2010-10-28T00:00:27 heh, antbot likes to timeout... aw 2010-10-28T00:00:53 *** TheMacca has joined #aichallenge 2010-10-28T00:02:38 ooh, game with CCCF 2010-10-28T00:03:05 *** seisatsu has joined #aichallenge 2010-10-28T00:03:37 ex9, 12, and 13 are his top ones 2010-10-28T00:03:45 who does cccf anyway? 2010-10-28T00:04:01 no idea 2010-10-28T00:04:13 people say felixcoto 2010-10-28T00:04:20 ooh, and a clutch victory. hooray! 2010-10-28T00:05:44 Wow. I love how the bot always wins by one. 2010-10-28T00:05:52 wins a planet* 2010-10-28T00:07:44 and defends with 0 left, that's why it's hard for a human to compete 2010-10-28T00:11:43 my bot is doing better than i would have liked given i discarded the changes i made last night 2010-10-28T00:12:17 Top 10 players: bocsimacko(3938), rsergio(3914), GreenTea(3905), dmj111(3842), rs5147(3828), felixcoto(3786), protocolocon(3764), luksian(3756), Hazard(3745), Arekku(3744) 2010-10-28T00:13:00 why is bocsimacko so good... 2010-10-28T00:16:06 I'm turning in, have fun with all the pokemon 2010-10-28T00:16:16 i'm playing you now :P 2010-10-28T00:16:28 and cccf's best bot 2010-10-28T00:16:40 *** Titankiller has quit IRC (Ping timeout: 265 seconds) 2010-10-28T00:17:01 Awww: INFO You LOSE against davidjliu 2010-10-28T00:17:08 heheheh 2010-10-28T00:17:39 my bot's been gradually slipping though... 2010-10-28T00:18:10 *** flag has joined #aichallenge 2010-10-28T00:18:28 Does anyone know what I can do if I forget my password on the official server? 2010-10-28T00:18:47 I don't see a "forgot your password?" option anywhere 2010-10-28T00:19:23 I don't think so... 2010-10-28T00:19:29 you can ask the admins on the forums though, I think... 2010-10-28T00:19:43 Yeah, that's probably the only solution 2010-10-28T00:19:52 The rules say I'm only supposed to have one account aswell 2010-10-28T00:20:43 they can't really easily enforce that 2010-10-28T00:20:50 it's just intended that people don't have more than one bot 2010-10-28T00:20:57 don't take that as official blessing or anything though 2010-10-28T00:21:54 oh no 2010-10-28T00:22:00 i've seen one guys profile where he apologised on his about me about it being his second account 2010-10-28T00:22:02 a loss to charmander 2010-10-28T00:22:04 for the same reason 2010-10-28T00:22:14 is it mistmanovx? 2010-10-28T00:22:16 ok 2010-10-28T00:22:23 I'll make a forum post first and see how that goes 2010-10-28T00:23:28 boscimako's bot amazes me 2010-10-28T00:23:51 you're not the only one 2010-10-28T00:24:02 *** Titankiller has joined #aichallenge 2010-10-28T00:24:12 heh 2010-10-28T00:24:26 *** Mjothvitnir has joined #aichallenge 2010-10-28T00:25:16 calling it a night... see you all later 2010-10-28T00:25:23 g'night 2010-10-28T00:25:39 *** davidjliu has quit IRC (Quit: Page closed) 2010-10-28T00:29:31 ooh nice move at turn 25: http://72.44.46.68/canvas?game_id=741498 2010-10-28T00:30:32 Yesh! INFO You WIN against bocsimacko-b0.30 , http://72.44.46.68/canvas?game_id=741788 2010-10-28T00:30:37 *** perestrelka has quit IRC (Ping timeout: 265 seconds) 2010-10-28T00:32:02 That's like the one he lost on the official server 2010-10-28T00:32:46 well done 2010-10-28T00:35:54 *** Mjothvitnir has quit IRC (Quit: Page closed) 2010-10-28T00:52:17 *** hellman has joined #aichallenge 2010-10-28T00:52:27 *** amstan has quit IRC (Remote host closed the connection) 2010-10-28T01:01:40 *** narnach has joined #aichallenge 2010-10-28T01:09:13 *** Titankiller has quit IRC (Ping timeout: 255 seconds) 2010-10-28T01:12:18 Top 10 players: bocsimacko(3922), rsergio(3903), GreenTea(3895), dmj111(3829), rs5147(3818), felixcoto(3782), protocolocon(3753), luksian(3746), Arekku(3743), shangas(3731) 2010-10-28T01:15:15 *** delt0r has joined #aichallenge 2010-10-28T01:16:31 *** delt0r___ has quit IRC (Ping timeout: 272 seconds) 2010-10-28T01:17:31 *** Appleman1234 has joined #aichallenge 2010-10-28T01:20:55 *** perestrelka has joined #aichallenge 2010-10-28T01:30:36 is it always going to be just 200 turns max? 2010-10-28T01:30:54 *** Appleman1234 has quit IRC (Ping timeout: 240 seconds) 2010-10-28T01:31:34 *** Appleman1234 has joined #aichallenge 2010-10-28T01:36:22 they're talking about that atm 2010-10-28T01:36:29 wait a day or so and we should have a definitive answer 2010-10-28T01:53:29 *** espes has joined #aichallenge 2010-10-28T01:55:57 *** smellyhippy has quit IRC (Ping timeout: 245 seconds) 2010-10-28T02:00:22 *** smellyhippy has joined #aichallenge 2010-10-28T02:00:23 *** smellyhippy has joined #aichallenge 2010-10-28T02:03:38 *** TheMacca has quit IRC (Quit: Page closed) 2010-10-28T02:05:32 *** narnach has quit IRC (Quit: Leaving.) 2010-10-28T02:07:41 *** bhasker has joined #aichallenge 2010-10-28T02:08:15 *** airhorns has joined #aichallenge 2010-10-28T02:11:22 *** hornairs has quit IRC (Ping timeout: 245 seconds) 2010-10-28T02:12:18 Top 10 players: bocsimacko(3926), rsergio(3907), GreenTea(3894), dmj111(3828), rs5147(3803), felixcoto(3786), luksian(3762), protocolocon(3753), Arekku(3742), shangas(3728) 2010-10-28T02:18:42 *** sigh has quit IRC (Read error: Connection reset by peer) 2010-10-28T02:19:06 *** sigh has joined #aichallenge 2010-10-28T02:31:57 *** sigh has quit IRC (Remote host closed the connection) 2010-10-28T02:36:22 *** yasith has quit IRC (Ping timeout: 245 seconds) 2010-10-28T02:40:03 *** Cyndre has joined #aichallenge 2010-10-28T02:41:46 antimatroid: where is that conversation? 2010-10-28T02:42:10 *** hightime has joined #aichallenge 2010-10-28T02:43:12 hi all 2010-10-28T02:43:32 *** Utkarsh has quit IRC (Ping timeout: 250 seconds) 2010-10-28T02:48:26 *** Utkarsh has joined #aichallenge 2010-10-28T02:50:45 mega1: mailing list 2010-10-28T02:51:01 antimatroid: I haven't found the mailing list 2010-10-28T02:51:12 mathnerd wanted to know about turn limits, max minimum distance and prizes 2010-10-28T02:53:48 it looks like max minimum distance will be 2 rather than 1 :( 2010-10-28T02:53:50 i wanted 1 2010-10-28T02:56:30 *** Palmik has joined #aichallenge 2010-10-28T02:56:36 antimatroid: what about turn limits? 2010-10-28T02:56:39 sticking to 200? 2010-10-28T02:57:03 because that could be a factor for good bots playing each other 2010-10-28T02:57:37 it was explicitly asked not to post this kind of stuff on the forums, so don't take it as fact 2010-10-28T02:57:47 jcameron suggest raising it to 400, others said 200 was fine 2010-10-28T02:58:06 i think a higher one is necessary, i've seen quite a few games in the last week or two where I think the result would have ended differently given more time 2010-10-28T02:58:17 well, not necessary, but desirable 2010-10-28T02:59:49 but where is the mailing list? 2010-10-28T02:59:55 can't seem to find it 2010-10-28T03:00:17 i'm not sure sorry 2010-10-28T03:00:22 i'll have a bit of a look 2010-10-28T03:00:56 ai-contest-discuss@googlegroups.com if you can get there with that information? 