2011-12-02T00:03:05 *** antimatroid has joined #aichallenge 2011-12-02T00:04:08 *** avdg has quit IRC (Quit: Leaving.) 2011-12-02T00:06:09 *** Antimony has quit IRC (Ping timeout: 244 seconds) 2011-12-02T00:10:24 *** delt0r_ has joined #aichallenge 2011-12-02T00:14:01 *** yoden has quit IRC (Quit: Leaving.) 2011-12-02T00:18:56 *** jmcarthur has quit IRC (Read error: Connection reset by peer) 2011-12-02T00:19:46 *** jmcarthur has joined #aichallenge 2011-12-02T00:21:59 lots of the people on fluxid's tcp server seem to use like 4s a turn now :\ 2011-12-02T00:22:06 games are taking forever 2011-12-02T00:25:59 i just realized my combat code is making my bot worse. I am losing to HunterBot in the default test :/ 2011-12-02T00:26:08 would that mean that skils on tcp server are overrated compared to skills on the aichallenge due to more time for their bots? 2011-12-02T00:26:19 skills* 2011-12-02T00:27:38 possibly 2011-12-02T00:27:44 people may just be testing stuff out 2011-12-02T00:28:34 http://xkcd.com/974/ :) 2011-12-02T00:28:58 my bot average tiem per move is 20ms to max of 200ms 2011-12-02T00:29:03 time* 2011-12-02T00:29:08 *** ChrisH_ has joined #aichallenge 2011-12-02T00:29:44 i sped mine up last night 2011-12-02T00:29:54 going through and working out what's taking the longest 2011-12-02T00:29:55 maybe i should implement some sort of search also with alpha-beta/negascout, is anyone using them? 2011-12-02T00:30:13 not sure how you'd use alpha beta for searching? 2011-12-02T00:30:26 i use a* while collecting food but just variations of bfs otherwise 2011-12-02T00:30:56 it's just mini-max with a smaller branching factor 2011-12-02T00:31:42 AntDroid: I use A-B pruning in combat calcs. 2011-12-02T00:32:11 problem is that the branching factor in this game is too big, maybe if you could prune some of the ants useless moves the branching factor would go sown 2011-12-02T00:32:26 yes 2011-12-02T00:32:28 *** TheLinker has quit IRC (Quit: Bye) 2011-12-02T00:32:37 *** jix has quit IRC (Ping timeout: 240 seconds) 2011-12-02T00:32:41 ChrisH: nice to know that. how deep do you score? how do you evaluate a node? 2011-12-02T00:32:52 anything you can do to reduce the positions you actually need to evaluate will help 2011-12-02T00:33:06 i'm not telling 2011-12-02T00:33:08 do you search? not score. sorry 2011-12-02T00:33:19 *** jix has joined #aichallenge 2011-12-02T00:33:28 gotta have some secrets at least until the finals 2011-12-02T00:33:37 haha yeah 2011-12-02T00:33:41 *** Antimony has joined #aichallenge 2011-12-02T00:33:59 but i don't search very deep 2011-12-02T00:34:22 because I haven't squeezed enough time out of it yet 2011-12-02T00:34:42 and I'm not sure how much more I can squeeze out 2011-12-02T00:34:47 ok. i'm thinking of using number of squares controlled in the map as score. maybe this way it will deprive the enemy with space. do you do iterative deepening? 2011-12-02T00:34:49 i changed my bot last night to only do closest path info for stuff like ants/hills etc. up to a max length 2011-12-02T00:35:04 i want my bot to be able to run on 300x300 2011-12-02T00:35:33 *** Lithosphere has joined #aichallenge 2011-12-02T00:35:37 *** Lithosphere has joined #aichallenge 2011-12-02T00:36:17 AntDroid: I only use mini-max for ants engaged in combat, and I score primarily based on how many ants each side has in the combat. 2011-12-02T00:36:37 anyone know how long set up time is on the server? 2011-12-02T00:36:43 3000ms 2011-12-02T00:36:46 cool 2011-12-02T00:36:49 i'm still under 1000 2011-12-02T00:36:53 but just wanted to make sure 2011-12-02T00:37:19 under 1000 for 300x300 that is 2011-12-02T00:37:31 i turn my grid into a graph during that turn 2011-12-02T00:37:48 I actually timed out on that in testing working on my last version, but it was an easy fix to cut my setup time in half, I had never optimized it until then. 2011-12-02T00:38:11 yikes, what the hell are you doing :p 2011-12-02T00:38:41 *** bmh has joined #aichallenge 2011-12-02T00:39:02 ChrisH: thanks. gives me some idea. my problem now is how to determine if an ant or group of ants is to be included for computation of combat . i currently only use ants within attack radius + 2 2011-12-02T00:39:05 I do a lot of caching of attackable, and visible cells for each map cell up front to avoid doing that over and over during each turn. 2011-12-02T00:39:20 AntDroid: that's pretty standard 2011-12-02T00:39:22 AntDroid: why +2? I was only going to do +1 2011-12-02T00:39:39 Oh, 85% from latest ai class homework, I was expecting somehting like 50% 2011-12-02T00:39:56 pairofdice: do they do automated grading? 2011-12-02T00:39:58 bmh: because both ants can move towards each other 2011-12-02T00:40:06 bmh: in case enemy ants go into your direction also. the ants in the +2 radius will get into you attack radius 2011-12-02T00:40:07 ChrisH_: duh! :) 2011-12-02T00:40:12 Yeah 2011-12-02T00:40:56 ChrisH: when do you do the caching? at the start of the game? 2011-12-02T00:40:59 i don't like people's +1 things 2011-12-02T00:41:02 it's not quite right 2011-12-02T00:41:15 or +2 2011-12-02T00:41:22 AntDroid: yes, in turn 0 after you get the board dimensions before the game begins 2011-12-02T00:41:29 eg. a%..b and a..%b can't land in battle next turn 2011-12-02T00:41:45 it's worth getting the border of attack and vision too imo 2011-12-02T00:41:51 the translation values that is 2011-12-02T00:42:05 antimatroid: sure, take that into account in your model? 2011-12-02T00:42:11 ChrisH_: or do you get the translation and attack locations for each location? 2011-12-02T00:42:24 bmh: yes i wont put those ants in the same battle partition 2011-12-02T00:42:25 antimatroid: yes 2011-12-02T00:42:37 ChrisH_: how large is that? :P 2011-12-02T00:42:42 and I update them to remove invalid ones as I see water 2011-12-02T00:42:48 yeah i do that 2011-12-02T00:43:02 well, for land neighbours 2011-12-02T00:43:04 ah, I'm not going to attempt to do a partition. I'm feeling clock pressure, but I'm going to take a shot at a sampling based approach 2011-12-02T00:43:25 i just went for fully sized partitions and quicker combat code 2011-12-02T00:43:27 at least this will be over before Christmas 2011-12-02T00:43:34 well it depends on map size, but even a 200x200 map is not huge 2011-12-02T00:44:03 i would have liked to bomb vision information with water blocking in some instances, but it's really slow 2011-12-02T00:44:09 but your way might make that possible 2011-12-02T00:45:07 I use that a lot for exploration planning 2011-12-02T00:45:25 i wanted to :P 2011-12-02T00:45:29 and now might :P 2011-12-02T00:46:25 so, 177 visible cells from each cell times 40k for a 200x200, comes to ~7M entries max, if you store 4byte ints, that's 28MB, 2011-12-02T00:46:55 i take it you aren't passing states around for your game tree? :P 2011-12-02T00:48:30 not copies of full board states, no way, there is one master copy of the cached indexes, and a point to that is passed everywhere. 