T19:17:11 *** grwip_ has joined #aichallenge 2011-11-07T19:17:24 *** grwip_ has left #aichallenge 2011-11-07T19:19:04 *** Jak_o_Shadows has joined #aichallenge 2011-11-07T19:19:10 Minthos: oh, just for the record, a lot of people asked me for help in private 2011-11-07T19:19:31 and while i helped i tend to be "bitter" now :D 2011-11-07T19:20:47 is it a common strategy to find paths and save that kind of data between turns? i'm having a lot of issues with doing this and adapting to ants nearby, newly found food/hills, etc. 2011-11-07T19:21:02 trying to decide if my approach is wrong or if i'm just not thinking of a good solution 2011-11-07T19:22:29 *** grwip has quit IRC (Quit: Leaving) 2011-11-07T19:22:52 twymer: i dont save anything, its too cheap to recalculate 2011-11-07T19:23:11 but later maybe 2011-11-07T19:23:25 hm.. i'm timing out in a lot of my official games (though not on the tcp server) 2011-11-07T19:23:34 which was why i started caching my paths found 2011-11-07T19:23:47 *** TheDigitalNinja has joined #aichallenge 2011-11-07T19:23:54 i don't do any checking of turn time remaining, which i believe i should start doing probably 2011-11-07T19:24:34 twymer: The official servers have a turn time of 500ms and I think the tcp servers are several seconds, I'm sure to help with latency. 2011-11-07T19:24:39 i find it surprising that many people reach timeouts so soon 2011-11-07T19:24:40 oh 2011-11-07T19:24:53 5000 2011-11-07T19:24:59 mcstar: they did not code in C? 2011-11-07T19:24:59 mcstar: well, that's because we write bad code ;) 2011-11-07T19:25:07 *** Jak_o_Shadows has quit IRC (Ping timeout: 256 seconds) 2011-11-07T19:25:20 twymer: not necessarily, but you should profile 2011-11-07T19:25:24 and build test cases 2011-11-07T19:25:34 before you put an algorith into the bot 2011-11-07T19:25:58 mcstar: what language are you using? 2011-11-07T19:26:17 c++11 2011-11-07T19:26:29 that's a really good idea, i've been putting things in and at that point it's hard to profile or test 2011-11-07T19:28:41 anyone using python here that has had luck not saving results of path finding between turns? 2011-11-07T19:29:26 *** Jak_o_Shadows has joined #aichallenge 2011-11-07T19:29:39 *** Antimony has quit IRC (Ping timeout: 256 seconds) 2011-11-07T19:30:28 *** tobym has quit IRC (Remote host closed the connection) 2011-11-07T19:30:51 twymer: you should ask Fluxid he is well-ranked and does it in python 2011-11-07T19:33:01 *** replore has joined #aichallenge 2011-11-07T19:35:18 *** Glival has joined #aichallenge 2011-11-07T19:35:26 *** ecarpenter84 has joined #aichallenge 2011-11-07T19:38:13 is there any good channels for dicussing ai that isnt related to this challenge 2011-11-07T19:38:56 *** Glival has quit IRC (Client Quit) 2011-11-07T19:38:58 *** VoidXC has joined #aichallenge 2011-11-07T19:40:09 hey everyone. Anyone available to help me with a problem in my algorithm for going to food? 2011-11-07T19:42:10 *** olexs1 has quit IRC (Quit: Leaving.) 2011-11-07T19:43:46 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-07T19:43:55 *** Antimony has joined #aichallenge 2011-11-07T19:46:54 ecarpenter84, #ai-class possibly 2011-11-07T19:47:46 *** Antimony_ has joined #aichallenge 2011-11-07T19:48:05 *** datachomper1 has joined #aichallenge 2011-11-07T19:48:44 *** Antimony has quit IRC (Ping timeout: 258 seconds) 2011-11-07T19:49:52 anyway, thanks for the help mcstar.. time to do some profiling outside of the bot :) 2011-11-07T19:50:17 good luck 2011-11-07T19:53:01 *** Antimony_ has quit IRC (Ping timeout: 255 seconds) 2011-11-07T19:53:07 *** troxel has joined #aichallenge 2011-11-07T19:53:35 *** Antimony has joined #aichallenge 2011-11-07T19:53:40 *** troxel