2010-10-28T03:01:13 *** Utkarsh has quit IRC (Ping timeout: 276 seconds) 2010-10-28T03:01:13 i don't really have any idea about how mailing lists work, i just signed up to it ages ago 2010-10-28T03:01:48 *** Bobng has joined #aichallenge 2010-10-28T03:02:04 http://groups.google.com/group/ai-contest-discuss that might be it 2010-10-28T03:03:16 yeah, discussions are here http://groups.google.com/group/ai-contest-discuss/topics 2010-10-28T03:04:43 You cannot view the group's content or participate in the group because you are not currently a member. Members must be approved before joining. 2010-10-28T03:04:51 I've applied for membership. 2010-10-28T03:04:57 What a strange default. 2010-10-28T03:05:10 i think it's to keep it private 2010-10-28T03:05:32 What for? In a public contest? 2010-10-28T03:06:14 i guess there's still reason for having some private conversations 2010-10-28T03:06:19 it's still fairly open this time 2010-10-28T03:06:41 *** Utkarsh has joined #aichallenge 2010-10-28T03:07:30 I guess there might be, but the Mathnerd mentioned was certainly of public interest. 2010-10-28T03:08:25 yeah, the information will be released in the next couple of days 2010-10-28T03:08:53 he was pointing out to jeff that it's a month before the contest and we need some answers 2010-10-28T03:10:46 pointing out in all caps, heh 2010-10-28T03:11:46 antimatroid: ah ok 2010-10-28T03:12:19 Top 10 players: bocsimacko(3920), rsergio(3900), GreenTea(3888), dmj111(3820), rs5147(3811), felixcoto(3779), luksian(3757), protocolocon(3748), shangas(3743), Arekku(3729) 2010-10-28T03:18:16 *** Mjothvitnir has joined #aichallenge 2010-10-28T03:26:41 *** Bobng has quit IRC (Quit: Leaving) 2010-10-28T03:27:39 *** vizier has joined #aichallenge 2010-10-28T03:34:20 *** narnach has joined #aichallenge 2010-10-28T03:34:20 *** narnach is now known as narnach_mb 2010-10-28T03:36:17 *** Mjothvitnir has quit IRC (Quit: Page closed) 2010-10-28T03:41:48 *** SamothRefoe has joined #aichallenge 2010-10-28T03:43:27 *** perror has joined #aichallenge 2010-10-28T03:45:25 *** Utkarsh has quit IRC (Ping timeout: 276 seconds) 2010-10-28T03:46:52 *** aerique has joined #aichallenge 2010-10-28T03:51:19 *** antimatroid1 has joined #aichallenge 2010-10-28T03:52:32 *** Utkarsh has joined #aichallenge 2010-10-28T03:52:44 *** antimatroid has quit IRC (Read error: Connection reset by peer) 2010-10-28T03:57:20 yo dog, i heard you liked trees 2010-10-28T03:58:51 *** boegel has joined #aichallenge 2010-10-28T04:02:35 *** phreeza has joined #aichallenge 2010-10-28T04:03:11 I do like trees.... but only balanced binary ones... 2010-10-28T04:05:51 *** Error323 has joined #aichallenge 2010-10-28T04:06:04 *** Error323 has left #aichallenge 2010-10-28T04:06:08 *** Florian95 has joined #aichallenge 2010-10-28T04:06:08 *** Error323 has joined #aichallenge 2010-10-28T04:10:37 where is the almighty cloud 2010-10-28T04:12:19 Top 10 players: bocsimacko(3933), rsergio(3907), GreenTea(3894), dmj111(3825), rs5147(3818), felixcoto(3784), luksian(3766), protocolocon(3756), shangas(3752), sin_sun(3742) 2010-10-28T04:12:57 *** thord has joined #aichallenge 2010-10-28T04:16:24 *** thord is now known as egauo 2010-10-28T04:17:31 *** hightime has quit IRC (Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org) 2010-10-28T04:28:33 *** Naktibalda has joined #aichallenge 2010-10-28T04:34:40 *** yasith has joined #aichallenge 2010-10-28T05:12:21 Top 10 players: bocsimacko(3968), rsergio(3916), GreenTea(3902), dmj111(3839), rs5147(3839), felixcoto(3792), luksian(3771), shangas(3769), protocolocon(3763), sin_sun(3749) 2010-10-28T05:17:43 *** Arthur has joined #aichallenge 2010-10-28T05:26:59 *** jaspervdj has joined #aichallenge 2010-10-28T05:50:57 *** dmj111 has quit IRC (Ping timeout: 245 seconds) 2010-10-28T06:06:34 *** egauo has quit IRC (Quit: Page closed) 2010-10-28T06:12:20 Top 10 players: bocsimacko(3953), rsergio(3889), GreenTea(3880), rs5147(3826), dmj111(3812), felixcoto(3765), shangas(3756), protocolocon(3748), sin_sun(3735), Accoun(3730) 2010-10-28T06:21:25 *** iNtERrUpT has joined #aichallenge 2010-10-28T06:25:17 *** SamothRefoe has quit IRC (Read error: Connection reset by peer) 2010-10-28T06:25:30 *** SamothRefoe has joined #aichallenge 2010-10-28T06:25:56 *** SamothRefoe has joined #aichallenge 2010-10-28T06:28:42 *** SamothRefoe has quit IRC (Client Quit) 2010-10-28T06:37:08 *** hellman has quit IRC (Remote host closed the connection) 2010-10-28T06:47:37 *** boegel has quit IRC (Ping timeout: 240 seconds) 2010-10-28T06:48:55 *** Arthur has quit IRC (Ping timeout: 240 seconds) 2010-10-28T06:50:23 *** Arthur has joined #aichallenge 2010-10-28T07:05:56 you people are awfully quiet 2010-10-28T07:06:39 we are bots 2010-10-28T07:09:27 whyyyyy is there only one server active :/ 2010-10-28T07:09:30 this takes aaaaaages 2010-10-28T07:12:20 Top 10 players: bocsimacko(3955), rsergio(3901), GreenTea(3887), rs5147(3829), dmj111(3818), felixcoto(3778), shangas(3760), protocolocon(3757), sin_sun(3741), Accoun(3738) 2010-10-28T07:19:56 *** n9thbit has joined #aichallenge 2010-10-28T07:29:03 *** dmj111 has joined #aichallenge 2010-10-28T07:43:12 meh i should stop designing my bot and actually getting to do something 2010-10-28T07:43:26 ok resolution for today, get it to run 2010-10-28T07:43:27 :P 2010-10-28T07:43:42 brb transporting myself 2010-10-28T07:43:50 *** bhasker has quit IRC (Ping timeout: 250 seconds) 2010-10-28T07:50:40 *** bhasker has joined #aichallenge 2010-10-28T07:54:54 *** Arthur has quit IRC (Disconnected by services) 2010-10-28T07:55:19 *** Arthur_ has joined #aichallenge 2010-10-28T08:00:40 i just realised, if you realised early enough that you were going to lose, you could lead the other bot on quite the goose chase 2010-10-28T08:00:58 assuming they strictly take a neutral you're planning to take one move after you, and you have a few ships to spare with claiming neutrals 2010-10-28T08:11:46 *** Arthur_ has quit IRC (Ping timeout: 276 seconds) 2010-10-28T08:12:22 Top 10 players: bocsimacko(3949), rsergio(3893), GreenTea(3876), rs5147(3817), dmj111(3801), felixcoto(3765), shangas(3762), protocolocon(3750), sin_sun(3735), Accoun(3733) 2010-10-28T08:12:51 what's a point of it? 2010-10-28T08:13:15 *** Arthur_ has joined #aichallenge 2010-10-28T08:13:16 lose faster? 