2011-12-02T00:51:04 *** sf17k has joined #aichallenge 2011-12-02T00:56:48 *** Jak_o_Shadows has quit IRC (Remote host closed the connection) 2011-12-02T01:02:50 *** ltriant has quit IRC (Read error: No route to host) 2011-12-02T01:04:12 *** Palmik has joined #aichallenge 2011-12-02T01:06:11 *** ltriant has joined #aichallenge 2011-12-02T01:10:36 *** bhasker has joined #aichallenge 2011-12-02T01:10:40 evening 2011-12-02T01:13:19 evening 2011-12-02T01:13:26 how do we match that one up? 2011-12-02T01:13:32 must be almost midnight there? 2011-12-02T01:13:37 it's like 5 15pm here 2011-12-02T01:13:46 well its night here sorry! 2011-12-02T01:13:50 almost 10pm! 2011-12-02T01:14:06 but saying good night would be like signing off! 2011-12-02T01:14:09 so evening it is! 2011-12-02T01:14:15 bhasker: you aren't in a timezone..? 2011-12-02T01:14:36 i am a timezone 2011-12-02T01:14:48 in chuck norris style timezone is where chuck is 2011-12-02T01:16:29 Fun-fact: Kathmandu time is at UTC+5:45 because solar noon occurs in Kathmandu at UTC+5:43, so they rounded. 2011-12-02T01:16:42 %s/Kathmandu time/Nepali time/g 2011-12-02T01:16:48 indian tz is +5:30! 2011-12-02T01:18:09 *** moises has joined #aichallenge 2011-12-02T01:19:14 bhasker: you're back in india? 2011-12-02T01:19:20 nope 2011-12-02T01:19:32 in india it would be 11:45am right now! 2011-12-02T01:19:40 yeah i realised that after i said it :P 2011-12-02T01:25:48 so what is the name of your bot on TCP? 2011-12-02T01:26:19 not on there 2011-12-02T01:26:24 on official? 2011-12-02T01:27:11 nope 2011-12-02T01:27:35 i get distracted when i do either of those things :P 2011-12-02T01:36:36 newfoundland also has weird time 2011-12-02T01:36:37 I think 2011-12-02T01:37:07 *** Areks has joined #aichallenge 2011-12-02T01:38:20 *** ltriant has quit IRC (Quit: Get MacIrssi - http://www.sysctl.co.uk/projects/macirssi/) 2011-12-02T01:43:35 *** ChrisH_ has quit IRC (Quit: Page closed) 2011-12-02T01:44:32 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-12-02T01:47:59 *** analyst74 has quit IRC (Ping timeout: 245 seconds) 2011-12-02T01:48:14 *** analyst74 has joined #aichallenge 2011-12-02T02:07:49 *** UncleVasya has joined #aichallenge 2011-12-02T02:12:49 *** moises has quit IRC (Remote host closed the connection) 2011-12-02T02:13:23 *** jstrong has joined #aichallenge 2011-12-02T02:14:09 *** bmh has quit IRC (Quit: Page closed) 2011-12-02T02:14:23 *** roflmao has quit IRC (Ping timeout: 248 seconds) 2011-12-02T02:14:48 *** jstrong is now known as roflmao 2011-12-02T02:24:40 @seen Zaphus 2011-12-02T02:24:40 UncleVasya: Zaphus was last seen in #aichallenge 2 weeks, 2 days, 20 hours, 20 minutes, and 5 seconds ago: yeah, off by default is odd - but I must admit after a while I dont want to see it until something odd happens 2011-12-02T02:24:44 @seen Zaph 2011-12-02T02:24:44 UncleVasya: Zaph was last seen in #aichallenge 1 year, 11 weeks, 0 days, 9 hours, 36 minutes, and 9 seconds ago: If player 1 is based off games, player 2 could be based off time + elo and it solves both issues 2011-12-02T02:27:45 *** mleise has joined #aichallenge 2011-12-02T02:28:51 ahh the perennial topic, matchmaking 2011-12-02T02:34:41 *** analyst74 has quit IRC (Ping timeout: 244 seconds) 2011-12-02T02:39:42 *** UncleVasya has quit IRC () 2011-12-02T02:42:37 *** JorgeB has quit IRC (Quit: Textual IRC Client: http://www.textualapp.com/) 2011-12-02T02:42:58 *** JorgeB has joined #aichallenge 2011-12-02T03:02:32 *** roflmao has quit IRC (Quit: Leaving.) 2011-12-02T03:02:57 *** mravlja has joined #aichallenge 2011-12-02T03:06:25 !misc last --from thestinger 2011-12-02T03:06:33 @misc last --from thestinger 2011-12-02T03:06:33 thestinger: [03:06:25] !misc last --from thestinger 2011-12-02T03:07:05 @misc last --from thestinger --regexp battle. 2011-12-02T03:07:05 thestinger: Error: 'battle.' is not a valid regular expression. 2011-12-02T03:08:00 @misc last --from thestinger --with battle 2011-12-02T03:08:00 thestinger: [03:07:05] @misc last --from thestinger --regexp battle. 2011-12-02T03:09:57 *** sigh has joined #aichallenge 2011-12-02T03:10:32 *** aerique has joined #aichallenge 2011-12-02T03:14:34 *** kapoc has joined #aichallenge 2011-12-02T03:17:36 *** Ashoka has joined #aichallenge 2011-12-02T03:18:51 *** Kingpin13 has quit IRC (Quit: quit) 2011-12-02T03:19:07 *** emilk has quit IRC (Ping timeout: 244 seconds) 2011-12-02T03:19:12 *** emilk has joined #aichallenge 2011-12-02T03:24:59 *** jon1 has joined #aichallenge 2011-12-02T03:33:48 janzert: is this game resolution ok ? http://aichallenge.org/visualizer.php?game=153542&user=66 2011-12-02T03:39:31 *** ikaros has joined #aichallenge 2011-12-02T03:44:59 *** kapoc has left #aichallenge 2011-12-02T03:45:07 *** mj41 has joined #aichallenge 2011-12-02T03:46:31 Had ender sent any of the ants in the area to attack he would have won, so yes 2011-12-02T03:47:11 *** kire has joined #aichallenge 2011-12-02T03:48:24 I really don't understand why he sent the group of ants on the upper right of the map back past his own hill to attack instead of taking the short route that he had definitely already seen 2011-12-02T03:50:30 *** epicmonkey has joined #aichallenge 2011-12-02T03:52:34 yeah, game clearly would have ended the same way with more time anyway 2011-12-02T03:52:36 *** g0llum has joined #aichallenge 2011-12-02T03:53:51 *** kire has quit IRC (Remote host closed the connection) 2011-12-02T04:03:06 *** UncleVasya has joined #aichallenge 2011-12-02T04:05:52 *** mikewintermute has joined #aichallenge 2011-12-02T04:06:10 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-12-02T04:07:21 janzert : I meant the game was abruptly ended at just 431 turns. when one player was moving on and close towards the hill as well. what's the harm allowing games to continue a bit longer in cases like these... 2011-12-02T04:09:33 time is the harm 2011-12-02T04:09:44 the bot should have won the game way before then 2011-12-02T04:10:36 antimatroid : no. from turn 300 onwards, player1 was clearly marching towards the hill and almost reached there with a big army by 431. it would have been finished in abt 10-15 more turns 2011-12-02T04:24:50 *** mravlja has quit IRC (Ping timeout: 265 seconds) 2011-12-02T04:27:19 *** Garf has joined #aichallenge 2011-12-02T04:30:23 *** jcazevedo has quit IRC (Ping timeout: 248 seconds) 2011-12-02T04:36:51 * antimatroid has stripped his bot back to no strategy 2011-12-02T04:36:55 rewrite time :D 2011-12-02T04:39:39 *** chris__0076 has quit IRC (Ping timeout: 252 seconds) 2011-12-02T04:43:47 *** ikaros has joined #aichallenge 2011-12-02T04:47:03 *** sigh has quit IRC (Remote host closed the connection) 2011-12-02T05:02:55 *** epicmonkey has quit IRC (Ping timeout: 248 seconds) 2011-12-02T05:04:41 *** sigh has joined #aichallenge 2011-12-02T05:05:25 *** UncleVasya has joined #aichallenge 2011-12-02T05:06:59 amstan, this page http://aichallenge.org/starter_packages.php says that OCaml starter package has no hills support yet. But it is wrong, hills are implemented already. 