has quit IRC (Client Quit) 2011-11-07T19:54:14 I implemented a function to count the time, and now it get timed out at first turn 2011-11-07T19:54:31 *** pedrosorio has quit IRC (Quit: Page closed) 2011-11-07T19:54:38 *** McLeopold has left #aichallenge 2011-11-07T19:55:44 *** hacklash has joined #aichallenge 2011-11-07T19:55:44 shrodinger's time 2011-11-07T19:57:31 *** Jak_o_Shadows has quit IRC (Remote host closed the connection) 2011-11-07T19:57:36 presumably your function to count the time is doing something wrong 2011-11-07T19:57:49 more like heisenberg's time 2011-11-07T20:00:40 *** Antimony has quit IRC (Ping timeout: 255 seconds) 2011-11-07T20:01:15 i dont get it 2011-11-07T20:01:54 i think it was heisenberg 2011-11-07T20:02:04 who did what? 2011-11-07T20:02:11 i dot get the time reference 2011-11-07T20:04:47 heisenberg showed that you can't accurately measure both position and speed of an electron, iirc 2011-11-07T20:05:04 position and linear momentum 2011-11-07T20:05:17 and somewhat related is the quantum quirk of being unable to measure a result without changing it 2011-11-07T20:05:27 no 2011-11-07T20:05:31 not related 2011-11-07T20:05:46 okay, mr. smarty :P 2011-11-07T20:06:10 after all this i still dont get this time thing 2011-11-07T20:06:20 but nvm 2011-11-07T20:06:44 anyway, schrodinger's was a thought experiment where he postulated that the system was in every possible state at once until collapsed into one by observation 2011-11-07T20:06:52 Zannick: maybe he meant Plack's time 2011-11-07T20:06:53 i'm not sure what you mean by this time thing 2011-11-07T20:07:17 you mean < capa> I implemented a function to count the time, and now it get timed out at first turn 2011-11-07T20:07:27 "shrodinger's time" or "heisenberg's time" in relation to timeouts 2011-11-07T20:07:37 measuring the time => timeout 2011-11-07T20:07:56 hm 2011-11-07T20:08:31 actually, you can measure time infinitely precisely 2011-11-07T20:08:34 *** Zaphus has joined #aichallenge 2011-11-07T20:08:47 thats not a parameter that has inherent uncertainty 2011-11-07T20:09:35 only the physical processes, i.e. excited states has energy-time uncertainty relations 2011-11-07T20:09:41 have* 2011-11-07T20:10:23 *** VoidXC has quit IRC (Quit: Page closed) 2011-11-07T20:10:34 heisenjoke 2011-11-07T20:11:26 *** tmandry has joined #aichallenge 2011-11-07T20:11:26 *** tmandry has joined #aichallenge 2011-11-07T20:12:07 *** cyphase has quit IRC (Ping timeout: 258 seconds) 2011-11-07T20:12:32 *** redblue_ has quit IRC (Ping timeout: 265 seconds) 2011-11-07T20:13:07 *** rtward has joined #aichallenge 2011-11-07T20:14:42 *** cyphase has joined #aichallenge 2011-11-07T20:21:01 *** JorgeB has quit IRC (Ping timeout: 240 seconds) 2011-11-07T20:21:43 *** capa has quit IRC (Ping timeout: 265 seconds) 2011-11-07T20:22:41 *** DeGi has quit IRC (Ping timeout: 265 seconds) 2011-11-07T20:25:00 *** TheLinker has quit IRC (Quit: Bye) 2011-11-07T20:25:58 *** tmandry has quit IRC (Ping timeout: 260 seconds) 2011-11-07T20:27:25 *** janzert has quit IRC (Read error: Connection reset by peer) 2011-11-07T20:28:20 *** janzert has joined #aichallenge 2011-11-07T20:32:33 *** TheDigitalNinja has quit IRC (Ping timeout: 256 seconds) 2011-11-07T20:36:25 *** treeform_ has joined #aichallenge 2011-11-07T20:37:13 *** eccentric has joined #aichallenge 2011-11-07T20:39:29 *** p_l has joined #aichallenge 2011-11-07T20:39:56 *** treeform has quit IRC (Ping timeout: 260 seconds) 2011-11-07T20:43:56 So I have a 2d array that I'm bfs-ing. Would it be faster for me to create an "edge map" upfront and reference that for finding a node's neighbors, or calculating the n,e,s,w neighbors once per node? 2011-11-07T20:48:00 