2010-10-28T08:14:27 lose slower 2010-10-28T08:14:41 the idea would be to zig zag back and forth across the map 2010-10-28T08:15:07 isn't it yet another idea that works agains't stupid bots only? 2010-10-28T08:21:04 *** boegel has joined #aichallenge 2010-10-28T08:21:09 you'd still lose on number of ships probably 2010-10-28T08:23:19 at the same time you have to defend and expand 2010-10-28T08:25:21 morning 2010-10-28T08:25:29 http://72.44.46.68/canvas?game_id=746301 2010-10-28T08:27:12 you're too early :P 2010-10-28T08:27:16 it's still evening 2010-10-28T08:27:40 *** iNtERrUpT has quit IRC (Remote host closed the connection) 2010-10-28T08:28:28 *** Utkarsh has quit IRC (Ping timeout: 250 seconds) 2010-10-28T08:29:31 *** Utkarsh has joined #aichallenge 2010-10-28T08:29:39 heh 2010-10-28T08:29:43 its 5:30 am! 2010-10-28T08:30:09 11 30pm in australia 2010-10-28T08:30:32 heh 2010-10-28T08:30:39 how is your new bot coming along 2010-10-28T08:30:45 is retard still retarded 2010-10-28T08:30:49 *** Appleman1234 has quit IRC (Ping timeout: 255 seconds) 2010-10-28T08:32:21 yeah i'm working on a completely different one now aha 2010-10-28T08:32:59 *** bhasker_ has joined #aichallenge 2010-10-28T08:33:10 i'm going to throw stuff into a simultaneous game 2010-10-28T08:33:15 and make my decisions based off that 2010-10-28T08:33:19 i just restarted my bot 2010-10-28T08:33:33 lets see how it does, made a few changes 2010-10-28T08:33:40 i already have most of the functions written for a library i'm writing, it's just a matter of making it take the game states etc. 2010-10-28T08:33:58 what do you mean a simultaneous game 2010-10-28T08:34:03 i'm not entirely sure as to what to do when there aren't any saddle points 2010-10-28T08:34:05 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-10-28T08:34:06 you going to run the two instances 2010-10-28T08:34:08 one playing as you 2010-10-28T08:34:11 and one as the enemy 2010-10-28T08:34:14 yep 2010-10-28T08:34:16 ah k 2010-10-28T08:34:23 and each will pick moves, then create payoff matrix 2010-10-28T08:34:40 i'm creating the matrix with nodes, then i can run each of those instances too before working out the payoff 2010-10-28T08:34:54 i'll probably use a tree for each player to pick possible pure strategies too 2010-10-28T08:35:07 may even try to set up something that'll try to converge to mixed strategies when it's needed 2010-10-28T08:35:22 converge to mixed strategy nash equilibrium** 2010-10-28T08:35:38 ah k 2010-10-28T08:35:49 hmm my bot just got owned by testingFromPiter:( 2010-10-28T08:36:21 i considered trying to use something like minimax regret, but can't work out how to do it for things other than decision choices and not just get the same result, but have to compute nash equilibria for non-zero sum games 2010-10-28T08:36:55 i found a cool paper on markov games before too, but i think it mostly went over my head 2010-10-28T08:37:05 <--- reading up on nash equilibrium 2010-10-28T08:37:14 <--- not very mathematically inclined 2010-10-28T08:37:27 it's basically just the name attached to fixed points in games 2010-10-28T08:37:44 in that, neither player can change their strategy given those of the other players to increase their payoff 2010-10-28T08:38:31 ah k 2010-10-28T08:39:26 unless one can pick good potential moves from both players i don't think it's going to matter how well you do things on the macro scale for this anyway though 2010-10-28T08:39:27 morning 2010-10-28T08:39:54 and that is incredibly tricky imo 2010-10-28T08:40:14 hmm 2010-10-28T08:40:15 I was looking at http://72.44.46.68/canvas?game_id=746301 -- is red's move at turn 19 bad? 2010-10-28T08:40:59 hard to say, they were behind 2010-10-28T08:41:41 i think red made a conservative choice 2010-10-28T08:41:46 it would have been better not to acquire 2010-10-28T08:41:47 we were even 2010-10-28T08:41:57 at 18 2010-10-28T08:42:03 we are both at 210/16 2010-10-28T08:42:13 there was no reason to acquire a new neutral with an enemy this close 2010-10-28T08:42:35 so when he shunted ships away from center 2010-10-28T08:42:43 i could take a bigger neutral with no threat 2010-10-28T08:43:08 if he hadn't have done that it would have been a draw I guess 2010-10-28T08:43:58 yes 2010-10-28T08:44:47 it's interesting to note that almost everybody uses the same opening move or two. What do you think of taking the two center 34's as an opening move? 2010-10-28T08:45:39 *** perestrelka has quit IRC (Read error: Connection reset by peer) 2010-10-28T08:46:02 if they wait they'd own you 2010-10-28T08:46:07 i think 2010-10-28T08:46:10 that's a lot of cost 2010-10-28T08:46:15 *** perestrelka has joined #aichallenge 2010-10-28T08:47:22 the 34's are expensive 2010-10-28T08:47:23 and 2010-10-28T08:47:26 are too close 2010-10-28T08:47:27 to each other 2010-10-28T08:47:31 *** lavalamp_ has quit IRC (Quit: Page closed) 2010-10-28T08:47:33 like 1 turn 2010-10-28T08:47:46 so if you launch he can easily snipe it 2010-10-28T08:48:04 my bot explicitly ignores planets at similar distances unless the enemy makes a move 2010-10-28T08:48:07 to them 2010-10-28T08:48:22 or if i have enough ships to take it and hold it 2010-10-28T08:51:33 if you send more than double the troops present at a neutral planet 2010-10-28T08:51:50 the planet never lessens in strength 2010-10-28T08:53:12 Call for uploads: http://ai-contest.com/forum/viewtopic.php?f=17&t=959 - I'm tired of playing against myself in local matches 2010-10-28T08:53:22 eburnette tcpserver 2010-10-28T08:53:23 ? 2010-10-28T08:54:08 tcpserver can only get about 100 games an hour but if I run locally I can get thousands 2010-10-28T08:54:35 what are you going to do with thousands games? 2010-10-28T08:55:08 benchmark changes to see if they improve or degrade play over many different maps 2010-10-28T08:55:23 do you run 64bit debian? 2010-10-28T08:55:32 than i could give you my binary 2010-10-28T08:56:02 win7 64-bit :( 2010-10-28T08:57:30 i dunno about sharing binaries anyway 2010-10-28T08:57:40 are they still considering allowing binary submissions? 2010-10-28T08:57:54 wha that would be sweet 2010-10-28T08:58:08 don't get your hopes up, it's fairly late to allow that 2010-10-28T08:58:14 but it was being talked about at some point 2010-10-28T08:58:17 if the sandbox is sound i don't see why not 2010-10-28T08:58:19 true 2010-10-28T08:58:23 next time :) 2010-10-28T08:58:40 they were also talking about having people run their own bots like tcp 2010-10-28T08:58:50 I don't like binary submissions, they couldn't check to see if rules are being followed 2010-10-28T08:59:00 but you basically have to throw out time limits the moment you do that, and people with "good" connections get an advantage 2010-10-28T08:59:12 own bots through tcp is fail with a slow connection 2010-10-28T08:59:15 http://72.44.46.68/canvas?game_id=747569 2010-10-28T08:59:16 a tcp-only tourney would be interesting, then I could upload to supercomputer :) 2010-10-28T08:59:20 yes... and money can buy more comp power too 2010-10-28T08:59:21 hah my bot successfully snipes 2010-10-28T08:59:27 took me a long time to get that code working correctly 2010-10-28T08:59:32 damn thing would either be a turn ahead 2010-10-28T08:59:38 or at the same turn 2010-10-28T08:59:48 bhasker_: are you Shampoo? 