2011-12-02T05:08:46 *** UncleVasya has quit IRC (Client Quit) 2011-12-02T05:12:44 *** grwip has joined #aichallenge 2011-12-02T05:14:57 *** antimatroid has quit IRC (Ping timeout: 260 seconds) 2011-12-02T05:16:12 *** antimatroid has joined #aichallenge 2011-12-02T05:22:44 *** sf17k has quit IRC (Quit: Page closed) 2011-12-02T05:23:51 *** Ashoka has quit IRC (Quit: Page closed) 2011-12-02T05:25:26 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-12-02T05:40:54 *** ikaros has quit IRC (Ping timeout: 252 seconds) 2011-12-02T05:45:49 *** Areks has quit IRC (Ping timeout: 240 seconds) 2011-12-02T05:50:24 *** AntDroid has quit IRC (Ping timeout: 265 seconds) 2011-12-02T05:53:37 *** Garf has quit IRC (Read error: Connection reset by peer) 2011-12-02T05:54:16 *** Garf has joined #aichallenge 2011-12-02T05:56:58 *** mviel_ has joined #aichallenge 2011-12-02T06:00:02 *** mviel__ has quit IRC (Ping timeout: 255 seconds) 2011-12-02T06:01:34 *** Areks has joined #aichallenge 2011-12-02T06:06:45 *** Sir_Ragnarok has quit IRC (Ping timeout: 260 seconds) 2011-12-02T06:08:39 *** Sir_Ragnarok has joined #aichallenge 2011-12-02T06:11:06 *** lawmax has joined #aichallenge 2011-12-02T06:12:02 *** mikewintermute has quit IRC (Quit: mikewintermute) 2011-12-02T06:14:41 *** b0rder has quit IRC (Ping timeout: 260 seconds) 2011-12-02T06:20:28 *** ikaros has joined #aichallenge 2011-12-02T06:35:56 *** Conorach has joined #aichallenge 2011-12-02T06:37:36 note to self, the best time to start the timer is when you read "turn" 2011-12-02T06:51:48 *** Chris_0076 has joined #aichallenge 2011-12-02T07:00:04 *** raemde has joined #aichallenge 2011-12-02T07:02:23 *** raemde_ has quit IRC (Ping timeout: 248 seconds) 2011-12-02T07:21:29 *** g0llum has joined #aichallenge 2011-12-02T07:22:10 *** Nullpointer_ has joined #aichallenge 2011-12-02T07:30:10 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-12-02T07:33:08 *** kire has joined #aichallenge 2011-12-02T07:40:14 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-12-02T07:45:16 *** u_ has joined #aichallenge 2011-12-02T07:47:01 =] 2011-12-02T07:51:29 *** liberforce has joined #aichallenge 2011-12-02T07:55:56 *** foRei has joined #aichallenge 2011-12-02T07:56:44 *** delt0r_ has quit IRC (Ping timeout: 245 seconds) 2011-12-02T08:09:19 *** delt0r_ has joined #aichallenge 2011-12-02T08:12:13 friggin americans 2011-12-02T08:12:20 5 of them in the top 10 2011-12-02T08:23:13 *** sigh has quit IRC (Remote host closed the connection) 2011-12-02T08:23:39 *** mikewintermute has joined #aichallenge 2011-12-02T08:36:55 *** QuirionPT has joined #aichallenge 2011-12-02T08:39:36 *** mviel__ has joined #aichallenge 2011-12-02T08:43:04 *** mviel_ has quit IRC (Ping timeout: 244 seconds) 2011-12-02T08:46:50 *** Antimony has joined #aichallenge 2011-12-02T08:56:43 *** george has joined #aichallenge 2011-12-02T08:57:01 hi, about map symmetry... 2011-12-02T08:58:09 will/is the python code in branch: amstam_mapgen /ants/mapgen/ be used for the final contest (with different seeds of course) ? 2011-12-02T08:58:55 (i'm trying to take advantage of symmetries, and the code seems to be the (only/best) documentation ... 2011-12-02T08:59:29 ... so i wanted to know whether the code is now fixed? ) 2011-12-02T08:59:39 ie will not change 2011-12-02T08:59:48 just assume any symmetry cna hpapen 2011-12-02T09:00:08 Anti - 2011-12-02T09:00:23 for all i know they might use Penrose Tilings or something :/ 2011-12-02T09:00:41 i am not going to write code for every possible symmetry if i can help it 2011-12-02T09:02:45 i was reading the webpage, but the info there wasn't definitive - mentioned 'blocks' for eg the food. but some of the translation maps are not like that. 2011-12-02T09:03:21 which is when amstam suggested i read the python code 2011-12-02T09:04:14 there's only 8 possible symmetries 2011-12-02T09:07:30 hm. not sure that helps me. but thanks for the reply. 2011-12-02T09:07:40 eg 2-4-8 fold rotations 2011-12-02T09:08:00 are presumably possible on grid? 2011-12-02T09:08:22 *** Nullpointer_ has quit IRC (Quit: Page closed) 2011-12-02T09:08:52 but i'f they are not implemented, then i want to know that - allows me to eliminate possibilities earlier in the game, and so know the map sooner 2011-12-02T09:08:52 etc 2011-12-02T09:08:56 isn't it that from the perspective of each player the map has to look exactly the same? 2011-12-02T09:09:07 yes, sure 2011-12-02T09:09:19 so there are only 4 possibilities, right? 2011-12-02T09:10:08 well. eg do we have symmetry lines running between squares (yes, it seems) 2011-12-02T09:10:18 and also through squares? i dont know 2011-12-02T09:10:28 ... 2011-12-02T09:10:44 i'm just trying to find what the rules are. 2011-12-02T09:10:58 the webpage gives info that seems to be innaccurate 2011-12-02T09:11:13 which is why i read code and ask if code is what organisers will use. 2011-12-02T09:13:22 anything that passes the symmetry test in the engine is acceptable 2011-12-02T09:13:28 you can't assume the mapgens will stay the same 2011-12-02T09:16:15 hi Antimony. thanks. the 'engine' is mapgen.py? (btw, am i right in assuming your are an organiser?) 2011-12-02T09:16:40 no 2011-12-02T09:16:46 I'm not an organizer 2011-12-02T09:17:01 but that's what they said before 2011-12-02T09:20:55 ok, thanks. 2011-12-02T09:21:39 i'll try and ask amstam when he's about, to be sure. 2011-12-02T09:46:18 *** NightExcessive has joined #aichallenge 2011-12-02T09:59:49 *** Areks has quit IRC (Ping timeout: 240 seconds) 2011-12-02T10:00:37 what is the point of specifying rows and columns in the map file if the game validates it? 2011-12-02T10:00:50 it should just use what it sees 2011-12-02T10:03:48 *** NightExcessive has quit IRC (Remote host closed the connection) 2011-12-02T10:04:25 *** Harpyon has joined #aichallenge 2011-12-02T10:09:07 probably just convenience for the windows people who were born without basic text processing tools 2011-12-02T10:10:13 I literally make the map and run it to get the numbers it wants since it tells you lol 2011-12-02T10:11:05 I have so many maps for testing now, I should put a tcp server up with these random scenarios, they would be interesting to see who can win 2011-12-02T10:11:22 like 5v5v5v5 scenarios with no food 2011-12-02T10:11:39 chokepoint battles, all the fun stuff 2011-12-02T10:15:17 *** b0rder has joined #aichallenge 2011-12-02T10:25:38 *** Antimony has quit IRC (Ping timeout: 244 seconds) 2011-12-02T10:30:13 *** aerique has quit IRC (Quit: ...) 