datachomper1, depends if you have the data upfront or not 2011-11-07T20:48:44 *** ecarpenter84 has quit IRC (Quit: Page closed) 2011-11-07T20:51:55 *** Wondering has joined #aichallenge 2011-11-07T20:52:13 If you time out, is it generally because you run out of memory? How much memory are we given 2011-11-07T20:52:34 it's because you took longer than the turn limit 2011-11-07T20:52:41 :D 2011-11-07T20:52:44 you should have at least 1gb of memory which is heaps 2011-11-07T20:52:54 I print the time left as I go, and I never seem to get below 900 2011-11-07T20:53:03 datachomper1: i can't imagine using extra memory for an edge map would be advantageous unless it's expensive to look up squares 2011-11-07T20:53:18 but it's certainly cheap to compute adjacent directions 2011-11-07T20:53:19 a1k0n: it makes coding aesthetically nicer :P 2011-11-07T20:53:33 antimatroid: i did what you do 2011-11-07T20:53:38 possibly, yes... though you could just write code that generates the "edges" on the fly 2011-11-07T20:53:44 when i get water i delete pointers 2011-11-07T20:53:48 logging.debug("Time Remaining After Move: %s"% (str(ants.time_remaining()))) is how I check 2011-11-07T20:54:01 i just added a vector of neighbour locations to each squrae and remove appropriate ones as i get water information 2011-11-07T20:54:14 don't have to worry about checking if a location is water then either :P 2011-11-07T20:54:21 if that never seems I get low, so that is why I am asking about memory 2011-11-07T20:54:49 almost 3 2011-11-07T20:54:53 gotta sleep 2011-11-07T20:55:02 *** mcstar has quit IRC (Quit: WeeChat 0.3.6) 2011-11-07T20:55:30 a1k0n: thanks 2011-11-07T20:55:37 *** Antimony has joined #aichallenge 2011-11-07T20:55:53 is there a way to check if I timeout due to time versus memory? I am using python 2011-11-07T20:56:09 how would one timeout from memory? 2011-11-07T20:56:19 wouldn't running out of memory result in 'crashed'? 2011-11-07T20:57:24 ah, it would crash despite having enough local memory despite running out of alloted memory by the program from gogle 2011-11-07T20:57:47 I wonder why when I print out the time I always have plenty of time using ants.time_remaining() :-/ 2011-11-07T20:58:23 unless that is not the proper way to check 2011-11-07T20:59:10 *** RobotCaleb has quit IRC (Ping timeout: 255 seconds) 2011-11-07T21:03:46 *** null_rad1x has quit IRC (Ping timeout: 265 seconds) 2011-11-07T21:06:39 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-11-07T21:08:16 *** ccc has joined #aichallenge 2011-11-07T21:10:32 Wondering: are you using node? 2011-11-07T21:10:45 *** GeorgeSebastian has joined #aichallenge 2011-11-07T21:10:46 *** GeorgeSebastian has joined #aichallenge 2011-11-07T21:10:50 node? 2011-11-07T21:11:04 *** AVAVT has joined #aichallenge 2011-11-07T21:13:06 treeform_ most likely not, I am using python 2.7 nothing else in particular 2011-11-07T21:13:49 *** djr_ has quit IRC (Read error: Connection reset by peer) 2011-11-07T21:14:18 *** djr_ has joined #aichallenge 2011-11-07T21:15:33 *** Antimony has joined #aichallenge 2011-11-07T21:16:36 *** roflmao has joined #aichallenge 2011-11-07T21:17:11 *** tobym has joined #aichallenge 2011-11-07T21:17:11 if i do print('go') I time out within 5 turns, it seems I would never time out if I end my turn on the first line of doturn 2011-11-07T21:20:54 the reason i say this is because i am using node, and i have simmilar issues some times 2011-11-07T21:21:12 my thinking is that my i/o buffers are not flushing 2011-11-07T21:21:17 to keep up with the server 2011-11-07T21:21:30 i might be starting the server way to late 2011-11-07T21:22:47 ah, I timeout on my local machine just facing bots 2011-11-07T21:22:56 i