2010-10-28T08:59:51 but there allready was a binary swap topic wasnt there 2010-10-28T08:59:56 no malazan 2010-10-28T08:59:58 what was wrong with those bins? 2010-10-28T09:00:01 ah ok 2010-10-28T09:00:02 eburnette 2010-10-28T09:00:20 bhasker can you snipe from multiple sources or just one? 2010-10-28T09:01:21 i hate shampoo 2010-10-28T09:01:29 he gives me endless frustration 2010-10-28T09:01:36 even when my bot isn't in shambles 2010-10-28T09:02:23 http://72.44.46.68/canvas?game_id=732062 sniper defense ^_^ 2010-10-28T09:03:11 it's surprisingly confusing to do minimax on a game matrix 2010-10-28T09:03:26 is there a list of who does what bot? 2010-10-28T09:03:36 not to mention the fact that i also need to make a move when there are no saddle points 2010-10-28T09:03:38 where is shampoo ranked currently on the official server? 2010-10-28T09:03:48 what's a saddle point 2010-10-28T09:04:12 a fixed point for both players 2010-10-28T09:04:17 i can snipe from multiple 2010-10-28T09:04:23 one player is maximising their payoff given current moves, the other minimising theirs 2010-10-28T09:04:45 locutus bot does weird stuff 2010-10-28T09:04:47 each player's objective is to max/min respectively 2010-10-28T09:04:50 nice bhasker_ that's hard 2010-10-28T09:05:09 why is that hard? 2010-10-28T09:05:13 i pick all my moves with a single source and mutliple destinations 2010-10-28T09:05:15 hmm its the same as single source sniping 2010-10-28T09:05:19 i thought everyone'd be doing that? 2010-10-28T09:05:21 if you have coordinated launches right 2010-10-28T09:05:23 it will work 2010-10-28T09:05:30 wait, reverse that 2010-10-28T09:05:35 yea 2010-10-28T09:05:36 :) 2010-10-28T09:05:36 single destination and multiple sources 2010-10-28T09:05:59 eburnette which one is your bot? 2010-10-28T09:06:05 charmander? 2010-10-28T09:06:10 well I found it hard because I don't want to attack with less than overwhelming force 2010-10-28T09:06:22 some bastard stole my charizard card out of my desk in grade 6 2010-10-28T09:06:37 then like 2 weeks later my secret santa got me a pack and i got another, was awesome 2010-10-28T09:07:09 lol'd 2010-10-28T09:07:10 bhasker_ yeah the pokemon ones like charmander and infernape 2010-10-28T09:07:13 'i hate shampoo' 2010-10-28T09:07:15 ah k 2010-10-28T09:07:17 'wut?!' 2010-10-28T09:07:35 i always think rape 2010-10-28T09:07:55 like inferno-rape or something 2010-10-28T09:08:12 ooookkkkkk 2010-10-28T09:08:25 *** boegel has quit IRC (Ping timeout: 264 seconds) 2010-10-28T09:08:29 seen any good blogs about the contest? 2010-10-28T09:08:50 i saw some "advanced strategy guide" the other day 2010-10-28T09:08:52 but it sucked 2010-10-28T09:08:59 was on that ai forum "offering prizes" 2010-10-28T09:09:15 of like 6 months membership to their website 2010-10-28T09:09:30 ebur you can see my bot attack from multuple points neatly in this game 2010-10-28T09:09:31 http://72.44.46.68/canvas?game_id=747682 2010-10-28T09:09:39 the top and bottom attacks work independently 2010-10-28T09:09:54 yes anti ai-gamedev 2010-10-28T09:10:08 this one? http://aigamedev.com/open/tutorials/planetwars-guide/ 2010-10-28T09:10:15 that's it 2010-10-28T09:10:36 it wasnt' that bad for someone starting u 2010-10-28T09:10:51 most people who have a decent bot would know most of what they are saying 2010-10-28T09:11:02 it says "advande strategy guide" :P 2010-10-28T09:11:13 its reasonably advanced :) 2010-10-28T09:11:28 not really :P 2010-10-28T09:11:28 if you look at the strategy guid on the official website 2010-10-28T09:11:34 i would say theirs is certainly better ^ 2010-10-28T09:11:39 there was a great strategy guide last contest 2010-10-28T09:11:53 i wasn't measuring it against the official one :P 2010-10-28T09:12:19 bhasker_ wow you pwned that one 2010-10-28T09:12:21 Top 10 players: bocsimacko(3944), rsergio(3887), GreenTea(3870), rs5147(3810), dmj111(3795), felixcoto(3758), shangas(3755), protocolocon(3744), Accoun(3728), sin_sun(3727) 2010-10-28T09:12:28 he made a mistake ebur 2010-10-28T09:12:36 he didn't take the 5 growth planet 2010-10-28T09:12:40 with 12 ships on it 2010-10-28T09:12:50 it would have been different if he had 2010-10-28T09:12:59 so i was up 4 growth almost instantly 2010-10-28T09:13:08 5 or 4 growth not sure 2010-10-28T09:13:27 once someone has that kind of an advantage its hard to level it on close maps 2010-10-28T09:13:51 yeah I think his taking that 50 planet on the right was a mistake too 2010-10-28T09:14:20 this http://72.44.46.68/canvas?game_id=747720 2010-10-28T09:14:22 he probably has something that doesn't take planets that are equidistant on first move 2010-10-28T09:14:22 is nice 2010-10-28T09:14:40 i do that too 2010-10-28T09:15:08 *** delt0r___ has joined #aichallenge 2010-10-28T09:16:17 *** delt0r has quit IRC (Ping timeout: 240 seconds) 2010-10-28T09:16:56 wouldn't that 55 on the left have been a better target at turn 17? 2010-10-28T09:18:11 let me check 2010-10-28T09:18:25 not really 2010-10-28T09:18:34 it would have put me back by 50 ships 2010-10-28T09:18:36 almost 2010-10-28T09:18:45 the payback time for that planet is much higher 2010-10-28T09:18:52 the planet with 11 ships 2010-10-28T09:18:56 has a payback time of 3 turns 2010-10-28T09:18:57 or so 2010-10-28T09:19:03 the planet with 50 ships 2010-10-28T09:19:08 has a payback time of almost 11 turns 2010-10-28T09:19:27 and its further away form the planets from which i was launching 2010-10-28T09:19:35 so it will take even longer to payback 2010-10-28T09:20:13 hah i like my neutral sniping code 2010-10-28T09:20:14 http://72.44.46.68/canvas?game_id=747775 2010-10-28T09:20:21 watch turn 34 2010-10-28T09:20:25 *** cfaftw has joined #aichallenge 2010-10-28T09:20:45 and his 56 launch at turn 16 preventing him from sniping the 11? 