2011-12-02T10:43:48 *** praveen_ has joined #aichallenge 2011-12-02T10:48:59 *** Anilm3 has joined #aichallenge 2011-12-02T10:49:33 Hi guys 2011-12-02T10:50:00 *** TheLinker has joined #aichallenge 2011-12-02T10:51:48 *** avdg has joined #aichallenge 2011-12-02T10:56:07 *** avdg has quit IRC (Ping timeout: 244 seconds) 2011-12-02T10:59:04 *** Antimony has joined #aichallenge 2011-12-02T11:12:36 *** tekking has joined #aichallenge 2011-12-02T11:16:48 Hi 2011-12-02T11:18:09 *** QuirionPT has quit IRC (Quit: Leaving) 2011-12-02T11:21:14 Hi 2011-12-02T11:23:04 *** avdg has joined #aichallenge 2011-12-02T11:24:49 hi 2011-12-02T11:30:59 *** dvladim has joined #aichallenge 2011-12-02T11:32:09 *** tekking has quit IRC (Quit: Page closed) 2011-12-02T11:35:45 *** mj41 has quit IRC (Quit: Leaving) 2011-12-02T11:57:17 *** jstrong has joined #aichallenge 2011-12-02T11:58:24 *** Anilm3 has quit IRC (Ping timeout: 245 seconds) 2011-12-02T11:59:59 *** Antimony has quit IRC (Ping timeout: 260 seconds) 2011-12-02T12:00:16 *** Antimony has joined #aichallenge 2011-12-02T12:19:10 *** jstrong has quit IRC (Quit: Leaving.) 2011-12-02T12:19:29 *** UncleVasya has joined #aichallenge 2011-12-02T12:20:56 *** Murashka has joined #aichallenge 2011-12-02T12:29:08 *** Murashka has quit IRC (Quit: Page closed) 2011-12-02T12:31:27 *** b0rder has quit IRC (Ping timeout: 248 seconds) 2011-12-02T12:32:07 *** UncleVasya has quit IRC (Ping timeout: 244 seconds) 2011-12-02T12:32:39 *** iglo has joined #aichallenge 2011-12-02T12:37:31 *** kire has quit IRC (Remote host closed the connection) 2011-12-02T12:38:46 *** mviel__ has quit IRC (Quit: Leaving) 2011-12-02T12:41:35 *** dr- has quit IRC (Ping timeout: 248 seconds) 2011-12-02T12:42:44 *** dr- has joined #aichallenge 2011-12-02T12:46:00 *** b0rder has joined #aichallenge 2011-12-02T12:46:46 *** b0rder has quit IRC (Max SendQ exceeded) 2011-12-02T12:47:30 *** smiley1983 has joined #aichallenge 2011-12-02T12:48:01 *** TheLinker has quit IRC (Remote host closed the connection) 2011-12-02T12:54:51 *** Antimony has quit IRC (Ping timeout: 244 seconds) 2011-12-02T12:55:11 *** Murashka has joined #aichallenge 2011-12-02T12:59:27 *** smiley1983 has quit IRC (Quit: leaving) 2011-12-02T13:26:05 *** Antimony has joined #aichallenge 2011-12-02T13:26:34 *** Murashka has quit IRC (Quit: Page closed) 2011-12-02T13:27:37 *** jon1 has quit IRC (Quit: Leaving.) 2011-12-02T13:28:02 *** goffrie_ has joined #aichallenge 2011-12-02T13:28:04 *** goffrie has quit IRC (Disconnected by services) 2011-12-02T13:28:05 *** goffrie_ is now known as goffrie 2011-12-02T13:31:05 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-12-02T13:32:35 *** mikewintermute has quit IRC (Quit: mikewintermute) 2011-12-02T13:39:22 *** Antimony has quit IRC (Ping timeout: 260 seconds) 2011-12-02T13:43:40 *** liberforce has left #aichallenge 2011-12-02T13:46:39 *** dvladim has quit IRC (Ping timeout: 248 seconds) 2011-12-02T14:02:41 *** roflmao has joined #aichallenge 2011-12-02T14:02:53 Next game should be within 975 minutes. 2011-12-02T14:02:55 haha 2011-12-02T14:09:01 next game should be before the end of the contest 2011-12-02T14:09:35 *** Ox86 has quit IRC (Read error: Connection reset by peer) 2011-12-02T14:16:53 hmm, i better resubmit now then 2011-12-02T14:18:50 *** bretep has joined #aichallenge 2011-12-02T14:21:40 *** cwc has joined #aichallenge 2011-12-02T14:21:42 *** roflmao has quit IRC (Quit: Leaving.) 2011-12-02T14:26:42 *** jstemmer has quit IRC (Quit: leaving) 2011-12-02T14:29:13 Next game should be within 31257 minutes. 2011-12-02T14:34:55 If you upload a new version of your bot, it will always rank around 7000 2011-12-02T14:35:07 no more matches :) 2011-12-02T14:35:49 *** ikaros has joined #aichallenge 2011-12-02T14:35:56 *** praveen_ has quit IRC (Quit: Page closed) 2011-12-02T14:43:01 *** george has quit IRC (Ping timeout: 265 seconds) 2011-12-02T14:48:34 *** kire has joined #aichallenge 2011-12-02T14:52:11 *** Conorach has quit IRC (Ping timeout: 252 seconds) 2011-12-02T14:52:24 *** Conorach has joined #aichallenge 2011-12-02T15:01:05 *** Accoun has quit IRC () 2011-12-02T15:01:58 *** liberforce has joined #aichallenge 2011-12-02T15:03:43 *** b0rder has joined #aichallenge 2011-12-02T15:05:04 *** b0rder has joined #aichallenge 2011-12-02T15:05:44 *** Antimony has joined #aichallenge 2011-12-02T15:08:16 *** b0rder has joined #aichallenge 2011-12-02T15:10:32 *** Redgis has joined #aichallenge 2011-12-02T15:17:09 *** b0rder has quit IRC (Ping timeout: 245 seconds) 2011-12-02T15:18:49 *** TheLinker has joined #aichallenge 2011-12-02T15:21:49 *** Accoun has joined #aichallenge 2011-12-02T15:30:46 *** bergmark has joined #aichallenge 2011-12-02T15:32:12 *** Sion has joined #aichallenge 2011-12-02T15:32:42 Hi 2011-12-02T15:33:58 hello? 2011-12-02T15:36:33 Hi 2011-12-02T15:36:34 *** Kingpin13 has joined #aichallenge 2011-12-02T15:39:53 *** McLeopold has quit IRC (Quit: Leaving.) 2011-12-02T15:42:27 xathis is doing well with version 1 2011-12-02T15:47:11 *** amstan has joined #aichallenge 2011-12-02T15:47:11 *** ChanServ sets mode: +o amstan 2011-12-02T15:47:31 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-12-02T15:51:19 *** liberforce has quit IRC (Ping timeout: 245 seconds) 2011-12-02T15:54:14 *** epicmonkey has joined #aichallenge 2011-12-02T15:54:34 yep 2011-12-02T15:54:44 I wonder how xathis got so good 2011-12-02T15:55:38 beta 2011-12-02T15:57:19 *** delt0r_ has quit IRC (Ping timeout: 248 seconds) 2011-12-02T15:59:44 *** kioko has joined #aichallenge 2011-12-02T15:59:57 *** nomadpenguin has joined #aichallenge 2011-12-02T16:00:38 O HAI 2011-12-02T16:00:41 whaaaup dudes 2011-12-02T16:00:55 *** nomadpenguin has quit IRC (Client Quit) 2011-12-02T16:01:27 nobody 2011-12-02T16:01:35 Im really interested in this. 2011-12-02T16:04:15 *** mikewintermute has joined #aichallenge 2011-12-02T16:06:59 *** NightExcessive has joined #aichallenge 2011-12-02T16:10:01 *** bretep has quit IRC (Ping timeout: 265 seconds) 2011-12-02T16:10:35 *** delt0r_ has joined #aichallenge 2011-12-02T16:13:24 *** kioko has quit IRC (Ping timeout: 265 seconds) 2011-12-02T16:18:07 *** bqf has quit IRC (Ping timeout: 248 seconds) 2011-12-02T16:18:07 *** murr4y has quit IRC (Ping timeout: 248 seconds) 2011-12-02T16:18:15 *** Harpyon has quit IRC (Read error: Connection reset by peer) 2011-12-02T16:18:39 *** kurti has quit IRC (Ping timeout: 248 seconds) 2011-12-02T16:18:51 *** Harpyon has joined #aichallenge 2011-12-02T16:19:11 *** mleise has quit IRC (Ping timeout: 248 seconds) 2011-12-02T16:19:16 *** bqf has joined #aichallenge 2011-12-02T16:19:21 *** kurti has joined #aichallenge 2011-12-02T16:19:27 *** murr4y has joined #aichallenge 2011-12-02T16:19:27 *** tcoppi has quit IRC (Ping timeout: 244 seconds) 2011-12-02T16:21:27 *** tcoppi has joined #aichallenge 2011-12-02T16:24:40 *** mleise has joined #aichallenge 2011-12-02T16:26:29 *** NightExcessive has left #aichallenge ("Leaving") 2011-12-02T16:27:44 oh god rewrites are fun :) 2011-12-02T16:31:56 I wonder why my ants jerk back and forth. http://aichallenge.org/visualizer.php?game=155026&user=3845 2011-12-02T16:33:23 haha all your ants just ran from one enemy ant 2011-12-02T16:33:37 when? 2011-12-02T16:34:14 around turn 