either might be getting the "go" too late or server does not get my "go" in time 2011-11-07T21:23:04 this is local as well 2011-11-07T21:23:17 my thinking is that when your CPU is at 100% buffers might move slower 2011-11-07T21:23:24 but i am using node it has an odd IO loop 2011-11-07T21:23:51 I wish I knew 2011-11-07T21:24:28 *** tobym has quit IRC (Remote host closed the connection) 2011-11-07T21:24:30 i want to put some logging in the game server to see who and what takes this time 2011-11-07T21:25:51 *** ccc has quit IRC (Quit: Computer has gone to sleep.) 2011-11-07T21:28:21 hoooooooooooly coooooooooow 2011-11-07T21:28:25 *** Wondering has quit IRC (Ping timeout: 265 seconds) 2011-11-07T21:29:18 *** Antimony has quit IRC (Ping timeout: 240 seconds) 2011-11-07T21:30:10 HOLY SHIT 2011-11-07T21:30:29 so I was doing minor tweaks to my weighted bfs 2011-11-07T21:30:37 but my new bot would always have struggle against old 2011-11-07T21:30:45 *** treeform_ has quit IRC (Remote host closed the connection) 2011-11-07T21:30:45 very even matches 2011-11-07T21:30:57 so … I added 4 new lines of code 2011-11-07T21:31:00 and — [0,1,0,0,207,7,0,0] 2011-11-07T21:31:03 207 being the new bot 2011-11-07T21:31:27 that's the score 2011-11-07T21:31:37 207 ? thats a lot of hills! 2011-11-07T21:31:43 ants 2011-11-07T21:31:50 so before even matchup 2011-11-07T21:31:55 4 new lines — new bot completely rapes the old one 2011-11-07T21:32:02 let's see if it wasn't a one time thing 2011-11-07T21:34:24 so yea technically not score but :P 2011-11-07T21:36:36 yea 2011-11-07T21:36:38 holy rape 2011-11-07T21:38:01 *** TheLinker has joined #aichallenge 2011-11-07T21:45:45 bring it to ants.fluxid.pl! 2011-11-07T21:46:37 yeah now notice it might rape my previous bot but actually be worse against fluxid bots 2011-11-07T21:46:47 basically I made a certain percentage of my ants aggressively attack hills 2011-11-07T21:46:56 that's their only purpose 2011-11-07T21:47:01 but in ACTUAL combat, they are very conservative 2011-11-07T21:47:03 causing a sort of moving wall 2011-11-07T21:47:12 generally a good idea 2011-11-07T21:47:24 *** gadhafi has quit IRC (Quit: Page closed) 2011-11-07T21:47:25 buuuuut I think I have a bug in my code where the same conservativeness is being applied to exploring making ants "wall-freeze" when they shouldn't 2011-11-07T21:47:40 but yeah my survivability increased amazing 2011-11-07T21:49:40 *** hjax has joined #aichallenge 2011-11-07T21:52:15 *** djr_ has quit IRC (Ping timeout: 258 seconds) 2011-11-07T21:52:35 a1k0n: are the bots you see playing on tcp and the main site representative of what they will look like for the final competition or do people keep their good ideas hidden? 2011-11-07T21:52:40 hi 2011-11-07T21:53:34 bugnuts: while i don't think anyone's hiding anything, i don't think the bots are representative of the final compo just because there's enough time between now and then for some decent innovations 2011-11-07T21:54:00 it's more that what you see in the last week is basically what you are up against... 2011-11-07T21:54:02 early on in the planetwars competition, i had a top-10 bot easily, and it took me like an hour to write 2011-11-07T21:54:08 by the end i was down to 200 or so 2011-11-07T21:54:21 *** hacklash has left #aichallenge 2011-11-07T21:54:51 I guess the real question is how different is the leader board a few days before the competition vs the final results... 2011-11-07T21:55:02 oh. not very different. 