2010-10-28T09:20:52 yes that too 2010-10-28T09:21:04 and this is why http://72.44.46.68/canvas?game_id=747788 2010-10-28T09:21:07 i don't go for middle planets 2010-10-28T09:21:17 they are almost always suicide unless you can take it safely 2010-10-28T09:21:21 with overwhelming force 2010-10-28T09:22:04 ooh those round maps are interesting, it's hard to decide which way to go so I go both ways and lose a lot :) 2010-10-28T09:23:02 they tend to lead to base swaps too 2010-10-28T09:23:12 i played a map where my lead force was attacking to the right 2010-10-28T09:23:17 while his was attacking to left 2010-10-28T09:23:23 so we were like chasing each other's tail 2010-10-28T09:23:30 quite a funny game to watch 2010-10-28T09:23:51 i have that on map1.txt 2010-10-28T09:24:00 that is a wide circular map 2010-10-28T09:25:13 move 34 is nice especially since you knew somehow to send more than the minimum to snipe 2010-10-28T09:25:22 yes 2010-10-28T09:25:23 i do that 2010-10-28T09:25:30 i calculate when launching at neutrals 2010-10-28T09:25:34 how many more ships can he send 2010-10-28T09:25:43 before mine land 2010-10-28T09:25:49 and i send that many 2010-10-28T09:25:53 instead of bare minimum 2010-10-28T09:26:17 i basically scan all planets of his which are at same distance or less than the planet i launch from 2010-10-28T09:26:27 and add that in as a factor to the ships required 2010-10-28T09:26:38 i calculate their free ships on each of those planets 2010-10-28T09:26:42 and assume that he will send them 2010-10-28T09:26:51 sometimes it can backfire though 2010-10-28T09:26:51 I've got something similar but I find it tricky not to overestimate his possible attack 2010-10-28T09:26:55 yea 2010-10-28T09:27:02 and sometimes i oversend to planets 2010-10-28T09:27:06 which are in the corner 2010-10-28T09:27:06 of the map 2010-10-28T09:27:07 or underestimate 2010-10-28T09:27:20 and then he never counters 2010-10-28T09:27:25 and now i have ships stuck in one corner 2010-10-28T09:27:30 which have nowhere to go 2010-10-28T09:27:45 bhasker_ like that 68 on the right side? 2010-10-28T09:27:52 *** mceier has joined #aichallenge 2010-10-28T09:28:23 let me check 2010-10-28T09:28:38 I mean the bottom right planet 2010-10-28T09:28:58 oh no thats because i don't move ships to staging 2010-10-28T09:29:02 if an enemy is closer than my staging 2010-10-28T09:29:14 i just accumulate ships 2010-10-28T09:29:19 till an attack can be launched 2010-10-28T09:29:28 it prevents snipes on edge planets 2010-10-28T09:29:35 if you keep moving ships to staging 2010-10-28T09:29:45 its a very easy target 2010-10-28T09:31:37 *** vizier has quit IRC (Quit: Page closed) 2010-10-28T09:32:09 hmm but at turn 27 on the viewer it looks like your staging area is closer 2010-10-28T09:33:15 *** Azrathud has quit IRC (Ping timeout: 252 seconds) 2010-10-28T09:33:57 my staging is at the planet 139 ships on nit 2010-10-28T09:34:09 which is almost the same distance as the enemy planet 2010-10-28T09:34:18 with 80 ships 2010-10-28T09:34:31 its not exact but the basic idea is it takes me 10 turns say to get to staging 2010-10-28T09:34:33 and enemy is at 11 2010-10-28T09:34:48 or if we are even equidistance 2010-10-28T09:34:52 staging/enemy 2010-10-28T09:34:55 then i lose 2010-10-28T09:34:59 cause i launch a turn later 2010-10-28T09:35:07 so even if i could counter it 2010-10-28T09:35:08 i would lose 2010-10-28T09:35:51 and mostly i use a strategy of cover, that is i forward only if the angle between src_planet -> staging ->closeEnemy > 120 2010-10-28T09:35:58 makes sense 2010-10-28T09:36:12 so if the angle is <120 i don't forward 2010-10-28T09:36:29 thats why you will see my bot at times has ships accumulating in more than 1 spot 2010-10-28T09:36:32 sometimes even 3-4 2010-10-28T09:36:47 and each of these spots can pretty much support the planets behind them 2010-10-28T09:36:53 I have an angle calc in mine too, works well 2010-10-28T09:38:11 I'm using 135 instead of 120, just trial and error to find the right value I guess 2010-10-28T09:38:29 bhasker_, your bot is sounding eerily similar to mine, now :) 2010-10-28T09:38:45 heh 2010-10-28T09:38:51 except yours in the top 10 2010-10-28T09:38:52 :p 2010-10-28T09:38:54 mine can't seem to break 50 2010-10-28T09:38:56 i don't have any angle calc :| 2010-10-28T09:38:57 i should do that 2010-10-28T09:39:14 hmm maybe is hould give 135 a shot 2010-10-28T09:39:18 oh well 2010-10-28T09:39:24 bhasker_, not anymore! the good bots have finally come out 2010-10-28T09:39:28 my bot is at 25 now 2010-10-28T09:39:33 which one is yours cfaftw? 2010-10-28T09:39:46 cfaftw :P 2010-10-28T09:39:52 oh k 2010-10-28T09:40:01 yea the new bots are very very god 2010-10-28T09:40:03 good* 2010-10-28T09:40:23 I don't see cfaftw on tcp ratings 2010-10-28T09:40:25 on the official server, cfaftw 2010-10-28T09:40:29 ah 2010-10-28T09:40:37 on the tcp, i've been masquerading as jb.exp.* 2010-10-28T09:40:47 ah that's a good one 2010-10-28T09:40:48 *** hellman has joined #aichallenge 2010-10-28T09:40:57 the * being the last 4 chars of my hg revision hash 2010-10-28T09:41:22 *** amstan has joined #aichallenge 2010-10-28T09:41:22 *** ChanServ sets mode: +o amstan 2010-10-28T09:43:48 ah yea the jb.exp seem to be doing quite well on tcp 2010-10-28T09:44:21 bhasker_, it does ok... there's a bunch of them because it's experimental. i'll try some new idea then let it run for a bit to see how well it works in practice 2010-10-28T09:45:07 ah k 2010-10-28T09:45:23 @rankings 2010-10-28T09:45:24 bhasker_: Top 10 players: bocsimacko(3947), rsergio(3889), GreenTea(3874), rs5147(3815), dmj111(3802), shangas(3761), felixcoto(3752), protocolocon(3749), Accoun(3732), sin_sun(3732) 2010-10-28T09:45:34 nice mega1's bot is rocking it 2010-10-28T09:45:43 which bot is his? 2010-10-28T09:45:49 I wonder how cccf keeps from playing itself on the tcp, maybe just alternates in a script? 2010-10-28T09:45:52 wow 1 loss 2010-10-28T09:46:04 the #1 bot is his 2010-10-28T09:46:09 bocsimacko 2010-10-28T09:46:15 oh! 2010-10-28T09:46:25 i love that bot 2010-10-28T09:46:44 i was wondering whose it was 2010-10-28T09:46:53 Accoun is pretty good too 2010-10-28T09:47:53 Is anybody blogging about their progress? 2010-10-28T09:48:12 does the contestbot know something like @tcprankings 2010-10-28T09:48:13 ? 2010-10-28T09:48:50 eburnette: i was thinking about it but my development has kindof stalled lately 2010-10-28T09:49:47 bhasker_, have you considered adjusting your possible staging angle based on the target planet risk? 2010-10-28T09:49:59 hmm nope 2010-10-28T09:50:14 bhasker_, i currently use an angle to cap neutrals... i've thought about adjusting the angle based on the risk of capping any one neutral 2010-10-28T09:50:29 not sure what you mean by capping neutrals? 