390 2011-12-02T16:34:33 400-410 ish 2011-12-02T16:35:17 lol 2011-12-02T16:35:21 it does look like that 2011-12-02T16:35:28 I wonder why my ants were doing what they did 2011-12-02T16:35:40 it sounds like seomthing I need to figure out and fix 2011-12-02T16:36:22 *** sigh has joined #aichallenge 2011-12-02T16:38:05 *** ztfw has joined #aichallenge 2011-12-02T16:40:38 *** Antimony has quit IRC (Ping timeout: 244 seconds) 2011-12-02T16:41:50 *** saptak has joined #aichallenge 2011-12-02T16:42:37 *** kire has quit IRC (Quit: Leaving) 2011-12-02T16:50:21 *** Garf has quit IRC (Read error: Connection reset by peer) 2011-12-02T16:50:28 *** Harpyon has quit IRC (Read error: Connection reset by peer) 2011-12-02T16:51:02 *** Harpyon has joined #aichallenge 2011-12-02T16:51:06 *** Sion has quit IRC (Ping timeout: 265 seconds) 2011-12-02T16:51:45 *** cowbandit has joined #aichallenge 2011-12-02T16:52:51 hey does anyone have any advice for efficiently exploring the map? 2011-12-02T16:53:17 currently i'm making a list of all unvisible tiles and removing them when they become visible 2011-12-02T16:53:32 and once that list is depleted i just reset the list 2011-12-02T16:53:39 however i feel that I don't get enough spread that way 2011-12-02T16:56:08 *** Palmik has quit IRC (Remote host closed the connection) 2011-12-02T17:02:25 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-12-02T17:05:45 *** Antimony has joined #aichallenge 2011-12-02T17:07:27 *** amstan has joined #aichallenge 2011-12-02T17:07:27 *** ChanServ sets mode: +o amstan 2011-12-02T17:15:45 *** Hexren has joined #aichallenge 2011-12-02T17:16:44 *** Hexren has left #aichallenge 2011-12-02T17:24:52 *** jacob_strauss has joined #aichallenge 2011-12-02T17:26:24 *** Redgis has quit IRC (Quit: ... mains libres) 2011-12-02T17:27:02 *** roflmao has joined #aichallenge 2011-12-02T17:30:32 *** jacob_strauss has quit IRC (Quit: Page closed) 2011-12-02T17:30:46 *** keith_ has joined #aichallenge 2011-12-02T17:32:59 *** TwistedLogic has joined #aichallenge 2011-12-02T17:33:40 *** Cyndre has joined #aichallenge 2011-12-02T17:38:45 *** Jak_o_Shadows has joined #aichallenge 2011-12-02T17:42:06 *** katebus has joined #aichallenge 2011-12-02T17:43:26 *** b0rder has joined #aichallenge 2011-12-02T17:45:33 *** mleise has quit IRC (Read error: Operation timed out) 2011-12-02T17:45:34 better to keep an array of the map 2011-12-02T17:45:40 cowbandit: ^^ 2011-12-02T17:45:55 increment all unvisible tiles per turn, 0 out visibles 2011-12-02T17:46:05 weight locations to explore based on the map 2011-12-02T17:46:13 ahh i see 2011-12-02T17:46:33 hmm i like that solution 2011-12-02T17:46:50 now it will just be implementing it.. T_T 2011-12-02T17:47:05 makes it so you want to explore tiles you have seen least recent 2011-12-02T17:47:20 wait so should i increment a tile 2011-12-02T17:47:26 each time it's seen by a different ant? 2011-12-02T17:47:30 no 2011-12-02T17:47:35 increment when not seen 2011-12-02T17:47:41 set to 0 when seen 2011-12-02T17:47:44 so every turn 2011-12-02T17:47:47 ohh 2011-12-02T17:47:58 so say you haven't seen a tile for 100 turns 2011-12-02T17:48:05 it would be weighted by 100 2011-12-02T17:48:15 ok i see what you're saying 2011-12-02T17:50:39 if you could not see an enemy hill there 100 turns ago, why go back ? 2011-12-02T17:50:52 g0llum: food 2011-12-02T17:50:56 *** bergmark has quit IRC (Ping timeout: 252 seconds) 2011-12-02T17:51:45 g0llum: but this increments from the start of the game, so odds are if something is the highest value, you haven't seen it yet 2011-12-02T17:52:06 otherwise you have seen the whole map and should be killing the hill 2011-12-02T17:53:19 ok, i thought you wood be separating exploring and food/enemy hunt 2011-12-02T17:53:24 *** nle85 has joined #aichallenge 2011-12-02T17:53:39 g0llum: well I don't because of the way my AI works 2011-12-02T17:54:41 I am modularizing everything now since I will probably be switching some parts to use a best first search using my diffusion map 2011-12-02T17:55:12 I already have the data, why not use it :-/ 2011-12-02T18:03:41 *** b0rder has quit IRC (Max SendQ exceeded) 2011-12-02T18:03:49 *** Jak_o_Shadows has quit IRC (Read error: Connection reset by peer) 2011-12-02T18:04:34 *** b0rder has joined #aichallenge 2011-12-02T18:05:12 let me clarify something... do your ants go for the highest or for the lowest value of the tiles? 2011-12-02T18:05:21 highest 2011-12-02T18:05:43 *** Jak_o_Shadows has joined #aichallenge 2011-12-02T18:05:57 sure, nice idea 2011-12-02T18:06:13 so there are many ideas for well leveled exploration 2011-12-02T18:06:26 now i have to pick one and implement 2011-12-02T18:07:50 my bot does not explore yet... well it does some random stuff, but my ants gets blocked by my other ants 2011-12-02T18:08:08 do you have problems with ants blocking each other? 2011-12-02T18:08:45 no, but I don't use any real navigation yet 2011-12-02T18:08:48 just an influence map 2011-12-02T18:09:02 I've read about flocks, and so on... but it does not fit my solution, at least not yet, maybe I should drop that starter bot and write it from scratch 2011-12-02T18:09:22 ohh, the influence, I see that this may be smarter 2011-12-02T18:09:35 I don't think any of the top maps use it 2011-12-02T18:09:41 at least not exclusively 2011-12-02T18:09:53 *** nle85 has quit IRC (Ping timeout: 265 seconds) 2011-12-02T18:09:54 *** Cyndre has quit IRC (Quit: Leaving) 2011-12-02T18:09:57 I am gonna transition to a hybrid approach most likely 2011-12-02T18:09:58 I am still thinking 2011-12-02T18:10:09 maybe i should stop and just try to do something 2011-12-02T18:10:47 I keep a list of where each ant will end up, then if the policy says to move to an inhabitted space I chose a different direction (or don't move if no possible choice). works alright, very rare do ants bump into each other 2011-12-02T18:11:33 i meant that i have a few ants, lets say 4 blocking 10 other ants - it works like a wall 2011-12-02T18:11:46 i do redirect too, but then it looks like a chaos 2011-12-02T18:12:03 the question is why are those 4 ants not moving 2011-12-02T18:12:08 yeah 2011-12-02T18:12:31 because they want to go into some other direction, and the decision changes with every turn 2011-12-02T18:13:00 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-12-02T18:13:03 ya, you need a relatively stable policy map 2011-12-02T18:13:06 I have that problem occasionally, but more ants coming in the area usually fixes the problem rather than worsen it 2011-12-02T18:14:31 the back and forth thing that isd 2011-12-02T18:14:32 is 2011-12-02T18:15:10 Silly skuto! Having more atns doesn't help if you don't defend! http://ants.fluxid.pl/replay.9875 2011-12-02T18:15:59 *** epicmonkey has quit IRC (Ping timeout: 248 seconds) 2011-12-02T18:17:27 I get back and forth action sometimes, really bad on some maps. probably because of two many conflicting goals 2011-12-02T18:20:00 ug, I need combat badly 2011-12-02T18:20:09 yes, deciding between many local goals sometimes makes the bot stupid, the cost changes with every turn, I wonder if we can easily try to compute some global cost function - for the whole map 2011-12-02T18:20:46 why do you have so many unstable goals? 