2011-11-07T21:55:13 bayeselo/trueskill seem to work pretty well. 2011-11-07T21:56:03 at least, that has been true in my vast experience of two other similar contests 2011-11-07T21:56:36 cool. It was partly since there are some things I know my bot would be susceptible to and you could code around it but if no one is using it then it's probably not worth it. 2011-11-07T21:57:27 e.g. decoys; some bots are very aggressive and will chase 1 ant with an army of 10 but if no one is using them into the finals then its sort of safe to assume that I don't need to worry about it 2011-11-07T21:58:37 oh, yeah, that's not a bad idea. i don't think it's safe to assume anything at this point 2011-11-07T21:58:43 re: tactics used in the future 2011-11-07T21:59:04 I am not but also don't want to preempt things we just are not seeing. 2011-11-07T21:59:17 here's me being way too aggressive against an impossible situation: http://ants.fluxid.pl/replay.25443 2011-11-07T22:01:10 i have no idea how to start coding bfs... 2011-11-07T22:01:27 heh 2011-11-07T22:01:44 *** Antimony has joined #aichallenge 2011-11-07T22:01:48 a1k0n: at least it was not a lone ant that came in and took your hill :) 2011-11-07T22:02:00 *** userjjb has joined #aichallenge 2011-11-07T22:03:00 yeah, i fought back but with not nearly enough ants. turning them away from the attack would have easily worked 2011-11-07T22:03:08 but my combat code isn't good enough to do it anyway 2011-11-07T22:04:02 yeah thats kind of where I am... 2011-11-07T22:04:45 I realised what I was doing for gathering made for a very poor starting point for adding combat and had to rip things apart. 2011-11-07T22:04:46 I'm trying to get to where you guys are at :D 2011-11-07T22:04:59 then I'll worry about what you are worring about LOL 2011-11-07T22:05:04 (I'll soon be there, I hope) 2011-11-07T22:05:56 roflmao: yay i got to play against your bot 2011-11-07T22:06:04 yeah, mine failed miserably 2011-11-07T22:06:15 but I noticed some deeeeep flaws in my new wall code 2011-11-07T22:06:19 currently fixing them :) 2011-11-07T22:06:21 it wasn't too bad 2011-11-07T22:06:23 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-11-07T22:06:28 my bot doesnt work :D 2011-11-07T22:06:35 i kept picking off your defenders without anything to back it up 2011-11-07T22:07:11 but i think your defense cost you expansion initally 2011-11-07T22:07:20 yes 2011-11-07T22:07:37 I'm seeing that too - because I wasn't lagging too far behind, but once you 'deadlocked' me in my base 2011-11-07T22:08:01 your bot looks a lot like my first one did 2011-11-07T22:08:42 heh you had a guy surrounded and you didn't press the advantage 2011-11-07T22:09:02 Have you seen lordap - I think he wins the gathering game with a bullet. 2011-11-07T22:09:29 a1k0n: yep, because of my deadlocking code that's all screwy 2011-11-07T22:09:41 a1k0n: my combat bots are a little - too - conservative now 2011-11-07T22:10:03 http://ants.fluxid.pl/replay.25441 2011-11-07T22:10:06 see center area 2011-11-07T22:10:10 wtf did my bots do just sitting there 2011-11-07T22:10:14 (I'm fixing this as we speak) 2011-11-07T22:10:19 ants** 2011-11-07T22:10:20 bugnuts: hm, i wonder if he's symmetry-aware 2011-11-07T22:10:52 don't think so. end distributions don't look like it 2011-11-07T22:10:56 though his expansion rate doens't look too different from anyone else 2011-11-07T22:11:58 he gets capped when people have decent combat but anyone without it and his better gathering just gets him big #s of ants. 2011-11-07T22:12:03 a1k0n: anything you'd recommend? :) 2011-11-07T22:13:18 well... i guess prefer doing things sooner than later? :) 2011-11-07T22:13:46 by deadlocking you mean your ants just sit there? 