2010-10-28T09:50:34 oh capture* 2010-10-28T09:50:37 for example, a neutral that will most likely be sniped, i will accept a larger routing angle 2010-10-28T09:50:38 haha, eyah 2010-10-28T09:50:41 yeah* 2010-10-28T09:50:56 hmm 2010-10-28T09:51:07 but neutrals that i know i can cap and hold, i would disallow routing 2010-10-28T09:51:17 the only problem is that i currently have no sane way of evaluating risk 2010-10-28T09:51:18 -_- 2010-10-28T09:54:03 *** boegel has joined #aichallenge 2010-10-28T09:56:12 hehe yea :-\ 2010-10-28T09:56:21 you have to factor in how many free ships enemy has 2010-10-28T09:56:29 which is quite hard to get right 2010-10-28T09:56:46 when you say disallow routing 2010-10-28T09:56:55 do you route ships to neutral via your planets? 2010-10-28T09:59:13 lol this one completely screwed up obligatory shampoo's bot 2010-10-28T09:59:17 http://72.44.46.68/canvas?game_id=748195 2010-10-28T09:59:20 check this game out 2010-10-28T09:59:36 not sure what my bot was thinking 2010-10-28T09:59:41 but it went after a neutral in his base 2010-10-28T09:59:46 and actually won it 2010-10-28T09:59:47 lol 2010-10-28T10:00:03 bhasker_, i have a method to calculate if a planet is "in line" with two planets 2010-10-28T10:00:13 oooh 2010-10-28T10:00:20 yea i thought of doing that 2010-10-28T10:00:21 "in line" means it is both between them and the angle is less than some amount 2010-10-28T10:00:29 so you just send 2x ships to first one? 2010-10-28T10:00:36 rather a+b to first one 2010-10-28T10:00:39 yes 2010-10-28T10:00:45 yea i thought of doing that 2010-10-28T10:00:51 except the way my code is structured 2010-10-28T10:00:57 it goes through destinations one at a time 2010-10-28T10:00:59 then when a+b arrive at that planet, that planet can decide what to do 2010-10-28T10:01:14 so its hard for me to combine launches 2010-10-28T10:01:27 bhasker_, my code isn't pretty :P 2010-10-28T10:02:09 neither is mine but i try lol 2010-10-28T10:02:15 actually i think i might be able to do it 2010-10-28T10:02:31 except i will have to track if i have launched ships for a neutral in line 2010-10-28T10:02:35 previously 2010-10-28T10:02:43 and just send the required ships to that neutral instead 2010-10-28T10:02:45 although, i don't actually use angle 2010-10-28T10:02:51 i use distance 2010-10-28T10:02:55 ah k 2010-10-28T10:03:16 probably better to use angle though since if the second planet is closer 2010-10-28T10:03:21 it maybe better to split it 2010-10-28T10:03:24 a/b 2010-10-28T10:03:33 unless you always launch to closer first 2010-10-28T10:03:34 yeah i'm thinking that now 2010-10-28T10:03:53 i do always launch to closer first 2010-10-28T10:03:59 ah 2010-10-28T10:04:03 i don't do that 2010-10-28T10:04:14 i prioritize my queue by value of the planet 2010-10-28T10:04:17 *** Cromulent has joined #aichallenge 2010-10-28T10:04:21 and i always go for high value planets 2010-10-28T10:04:25 even if they are further away 2010-10-28T10:04:32 i don't order by distance 2010-10-28T10:04:40 not for attacking enemy or neutrals 2010-10-28T10:04:59 the problem that happens with that is your bot tends to prefer defending than attacking 2010-10-28T10:05:06 so on maps where enemy is far away 2010-10-28T10:05:10 my bot would launch an attack 2010-10-28T10:05:13 enemy will counter 2010-10-28T10:05:20 and now my bot will decide it should send all ships back 2010-10-28T10:05:24 when they get there 2010-10-28T10:05:32 haha 2010-10-28T10:05:33 instead of taking over enemies base 2010-10-28T10:05:45 so i got rid of all distance based preference 2010-10-28T10:05:49 for attacks 2010-10-28T10:05:55 i rather use payoff points 2010-10-28T10:05:56 to deicde 2010-10-28T10:06:00 which neutral to go for 2010-10-28T10:06:10 and i go for neutrals which have better payoff points 2010-10-28T10:06:11 first 2010-10-28T10:06:30 i like my neutral sniping code 2010-10-28T10:06:31 http://72.44.46.68/canvas?game_id=748230 2010-10-28T10:06:35 watch turn 32 2010-10-28T10:06:56 and 56 2010-10-28T10:06:59 57 rather 2010-10-28T10:07:38 i'm trying to figure out how to use a payoffs now... found a book linked to on wikipedia about computational game theory 2010-10-28T10:08:27 i have a crude way of doing it 2010-10-28T10:09:06 999999999/NumShips/growth + log(2^dist) 2010-10-28T10:09:09 something like that 2010-10-28T10:09:27 999999999/(NumShips/growth + log(2^dist)) 2010-10-28T10:09:28 rather 2010-10-28T10:09:42 works reasonably well 2010-10-28T10:09:53 larger values indicate better payoff points 2010-10-28T10:10:11 i divide because i hate negating values to put in priority queue ^ 2010-10-28T10:10:57 *** amstan has quit IRC (Ping timeout: 245 seconds) 2010-10-28T10:12:04 so you do use distance? 2010-10-28T10:12:22 Top 10 players: bocsimacko(3949), rsergio(3891), GreenTea(3876), rs5147(3817), dmj111(3804), shangas(3763), felixcoto(3754), protocolocon(3751), Accoun(3734), sin_sun(3734) 2010-10-28T10:12:24 not for deciding planet Value 2010-10-28T10:12:31 planetValue is a simple metric 2010-10-28T10:12:43 this payoff point is used to decide which neutrals to prefer at a particular turn 2010-10-28T10:12:45 mostly 2010-10-28T10:12:49 not for enemy planets 2010-10-28T10:12:49 etc 2010-10-28T10:12:53 and not for ordering the queue 2010-10-28T10:13:08 i use it to discard neutrals if there are neutrals later in the queue with better payoff points 2010-10-28T10:13:23 mmm i'm using knapsack for determining neutrals 2010-10-28T10:13:30 i use knapsack only at first turn 2010-10-28T10:13:33 and growthrate/distance as a metric 2010-10-28T10:13:45 i was using all that earlier 2010-10-28T10:13:50 payoff points are a better metric 2010-10-28T10:13:55 and i compare my payoff against enemies 2010-10-28T10:13:59 to decide which planets to prefer 2010-10-28T10:14:02 or ignore 2010-10-28T10:14:17 so if the enemy has a better payoff point 2010-10-28T10:14:19 i ignore it 2010-10-28T10:14:23 unless enemy is attacking it 2010-10-28T10:14:28 in which case i see if i can snipe 2010-10-28T10:14:37 mm 2010-10-28T10:15:13 and what is your current rank if i may ask? ie how well does it wrok? :) 2010-10-28T10:15:19 i actually ignore planets with equal or worse payoff points unless enemy is attacking it 2010-10-28T10:15:33 on TCP its doing at around 411 elo 2010-10-28T10:15:34 right now 2010-10-28T10:15:45 on official my rank varies between 60-100 2010-10-28T10:15:45 oh nice 2010-10-28T10:16:13 right now its over 150 because i uploaded a more recent version 2010-10-28T10:16:18 and it has got 5 games:-\ 2010-10-28T10:16:20 since yesterday 2010-10-28T10:16:22 *** Cold-Phoenix has joined #aichallenge 2010-10-28T10:16:35 yeah the cloud has vanished into thin air -_- 2010-10-28T10:17:22 just watched that game you linked 2010-10-28T10:17:57 tinkering with my bot is a bad idea 2010-10-28T10:17:58 sigh 2010-10-28T10:18:01 i can't resist it though 2010-10-28T10:18:06 and now i seem to have screwed up something 2010-10-28T10:18:11 haha 2010-10-28T10:18:13 it happens 2010-10-28T10:19:34 in other news ebur 120>135 2010-10-28T10:19:40 135 is too narrow 2010-10-28T10:19:53 my ships just pile up 2010-10-28T10:20:00 on individual planets 2010-10-28T10:22:19 *** JamesMG has joined #aichallenge 2010-10-28T10:22:25 just started up my bot on TCP 2010-10-28T10:22:30 bhasker_: so you use a priority queue per planet that belongs to you holding neutrals? 