2011-12-02T18:20:52 *** IU-YO has joined #aichallenge 2011-12-02T18:21:03 hows it going 2011-12-02T18:21:39 Antimony: It's not so much goals are unstable, but turn by turn goals may get different weights based on map state 2011-12-02T18:22:02 it's not local goals for me I don't think. I think it is when I know about two hills, maybe I need to focus on one target at a time 2011-12-02T18:22:25 I have that problem when I send many ants to the nearest something, and that something can appear or disappear frequently 2011-12-02T18:22:28 use weights for each kind of goal 2011-12-02T18:22:34 such as exploration points and enemy ants 2011-12-02T18:23:03 Fluxid: this is where my influence map falls short, I blend my goals 2011-12-02T18:23:09 but I have solutions, I just need to find the right balance 2011-12-02T18:23:51 in my case it is because I just started with the starter package from java tutorial, and there the ants select the targets, while preferably the ant should be assigned to targets, 2011-12-02T18:24:19 the resulting solution is suboptimal and unstable, I may have much to improve 2011-12-02T18:25:29 not sure about the java starter package, but sounds like you are doing too much fine work. I think the best solution is to define global policies that apply to all ants (or subsections of ants) 2011-12-02T18:26:11 *** iglo has quit IRC (Remote host closed the connection) 2011-12-02T18:27:00 there we go successfully modularized 2011-12-02T18:27:05 now I can start doing work 2011-12-02T18:28:24 @cwc: probably you are right, 2011-12-02T18:28:25 katebus: You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2011-12-02T18:28:50 *** Antimony_ has joined #aichallenge 2011-12-02T18:29:25 *** Antimony has quit IRC (Ping timeout: 240 seconds) 2011-12-02T18:29:30 *** Antimony_ is now known as Antimony 2011-12-02T18:33:34 *** pguillory has joined #aichallenge 2011-12-02T18:33:39 *** amstan has joined #aichallenge 2011-12-02T18:33:39 *** ChanServ sets mode: +o amstan 2011-12-02T18:33:47 *** analyst74 has joined #aichallenge 2011-12-02T18:55:27 ugh these new small maps are so frustrating 2011-12-02T18:55:37 they seem to favor bots that rush an attack 2011-12-02T18:56:21 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-12-02T18:58:20 *** b0rder has quit IRC (Quit: 离开) 2011-12-02T18:59:55 *** Cali has joined #aichallenge 2011-12-02T19:00:24 why do the servers allow for such a limited amount of games to occur on average? 2011-12-02T19:01:35 ...games take a lot of CPU time 2011-12-02T19:06:13 Yes, but google could find a few more servers :) 2011-12-02T19:06:29 "google" 2011-12-02T19:06:38 Next game should be within 964 minutes. 2011-12-02T19:06:45 @cali - just try one of the tcp servers 2011-12-02T19:06:46 katebus: User error, it's not my fault. 2011-12-02T19:07:27 I have just uploaded my bot, and I see that ... next game in 1200 minutes 2011-12-02T19:07:37 yeah... 2011-12-02T19:09:15 well, The tcp server will be the solution to test in battle 2011-12-02T19:12:59 enough for today it's 1AM here, bye 2011-12-02T19:15:38 *** Antimony has quit IRC (Ping timeout: 255 seconds) 2011-12-02T19:18:10 *** katebus has left #aichallenge 2011-12-02T19:21:36 *** Anilm3 has joined #aichallenge 2011-12-02T19:21:52 Hi 2011-12-02T19:27:46 what's this guy's problem? http://forums.aichallenge.org/viewtopic.php?f=25&t=1989#p12387 2011-12-02T19:28:08 he sucks at cross-platform coding, yet he's calling us unprofessional 2011-12-02T19:28:15 *** rofer has quit IRC (Ping timeout: 260 seconds) 2011-12-02T19:31:37 *** rofer has joined #aichallenge 2011-12-02T19:33:10 aichallenge: Alexandru Stan epsilon * r50bfa8c / ants/dist/starter_bots/vb/Ants.vb : Turn time should not be hardcoded, there's a reason we have parameters in the protocol. Fixed. - http://git.io/cSbIiw 2011-12-02T19:35:17 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-12-02T19:35:53 *** analyst74 has quit IRC (Ping timeout: 244 seconds) 2011-12-02T19:39:23 *** nle85 has joined #aichallenge 2011-12-02T19:42:17 *** Anilm3 has quit IRC (Quit: Lost terminal) 2011-12-02T19:42:58 *** Antimony has joined #aichallenge 2011-12-02T19:45:18 *** Jak_o_Shadows has quit IRC (Remote host closed the connection) 2011-12-02T19:45:25 *** foRei has quit IRC (Read error: Connection reset by peer) 2011-12-02T19:46:37 *** matthewd has left #aichallenge 2011-12-02T19:51:55 *** ztfw has quit IRC (Read error: Connection reset by peer) 2011-12-02T19:52:31 *** ztfw has joined #aichallenge 2011-12-02T19:55:23 aichallenge: McLeopold epsilon * r1383e68 / (ants/mapgen/cell_maze.py ants/mapgen/map.py): improvements to mazing and hill distance checks - http://git.io/enAlvg 2011-12-02T19:55:46 janzert: so what's going on then? with those workers? 2011-12-02T19:56:25 we're at the exchanging emails point :) 2011-12-02T19:56:34 advanced, lol 2011-12-02T19:56:38 *** Harpyon has quit IRC (Quit: Textual IRC Client: http://www.textualapp.com/) 2011-12-02T19:56:55 btw, the current processor speed is generally 2.66 not 2 ghz, not that it really matters 2011-12-02T20:01:58 *** IU-YO has quit IRC (Quit: Page closed) 2011-12-02T20:03:27 aichallenge: McLeopold epsilon * r0660abe / (64 files): add a few maps - http://git.io/UmdTKw 2011-12-02T20:04:55 *** keith_ has quit IRC (Ping timeout: 265 seconds) 2011-12-02T20:11:48 *** keith_ has joined #aichallenge 2011-12-02T20:24:43 is there a way to watch tcp games live 2011-12-02T20:24:49 like from the players view at least 2011-12-02T20:25:02 short of tailing the logs 2011-12-02T20:28:00 is there any way to retrieve the chat messages from the last 24 hours or the last 7 days? i'm not online all the time and i may miss some interesting conversations 2011-12-02T20:29:01 topic 2011-12-02T20:29:10 @topic 2011-12-02T20:29:10 pairofdice: AI Challenge (sponsored by Google): http://aichallenge.org || Channel Logs: http://contestbot.aichallenge.org || Code Repo: http://github.com/aichallenge/aichallenge || About the game (Ants): http://aichallenge.org/problem_description.php || Submission deadline: December 18 23:59 EST 2011-12-02T20:32:17 @later tell McLeopold are the old maze and random walk generators close to being converted over? we're really really heavily weighted toward cell mazes now 2011-12-02T20:32:17 janzert: I think that worked... 