2011-11-07T22:13:51 yeah 2011-11-07T22:14:09 yeah. i dunno. it's hard, because both ants have to act in concert for it not to be a loss 2011-11-07T22:14:10 which is all horribly wrong when i have 5 ants that should be gathering food but they are sitting there because of one stupid enemy ant in the middle 2011-11-07T22:14:18 so if you optimize each ant independently you get stuck 2011-11-07T22:14:37 ...unless you have some way to assume the other guy will do the right thing 2011-11-07T22:14:57 yes 2011-11-07T22:15:01 i dunno, assign a probability to them going after you and flip a coin? 2011-11-07T22:15:15 that's actually not a bad idea 2011-11-07T22:15:20 I was thinking of counting allies vs enemies in region 2011-11-07T22:15:26 *** hjax has quit IRC (Quit: Page closed) 2011-11-07T22:15:35 and proceeding with combat if I overpower with X percentage (let's say 30% just for now) 2011-11-07T22:15:38 also a good idea 2011-11-07T22:15:38 otherwise staying back 2011-11-07T22:16:45 *** george has quit IRC (Ping timeout: 265 seconds) 2011-11-07T22:17:13 aight I think I've fixed this baby 2011-11-07T22:17:15 time to try again 2011-11-07T22:19:08 I was freezing ants if there were no enemy ants in region 2011-11-07T22:19:13 causing huge massive deadlock ha 2011-11-07T22:19:24 because when no enemy ants were found, by definition my bfs costmap would be all 0's 2011-11-07T22:19:37 so my code was saying "0 distance = really close enemy ants!! stop!" 2011-11-07T22:20:00 lol I just found exactly the same bug about 2 minutes ago. 2011-11-07T22:20:15 *** hjax has joined #aichallenge 2011-11-07T22:20:45 I can't write bfs, I think my code is going to be really slow 2011-11-07T22:20:53 but since I seed the map with exploration points it only happens when I have seen the entire map. 2011-11-07T22:21:43 heh bugnuts 2011-11-07T22:22:16 I have a system for exploration that I think works relatively well 2011-11-07T22:22:31 I seed map with exploration points like you, but I assign a fixed percentage of my ants to exploration 2011-11-07T22:22:36 that percentage decreases as I get more ants 2011-11-07T22:22:43 so early game it's mostly food and exploration 2011-11-07T22:22:54 as I gain critical mass of ants, the percentage of exploration decreases and combat increases 2011-11-07T22:23:16 yeah the seeding explore points is not as adaptive as I would like but I need to focus more on getting combat in before I address it. 2011-11-07T22:23:35 http://ants.fluxid.pl/replay.25490 — that's what I like to see 2011-11-07T22:25:00 *** hjax has quit IRC (Remote host closed the connection) 2011-11-07T22:25:18 *** hjax has joined #aichallenge 2011-11-07T22:25:25 *** djr_ has joined #aichallenge 2011-11-07T22:28:40 *** hjax has quit IRC (Remote host closed the connection) 2011-11-07T22:30:46 contestbot: seen janzert 2011-11-07T22:30:46 amstan: janzert was last seen in #aichallenge 5 hours, 41 minutes, and 26 seconds ago: conor_f: you'll generally be pretty close within 10 games 2011-11-07T22:31:03 ? 2011-11-07T22:31:19 this C guy is really insistent: http://aichallenge.org/forums/viewtopic.php?f=26&t=1797 2011-11-07T22:31:47 we could just tell him no then ;) 2011-11-07T22:32:01 I was looking into it a little bit earlier today 2011-11-07T22:32:02 he also found a memory leak in the C starter(which flag got fixed), and suggested some changes to the makefile 2011-11-07T22:32:10 hadn't made up my mind yes 2011-11-07T22:32:13 yet even 2011-11-07T22:32:48 it's probably going to be no though 2011-11-07T22:33:08 why not? 2011-11-07T22:33:19 i always compile my code with c99 2011-11-07T22:33:24 implicit int declaration goes away which has a fair chance of breaking current (poorly) written code 2011-11-07T22:34:20 if I get a chance I'm going to implement a test that would compile and test every current submission with the change to see if it breaks 2011-11-07T22:35:18 hjax: you there? 