2010-10-28T10:22:31 ah 2010-10-28T10:22:37 :X 2010-10-28T10:22:49 not sure your what you mean Error 2010-10-28T10:22:57 i use a priority queue for all planets 2010-10-28T10:23:00 i build a single queue 2010-10-28T10:23:03 and walk through it 2010-10-28T10:23:19 priority being a simple metric 2010-10-28T10:23:20 well since in your payoff formula you include distance 2010-10-28T10:23:29 ah that comes into play only for neutrals 2010-10-28T10:23:32 or is that your average distance? 2010-10-28T10:23:46 i use average distance checks as well 2010-10-28T10:23:55 my shouldAcquireneutral is like a big bunch of checks 2010-10-28T10:24:01 one after the other 2010-10-28T10:24:03 hehe 2010-10-28T10:24:38 lamo lost a game by 2 ships 2010-10-28T10:24:40 http://72.44.46.68/canvas?game_id=748435 2010-10-28T10:24:41 :-\ 2010-10-28T10:24:59 obligatory shampoo... my bot seems to be in sync with his game timer 2010-10-28T10:25:18 haha, not cool 2010-10-28T10:25:30 http://72.44.46.68/getplayer?player=jb.exp.2b8c 2010-10-28T10:25:30 i think he runs multiple instances 2010-10-28T10:25:39 i get like a bunch of matches with him 2010-10-28T10:27:06 bhasker_, i see what you mean about distances causing your bot to play defensively 2010-10-28T10:27:11 mine is pretty defensive 2010-10-28T10:27:15 yea 2010-10-28T10:27:19 so i took all that out 2010-10-28T10:27:26 with distances used to order planets 2010-10-28T10:27:29 you will always defend 2010-10-28T10:27:31 before attack 2010-10-28T10:27:42 which is not always the best 2010-10-28T10:28:47 http://72.44.46.68/canvas?game_id=748498 2010-10-28T10:29:02 watch my bot lose... after taking that center neutral like a fool 2010-10-28T10:29:34 yup 2010-10-28T10:29:42 also i think your routing code needs a fixt 2010-10-28T10:29:44 (probably would have lost, anyway, because he capped that other 5 planet) 2010-10-28T10:29:50 the corner 2 growth planet 2010-10-28T10:29:56 is routing via the bottom planet 2010-10-28T10:31:21 hmm 2010-10-28T10:31:56 i'll look into it later 2010-10-28T10:32:40 i'wait 2010-10-28T10:32:45 which bottom planet 2010-10-28T10:35:57 the bottom top 2010-10-28T10:36:00 with 2 growth 2010-10-28T10:36:14 it should just route to the big one in center 2010-10-28T10:36:18 why route down and then up 2010-10-28T10:36:41 same for the corner ones 2010-10-28T10:36:49 they are routing to the planet one below the middle 2010-10-28T10:36:56 more efficient to route to the middle straight? 2010-10-28T10:37:04 *** Azrathud has joined #aichallenge 2010-10-28T10:37:12 yeah i see 2010-10-28T10:37:24 wow, things have got ridiculous 2010-10-28T10:37:28 i have 25 separate files 2010-10-28T10:37:41 lol 2010-10-28T10:37:52 i started pulling a bunch of things from places for my next idea 2010-10-28T10:38:08 it's all separated into folders etc. like planetwars, gametheory, maths etc. 2010-10-28T10:38:53 what i really need to do is fix my move selection aha 2010-10-28T10:49:10 *** tapwater has joined #aichallenge 2010-10-28T10:51:24 *** narnach_mb has quit IRC (Quit: Leaving.) 2010-10-28T10:54:20 *** aerique has quit IRC (Quit: ...) 2010-10-28T10:56:44 *** seisatsu has quit IRC (Quit: Leaving) 2010-10-28T10:59:57 *** McLeopold has quit IRC (Quit: Leaving) 2010-10-28T11:00:18 *** hornairs has joined #aichallenge 2010-10-28T11:00:51 *** Sarin has joined #aichallenge 2010-10-28T11:01:55 *** Azrathud has quit IRC (Ping timeout: 265 seconds) 2010-10-28T11:02:44 *** Azrathud has joined #aichallenge 2010-10-28T11:03:54 *** airhorns has quit IRC (Ping timeout: 240 seconds) 2010-10-28T11:05:50 *** jezuz has joined #aichallenge 2010-10-28T11:05:52 that's so weird, the bot i have just went from winning in 70 moves to 100 consistently 2010-10-28T11:06:08 despite me not changing anything related to moves and i ctrl z'd anything recent as well 2010-10-28T11:07:30 i find it kinda funny that on the 'using the tools' page theres text to the effect of 'dont worry about using the cmd its easy!', and this is a programming contest 2010-10-28T11:07:35 *** lavalamp_ has joined #aichallenge 2010-10-28T11:12:24 Top 10 players: bocsimacko(3938), rsergio(3882), GreenTea(3866), rs5147(3825), dmj111(3795), shangas(3753), felixcoto(3745), protocolocon(3742), Accoun(3727), sin_sun(3725) 2010-10-28T11:14:58 weird, my bot is rated 50-100 Elo higher if I don't run the control bot... I guess it does badly against my old bot... 2010-10-28T11:17:49 *** Azrathud has quit IRC (Ping timeout: 276 seconds) 2010-10-28T11:24:37 *** boegel has quit IRC (Ping timeout: 264 seconds) 2010-10-28T11:28:09 *** narnach has joined #aichallenge 2010-10-28T11:28:09 *** narnach is now known as narnach_mb 2010-10-28T11:32:51 *** Cold-Phoenix has quit IRC () 2010-10-28T11:41:23 *** Olathe has quit IRC (Read error: Connection reset by peer) 2010-10-28T11:42:08 *** Olathe has joined #aichallenge 2010-10-28T11:42:43 bocsimacko makes the rest of us look like amateurs... 2010-10-28T11:42:57 I am an amateur so it's ok 2010-10-28T11:43:05 he is using a real AI language, not like the rest of your 2010-10-28T11:43:13 of you 2010-10-28T11:44:25 meh, i'ma beat him soon enough 2010-10-28T11:44:47 I have a hard time believing the language is why his bot is so good 2010-10-28T11:45:06 that's like saying ansel adams was awesome because he had an expensive camera... 2010-10-28T11:47:51 that's no excuse, he had to haul glass plates and turn of the century gear up to those vantage points 2010-10-28T11:47:51 http://stackoverflow.com/questions/130475/why-is-lisp-used-for-ai 2010-10-28T11:52:27 *** narnach_mb has quit IRC (Quit: Leaving.) 2010-10-28T11:54:30 AI has mostly become machine learning and symbolic computation is less relevant. 2010-10-28T11:55:10 What's good about Lisp is rapid development and speedy execution. 2010-10-28T11:55:13 and power 2010-10-28T11:57:14 I won't have to consider porting from python to c++ near the end of the contest. 2010-10-28T11:57:40 ah ok, mega1 does your bot learn? 2010-10-28T11:57:50 no 2010-10-28T11:57:54 ok 2010-10-28T11:57:54 I do 2010-10-28T11:57:57 ^^ 2010-10-28T11:58:14 *** irchs has joined #aichallenge 2010-10-28T11:58:32 is python having issues, or just execution speed? 2010-10-28T11:58:45 "...so what you see here in bocksimacko-0.50, is actually a binary dump of mega1's brain." 