2011-12-02T20:32:39 *** Zaphus has joined #aichallenge 2011-12-02T20:33:58 *** grwip has quit IRC (Remote host closed the connection) 2011-12-02T20:35:44 *** cowbandit has quit IRC (Quit: Page closed) 2011-12-02T20:40:06 "Next game should be within 1371 minutes." Guess I'll come back tomorrow then ! (Thats ~23 hrs) 2011-12-02T20:42:08 *** nle85 has quit IRC (Ping timeout: 265 seconds) 2011-12-02T20:46:09 *** bugnuts has quit IRC (Read error: Connection reset by peer) 2011-12-02T20:46:33 *** bugnuts has joined #aichallenge 2011-12-02T20:48:00 *** sigh has quit IRC (Read error: Connection reset by peer) 2011-12-02T20:48:35 *** sigh has joined #aichallenge 2011-12-02T21:02:12 *** Antimony has quit IRC (Ping timeout: 260 seconds) 2011-12-02T21:08:14 *** Cali has quit IRC (Ping timeout: 265 seconds) 2011-12-02T21:10:56 *** roflmao has quit IRC (Quit: Leaving.) 2011-12-02T21:23:46 *** jstrong has joined #aichallenge 2011-12-02T21:26:39 *** ChrisH_ has joined #aichallenge 2011-12-02T21:28:00 *** Conorach has quit IRC (Ping timeout: 244 seconds) 2011-12-02T21:30:45 amstan: meh, I bet if he had all the warnings turned up he would have spotted that anyway 2011-12-02T21:31:43 trying it myself atm 2011-12-02T21:35:47 thestinger: he didn't even use the same os/compiler 2011-12-02T21:36:39 *** ChrisH_ has quit IRC (Quit: Page closed) 2011-12-02T21:39:27 test.c:9:11: warning: implicit conversion changes signedness: 'long' to 'unsigned long' [-Wsign-conversion] 2011-12-02T21:39:28 return (tm.tv_sec * 1000) + (tm.tv_usec / 1000); 2011-12-02T21:39:51 that's with clang 2011-12-02T21:39:53 test.c:9:30: warning: conversion to ‘long unsigned int’ from ‘long int’ may change the sign of the result [-Wsign-conversion] 2011-12-02T21:39:55 gcc 2011-12-02T21:40:02 just need to turn on proper warnings if you expect them 2011-12-02T21:40:36 hmmm I still don't really get what his problems were 2011-12-02T21:41:01 his biggest problem seems to be he is a pompous ass 2011-12-02T21:43:11 *** Antimony has joined #aichallenge 2011-12-02T21:47:58 he needs to be told to just RTFM next time 2011-12-02T21:48:07 http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/sys_time.h.html 2011-12-02T21:48:18 http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/sys_types.h.html 2011-12-02T21:48:21 time_t can be a float. 2011-12-02T21:49:12 you're supposed to use uintmax_t if you want portability 2011-12-02T21:56:50 *** ericlavigne has joined #aichallenge 2011-12-02T21:56:51 I think he went over the top, but his frustration is understandable - he's stuck and there really is no way for anyone to help tell him what's going wrong, I imagine that's very frustrating and he vented. 2011-12-02T21:57:26 *** besh has joined #aichallenge 2011-12-02T21:59:24 spotted what? 2011-12-02T22:00:17 *** ztfw has quit IRC (Remote host closed the connection) 2011-12-02T22:00:57 *** besh has quit IRC (Client Quit) 2011-12-02T22:02:07 *** dom7b5 has quit IRC (Ping timeout: 248 seconds) 2011-12-02T22:02:21 *** Zaphus has quit IRC (Quit: Page closed) 2011-12-02T22:11:04 *** pguillory has quit IRC (Ping timeout: 265 seconds) 2011-12-02T22:12:40 *** Extrarius has joined #aichallenge 2011-12-02T22:16:40 *** afw314 has joined #aichallenge 2011-12-02T22:18:27 *** jstrong has quit IRC (Quit: Leaving.) 2011-12-02T22:19:47 *** yoden has joined #aichallenge 2011-12-02T22:22:18 amstan: Have you spent any time playing around with my generator? 2011-12-02T22:22:42 Extrarius: nope, sorry 2011-12-02T22:23:33 not a problem, just curious =-) I've been moving so I have't had time to work on an entry or my generator 2011-12-02T22:27:49 http://forums.aichallenge.org/viewtopic.php?f=25&t=1989&p=12398#p12398 hopefully that wasn't too harsh :) 2011-12-02T22:35:29 I like how the POSIX docs say "End of informative text." at the bottom as if I couldn't figure that out myself xD 2011-12-02T22:36:58 *** besh has joined #aichallenge 2011-12-02T22:37:50 *** roflmao has joined #aichallenge 2011-12-02T22:38:32 i slacked a couple of days and hit rock bottom on fluxid :( 2011-12-02T22:38:56 join the club 2011-12-02T22:39:06 *** keith_ has quit IRC (Ping timeout: 265 seconds) 2011-12-02T22:39:17 which club 2011-12-02T22:40:10 at this rate finishing top 100 is gonna be tough 2011-12-02T22:45:29 http://aichallenge.org/profile.php?user=2255 my bot climbed up again in 2 days :) 2011-12-02T22:46:16 mu is 88.45 so it's definitely better than last time (sigma just needs to drop from 3) 2011-12-02T22:52:51 @besh: the club of now-tcp-losers 2011-12-02T22:52:51 :P 2011-12-02T22:52:51 roflmao: I'm sorry Dave, err roflmao; I cannot 'besh:'. 2011-12-02T22:53:01 @learnhowto 2011-12-02T22:53:02 roflmao: An error occured while trying to show the previous error. 2011-12-02T23:00:34 wtf 2011-12-02T23:00:44 if i comment a line out input time is going up :\ 2011-12-02T23:00:50 and the line is doing something 2011-12-02T23:02:20 bhasker: did you ever work out your coordinate problem? I think I can help you otherwise 2011-12-02T23:02:45 oh, maybe not when the battles wrap 2011-12-02T23:03:31 antimatroid: gcc does some really strange stuff :P 2011-12-02T23:04:40 gcc -S -fverbose-asm will give you commented assembly which I find nice for optimizing :) 2011-12-02T23:04:53 sucks in C++ due to all the magic done by the containers/templates in the standard lib though 2011-12-02T23:05:31 well actually 'gcc -S -fverbose-asm -masm=intel' cause the default syntax is unreadable 2011-12-02T23:06:32 foflmao: yeah i know :) 2011-12-02T23:07:28 weakend is reserved for improving the bot one last time 2011-12-02T23:07:33 *** bmh has joined #aichallenge 2011-12-02T23:10:51 *** analyst74 has joined #aichallenge 2011-12-02T23:15:38 *** treeform has joined #aichallenge 2011-12-02T23:18:40 *** besh has quit IRC (Quit: Page closed) 2011-12-02T23:21:07 *** bmh_ has joined #aichallenge 2011-12-02T23:21:29 *** bmh has quit IRC (Disconnected by services) 2011-12-02T23:21:34 *** bmh_ is now known as bmh 2011-12-02T23:22:44 http://ants.fluxid.pl/replay.10134 no more combat bugs! 2011-12-02T23:23:11 thestinger: I hear you're a haskell whiz? 2011-12-02T23:23:16 no 2011-12-02T23:23:21 I just know haskell a bit :P 2011-12-02T23:23:45 wouldn't want to use it for this tbh, I don't know it well enough to make it fast 2011-12-02T23:23:46 any debugging pro-tips? I'm getting a runtime error (index out of bounds) and can't track the darn thing down 2011-12-02T23:24:10 unsafePerformIO? 2011-12-02T23:24:15 i stripped my bot yesterday 2011-12-02T23:24:26 i'm trying to get all the info i'm using in various places as cheaply as possible 2011-12-02T23:24:28 I was afraid that would be your answer. 2011-12-02T23:25:24 bmh: http://www.haskell.org/haskellwiki/Debugging the wiki has some good stuff 2011-12-02T23:25:45 I don't like using breakpoints even C for some reason though 2011-12-02T23:26:24 I prefer getting a stack trace. It would be wonderful if it could just decorate each thunk with its origin. 2011-12-02T23:26:35 (back in a bit) 2011-12-02T23:29:21 *** ericlavigne has quit IRC (Quit: Page closed) 2011-12-02T23:32:09 lol, got to be kidding me 2011-12-02T23:32:36 turns out i think making grids with vectors is quicker with ints than bools 2011-12-02T23:32:46 *** avdg has quit IRC (Quit: Leaving.) 