2011-11-07T22:35:30 janzert: poorly? what do you mean? random people's bots? 2011-11-07T22:35:41 or the starter package? 2011-11-07T22:35:53 poorly as in you really shouldn't be using an implicit int declaration anyway 2011-11-07T22:36:00 I hope the starter doesn't 2011-11-07T22:36:06 I wouldn't expect it to 2011-11-07T22:36:20 http://ants.fluxid.pl/replay.25492 - hahahahaha 2011-11-07T22:36:22 :P 2011-11-07T22:36:22 you really think it's going to break things so bad? 2011-11-07T22:36:41 <_flag> You could add another extension specifically for C99 code (basically for one person) 2011-11-07T22:36:44 I don't know but I'm pretty committed to not breaking current submissions 2011-11-07T22:36:53 _flag: it's really a hassle to do these things 2011-11-07T22:36:57 and since I don't know :) 2011-11-07T22:37:24 yeah I really don't want to keep proliferating minor language version differences 2011-11-07T22:37:27 janzert: i just don't think it should be our job to avoid breaking fragile code 2011-11-07T22:38:09 if we accept a submission we should if at all possible not then make changes that break it 2011-11-07T22:38:46 if it was beta I'd make the change for sure 2011-11-07T22:39:11 you're affraid about new workers breaking the current submissions? 2011-11-07T22:39:36 what if no bots actually get broken? 2011-11-07T22:39:38 yes, or current workers that haven't happened to get that submission yet 2011-11-07T22:39:50 see the test I mentioned above :) 2011-11-07T22:39:58 oh.. ic 2011-11-07T22:40:06 (10:34:23 PM) janzert: if I get a chance I'm going to implement a test that would compile and test every current submission with the change to see if it breaks 2011-11-07T22:40:18 i was imagining something more complicated.. like compiler script trying with and without c99 and taking the one which worked 2011-11-07T22:40:26 yeah, ok 2011-11-07T22:40:45 janzert: ok.. i suggest we just tell that guy exactly what we're doing 2011-11-07T22:40:55 go ahead :) 2011-11-07T22:42:18 I get " warning: type defaults to ‘int’ in declaration of ‘a’ [enabled by default]" with gcc -std=c99 (c89 gives no warning) 2011-11-07T22:42:19 have you heard anything out of jeff recently? 2011-11-07T22:42:24 that shouldn't break code 2011-11-07T22:42:25 janzert: nope 2011-11-07T22:42:39 thestinger: it would on the main server 2011-11-07T22:42:41 janzert: he's off the face of the earth again 2011-11-07T22:42:46 amstan: :( 2011-11-07T22:43:01 thestinger: basically warnings are treated as errors 2011-11-07T22:43:06 ah 2011-11-07T22:49:00 *** Antimony has joined #aichallenge 2011-11-07T22:50:00 *** djr_ has quit IRC (Quit: Leaving) 2011-11-07T22:54:57 *** AVAVT has quit IRC (Ping timeout: 265 seconds) 2011-11-07T22:57:42 http://ants.fluxid.pl/replay.25534 2011-11-07T22:57:45 :| 2011-11-07T22:59:08 *** Jak_o_Shadows has joined #aichallenge 2011-11-07T22:59:49 *** dvladim has joined #aichallenge 2011-11-07T23:03:17 *** Antimony_ has joined #aichallenge 2011-11-07T23:06:07 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-11-07T23:06:13 *** Antimony_ is now known as Antimony 2011-11-07T23:07:38 *** TheLinker has quit IRC (Quit: Bye) 2011-11-07T23:10:07 roflmao, you got some issues 2011-11-07T23:10:28 I think he just needed some motivation 2011-11-07T23:10:30 he was scared 2011-11-07T23:10:51 he just needed some friends 2011-11-07T23:11:05 yeah 2011-11-07T23:12:47 *** easystar has joined #aichallenge 2011-11-07T23:13:17 *** Zaphus has quit IRC (Quit: Page closed) 2011-11-07T23:14:00 *** yoden has quit IRC (Quit: Leaving.) 