2011-12-02T23:33:01 *** u_ has quit IRC (Quit: u_) 2011-12-02T23:33:15 antimatroid: vector is slow as hell 2011-12-02T23:33:26 why the hell is it any different if not to be faster? 2011-12-02T23:33:28 it isn't a vector of bools, it's a bunch of bitfields 2011-12-02T23:33:39 yeah i know it's different, that's a pita too 2011-12-02T23:33:42 never use vector 2011-12-02T23:33:45 but i expected it to be faster 2011-12-02T23:33:45 use vector 2011-12-02T23:34:04 i will be able to index again properly again then too :D 2011-12-02T23:34:14 you don't overload say [] for a matrix srtuct 2011-12-02T23:34:19 *** TheLinker has quit IRC (Quit: Bye) 2011-12-02T23:34:21 cause you can't return a reference to an element of a vector of bools 2011-12-02T23:34:32 well, for location indexing 2011-12-02T23:34:44 you can for indexing by [r] and returning a reference to a vector of bools 2011-12-02T23:35:09 *** avdg has joined #aichallenge 2011-12-02T23:35:54 uint8_t will work fine, but I wouldn't be surprised if unsigned/int are still faster 2011-12-02T23:36:25 no point using 4x/8x the space than uint8_t just for 2% speed or whatever though :) 2011-12-02T23:37:06 * antimatroid tests what's fastest 2011-12-02T23:37:20 when you're searching how do you mark what's searched? 2011-12-02T23:37:29 i figure these are faster than a set? 2011-12-02T23:37:42 but it depends on the board size i guess 2011-12-02T23:37:44 oh, for BFS? 2011-12-02T23:37:54 yeah, or some other kind of search 2011-12-02T23:38:07 well most of my searches just mark distances, so I just compare the new distance I would set that tile to with the old value 2011-12-02T23:38:39 for the ones that need to track visited I use std::vector> atm 2011-12-02T23:38:58 to mark path distance isn't that essentially the same thing? 2011-12-02T23:39:27 *** avdg has quit IRC (Ping timeout: 260 seconds) 2011-12-02T23:39:27 and you can't search off euclidean or manhatten 2011-12-02T23:39:31 because that wouldn't work :P 2011-12-02T23:39:32 well, since I'm already marking path distances it means I don't need a second 2d vector for marking which are visited 2011-12-02T23:39:38 yeah sure 2011-12-02T23:39:49 for the radius stuff I just paint it on with offsets 2011-12-02T23:40:12 yeah i have translation values for attack, attackBorder, view and viewBorder 2011-12-02T23:40:21 std::unordered_set might actually be faster for the few searches I do without making distance maps though 2011-12-02T23:40:40 but i want to store in each location the possible locations an ant residing in this turn it could battle next turn 2011-12-02T23:40:42 since my hash function basically cheats and there will never be collisions :) 2011-12-02T23:40:50 so i don't actually need to do those calculations when i make my battle partitions 2011-12-02T23:41:31 I bet a 2D array is as fast you can get though 2011-12-02T23:42:01 a 2D std::vector probably has a lot of overhead since you have a vector of pointers to vectors, when you really don't ever resize them 2011-12-02T23:42:53 the good old 'row * cols + col' trick works well though for using a 1D array 2011-12-02T23:43:15 I'm way too lazy to optimize like that though 2011-12-02T23:43:28 I could probably just use an array 2011-12-02T23:43:39 although i might not get indexing how i'd like then 2011-12-02T23:43:49 i'll update the matrix class to use a 1d vector 2011-12-02T23:44:01 would templates be slow? 2011-12-02T23:44:12 no, templates don't add overhead 2011-12-02T23:44:12 ie. should i just write a separate matrix class for anything i might use it for 2011-12-02T23:44:14 cool 2011-12-02T23:44:16 they're compile-time code generation 2011-12-02T23:44:40 i don't know very much about behind the scenes stuff, i just learn how to do what i need and leave it at that usually :P 2011-12-02T23:45:02 like template T min(T a, T b); will compile to the same thing as defining min for each type 2011-12-02T23:45:23 awesome, templates can be awesome 2011-12-02T23:45:31 i do wish c++ had type classes and variables though 2011-12-02T23:45:36 yeah 2011-12-02T23:45:52 'auto' in C++11 is nice though 2011-12-02T23:45:57 what's that? 2011-12-02T23:46:04 local type inference 2011-12-02T23:46:52 std::vector::const_iterator it = vec.begin(); -> auto it = vec.cbegin(); 2011-12-02T23:48:13 *** bmh has quit IRC (Ping timeout: 265 seconds) 2011-12-02T23:48:34 still type safe ofc 2011-12-02T23:48:43 *** ssclift has joined #aichallenge 2011-12-02T23:49:37 https://en.wikipedia.org/wiki/Concepts_(C++) they were going to add something like type classes too 2011-12-02T23:50:17 only GCC started to implement them 2011-12-02T23:50:45 the other compilers didn't want to I guess, so it was dropped 2011-12-02T23:51:07 pesky compilers 2011-12-02T23:51:17 * antimatroid gets back to coding 2011-12-02T23:51:22 xD 2011-12-02T23:51:31 s/compilers/compiler vendors/ 2011-12-02T23:51:37 was going to say that but it didn't really make sense either 2011-12-02T23:53:52 no indexing my matrix by [r][c] with this now 2011-12-02T23:53:53 *** Antimony_ has joined #aichallenge 2011-12-02T23:53:56 but that's a okay by me 2011-12-02T23:54:06 hmm i think 2011-12-02T23:54:20 overloading operator() for that is easier 2011-12-02T23:54:33 oh, i can overload [r,c] 2011-12-02T23:54:35 i forgot that 2011-12-02T23:55:25 oh, that's neat 2011-12-02T23:56:58 *** contestbot_ has joined #aichallenge 2011-12-02T23:57:19 less nice though :P 2011-12-02T23:57:20 *** Antimony has quit IRC (Ping timeout: 255 seconds) 2011-12-02T23:57:33 *** Antimony_ is now known as Antimony 2011-12-02T23:57:34 *** delt0r_ has quit IRC (Ping timeout: 244 seconds) 2011-12-02T23:57:46 oh yes, that's why i used operator() instead of [] 2011-12-02T23:57:56 matrix(row,col) matrix[row][col] <- 1 extra character 2011-12-02T23:58:00 I'm sure it compiles to the same code xD 2011-12-02T23:58:20 maybe not... 2011-12-02T23:58:20 yeah, but my preference would go m[r][c], m[r,c], m(r,c) 2011-12-02T23:58:26 but then your operator[] has to return some sort of sentinel with its own operator[] 2011-12-02T23:58:35 *** amstan_ has joined #aichallenge 2011-12-02T23:58:35 *** ChanServ sets mode: +o amstan_ 2011-12-02T23:58:54 a1k0n: I've found that gcc is good at getting rid of stuff like that 2011-12-02T23:58:55 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-12-02T23:58:55 *** contestbot has quit IRC (Ping timeout: 248 seconds) 2011-12-02T23:59:03 you really have to mark stuff as 'inline' and turn on -Winline though 2011-12-02T23:59:20 sometimes it decides not to inline a 1 line function 2011-12-02T23:59:20 *** keith_ has quit IRC (Ping timeout: 265 seconds) 2011-12-02T23:59:21 is -Winline on the server? 2011-12-02T23:59:23 i have these as inline 2011-12-02T23:59:33 -Winline just warns you if it doesn't inline something 2011-12-02T23:59:39 well sure but i mean. to implement it in the first place unless you just have vectors of vectors, you need to emulate that somehow 2011-12-02T23:59:45 so if you can force it to inline locally it will probably do it on the server