2011-11-07T23:14:07 *** tmandry has joined #aichallenge 2011-11-07T23:14:07 *** tmandry has joined #aichallenge 2011-11-07T23:16:00 *** treeform has joined #aichallenge 2011-11-07T23:19:29 *** califax has joined #aichallenge 2011-11-07T23:20:42 if you do: print('go') in do_turn(self, ants) in python, does it end your turn? 2011-11-07T23:20:54 *** tmandry has quit IRC (Ping timeout: 240 seconds) 2011-11-07T23:21:19 *** avdg1 has quit IRC (Quit: Leaving.) 2011-11-07T23:21:27 califax: i think so 2011-11-07T23:21:45 *** dvladim has quit IRC (Ping timeout: 260 seconds) 2011-11-07T23:22:18 califax, you're better off just returning from do_turn 2011-11-07T23:22:39 because ants.finish_turn() will run after that and it prints 'go' out as well 2011-11-07T23:22:51 *** tmandry has joined #aichallenge 2011-11-07T23:22:51 *** tmandry has joined #aichallenge 2011-11-07T23:23:01 *** avdg has joined #aichallenge 2011-11-07T23:23:59 andar: Thank you, I am doing this because my bot times out and I am trying to get it to not time out. Doing print('go') causes me to time out even sooner than without it if it is the first line in my do_turn method 2011-11-07T23:24:20 *** Antimony has quit IRC (Ping timeout: 260 seconds) 2011-11-07T23:24:32 i don't know how it all works.. but maybe it's a problem with writing 'go' out twice in row? 2011-11-07T23:24:38 just try 'return' instead 2011-11-07T23:31:11 *** twymer has quit IRC (Ping timeout: 248 seconds) 2011-11-07T23:35:41 *** Zaphus has joined #aichallenge 2011-11-07T23:36:18 *** cyphase has quit IRC (Quit: http://www.cyphase.com/) 2011-11-07T23:36:32 is fluxid stuck? 2011-11-07T23:37:47 it happens :), the tcp server is buggy 2011-11-07T23:38:18 it leaks sockets and threads 2011-11-07T23:38:32 and it can get stuck if someone opens a connection and never sends something 2011-11-07T23:39:03 (oooops, I'm a culprit of that) 2011-11-07T23:39:07 (i thought my bot would just timeout) 2011-11-07T23:39:08 well, one of my bots is playing a game still 2011-11-07T23:39:14 yea it started 2011-11-07T23:39:24 well, it would be hard to do it by accident 2011-11-07T23:39:30 if you ctrl-c you close the socket 2011-11-07T23:40:14 *** cyphase has joined #aichallenge 2011-11-07T23:41:11 i'm surprised no one has written a better tcpserver.. it shouldn't be very difficult 2011-11-07T23:42:28 *** u_ has quit IRC (Quit: u_) 2011-11-07T23:43:45 Hey guys, can you enter comments in your bot commands (with # comment or something like that) - it's not in the game spec but ants.py looks like it handles it (not intending the comments to get to the live server, but just checking to be safe) 2011-11-07T23:44:15 you can comment your code.. 2011-11-07T23:45:05 comment my code ? pigs might fly! but can I put comments in my command stream ? 2011-11-07T23:45:33 you're just looking for a way to print out some debug info? 2011-11-07T23:46:28 take a look at some of the example bots.. some of them use the included logutil.py to set up a logger you can use to print out debug info without interrupting the command stream 2011-11-07T23:46:50 or you could just use the python logging module directly 2011-11-07T23:47:10 sort of, but I want it in the onscreen stream that I see when running on the TCP server. 2011-11-07T23:47:15 yes 2011-11-07T23:47:20 that will do it 2011-11-07T23:47:51 *** hacklash has joined #aichallenge 2011-11-07T23:48:56 logutil.py sounds good... if only I used Python for my bot and not some other antiquated language :-) 2011-11-07T23:49:18 You should just be able to write to stderr 2011-11-07T23:49:27 Or use your language's logging library that does that for you 2011-11-07T23:50:45 http://ants.fluxid.pl/replay.25598 2011-11-07T23:51:23 getting better 2011-11-07T23:53:41 *** Antimony has joined #aichallenge 2011-11-07T23:57:22 just not sure if I should implement enemy avoidance for food and exploreres 2011-11-07T23:58:22 *** djr_ has joined #aichallenge