2010-10-28T12:00:00 I guess python is also good for rapid development, but lacks a true interactive dev env (correct me) 2010-10-28T12:00:41 ...and I learn by watching bocsimacko play... 2010-10-28T12:00:52 i use C because im familiar with it xD 2010-10-28T12:07:26 *** irchs has quit IRC (Quit: irchs) 2010-10-28T12:08:02 raedwulf: learn c++ 2010-10-28T12:08:06 it's like c but better 2010-10-28T12:09:12 learn go... it's like c but better... without all the gotchas that c++ has ;) 2010-10-28T12:12:23 Top 10 players: bocsimacko(3936), rsergio(3892), GreenTea(3870), rs5147(3846), dmj111(3805), felixcoto(3757), shangas(3747), protocolocon(3741), Accoun(3727), deepblue(3717) 2010-10-28T12:15:50 antimatroid1: i can write in C++, but just feel a lot more comfortable with C 2010-10-28T12:17:34 *** amstan has joined #aichallenge 2010-10-28T12:17:35 *** ChanServ sets mode: +o amstan 2010-10-28T12:18:05 also, i can list quite a few reasons i dislike C++ too lol :P xD, but i won't start a rant 2010-10-28T12:21:31 mega1, what do you mean by a "true" interactive dev environment? 2010-10-28T12:21:31 *** narnach has joined #aichallenge 2010-10-28T12:21:31 *** narnach is now known as narnach_mb 2010-10-28T12:21:45 *** narnach has joined #aichallenge 2010-10-28T12:22:08 *** Mjothvitnir has joined #aichallenge 2010-10-28T12:22:52 *** irchs has joined #aichallenge 2010-10-28T12:23:12 cfaftw: Lisp development happens in a long running lisp process that's rarely restarted 2010-10-28T12:23:41 once I load the program I'm tinkering with I usually just recompile functions, files. 2010-10-28T12:24:00 evaluate bits here and there to test 2010-10-28T12:24:26 in python you have the interactive console, and you can certainly redefine classes, methods, and reimport modules 2010-10-28T12:24:43 that I know 2010-10-28T12:24:44 might not be as fluid as the LISP shell you're using, though 2010-10-28T12:24:58 i definitely don't do my work in the python interactive shell :) 2010-10-28T12:25:01 it's an emacs mode (SLIME) 2010-10-28T12:25:14 it's more like an ide 2010-10-28T12:25:35 oh nice 2010-10-28T12:27:12 how long have you been programming with lisp? 2010-10-28T12:28:09 8 years 2010-10-28T12:28:35 haha, that's a long time... is it your favorite language? 2010-10-28T12:29:00 by a margin 2010-10-28T12:29:19 what's the runner up? 2010-10-28T12:30:32 *** bhasker_ has quit IRC (Quit: bhasker_) 2010-10-28T12:30:34 i've only used lisp briefly for one class at uni; a data structures class 2010-10-28T12:30:45 also used ocaml in another semester 2010-10-28T12:30:49 the rest are so far down the ladder I cannot even answer 2010-10-28T12:31:22 I had lisp classes at uni too, it just didn't make any sense at all. 2010-10-28T12:31:23 mega1, oh... i thought you meant by a small margin 2010-10-28T12:31:34 no :-) 2010-10-28T12:31:37 *** dhartmei has joined #aichallenge 2010-10-28T12:31:47 so what got you hooked? 2010-10-28T12:32:14 mostly that it doesn't get in my way as much as other languages 2010-10-28T12:32:14 mega1: im learning haskell 2010-10-28T12:32:33 and i learnt scheme in my first year 2010-10-28T12:32:47 i liked using ocaml better, but i think that was solely because i thought the syntax was prettier 2010-10-28T12:32:56 after getting hooked I learnt to value the interactive development environment, the object system, etc. 2010-10-28T12:33:04 lisp definitely has kickass compilers though 2010-10-28T12:33:51 im unfortunately not of the church of emacs, but use vim :P 2010-10-28T12:33:58 being an sbcl dev I tend to agree (-; 2010-10-28T12:33:59 and quite familiar with gdb 2010-10-28T12:34:07 mega1: ^^ 2010-10-28T12:34:54 syntax is what puts people off 2010-10-28T12:35:10 so C programming is almost second nature to me, i don't miss OO features that much; but i had a placement for a year with C# 2010-10-28T12:35:18 i quite liked scheme 2010-10-28T12:35:32 im told its very close to lisp 2010-10-28T12:35:43 I'm lazy too, so when confronted by something new I look for a reason to discard it. 2010-10-28T12:36:10 Syntax happens to be this reason for a lot of people without realizing why it is so. 2010-10-28T12:36:18 raedwulf: you really should check out go, it is similar to C in a lot of ways... 2010-10-28T12:36:22 yes, scheme is _a_ lisp 2010-10-28T12:36:31 lavalamp_: i've had a look, it looks nice 2010-10-28T12:36:44 but of course, C definitely is a bit more production ready atm ^^ 2010-10-28T12:36:58 unfortunately, yeah 2010-10-28T12:36:59 i just realised, you can potentially tempt the other bot into mistakenly sending the wrong amount to a neutral near them 2010-10-28T12:37:13 if you don't send enough to take the neutral, they might send to get there the same time you do seeing it as cheaper 2010-10-28T12:37:14 did a bit of vala programming too, thats quite nice, but it has some flaws (which may be fixed now) 2010-10-28T12:37:23 but not account properly for your fleets arriving at the same time 2010-10-28T12:37:40 antimatroid1: but isn't this exploiting flaws in other bots 2010-10-28T12:37:53 that's part of the contest :P 2010-10-28T12:38:01 not actually beating a more optimal opponent? 2010-10-28T12:38:18 hmmm ok 2010-10-28T12:38:24 i tend to disagree 2010-10-28T12:38:30 i wouldn't be surprised if some of the better bots got fooled by that 2010-10-28T12:38:39 mine certainly would i think 2010-10-28T12:38:39 im getting close, i've got two function stubs i need to fill 2010-10-28T12:38:42 xD 2010-10-28T12:38:50 but how do you gain from that? 2010-10-28T12:38:58 they waste ships 2010-10-28T12:39:10 so do you though 2010-10-28T12:39:15 you could send 3 ships and potentially fool them if you did it strategically enough 2010-10-28T12:39:30 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-10-28T12:39:45 oh I see you're counting on them not doing 3 way battles correctly 2010-10-28T12:40:01 no, they can do that alright 2010-10-28T12:40:02 lavalamp_ really loves go :P 2010-10-28T12:40:18 yup ^^ 2010-10-28T12:40:24 but if you do future states, they are going to have the incoming fleet arrive and update the ships yeah? 2010-10-28T12:40:31 then see it as cheaper than what it really is 2010-10-28T12:40:38 or maybe that counts as not doing it right 2010-10-28T12:40:46 at which point, i'm currently not doing it right 2010-10-28T12:40:47 yeah that counts as a math fail ^^ 2010-10-28T12:41:07 I don't think that will fool my bot. 2010-10-28T12:41:31 Of course my bot already wastes ships on bad neutral planets without even taking them, so it's not even necessary 2010-10-28T12:41:42 cfaftw: what about you? do you think you'd account for that properly? 2010-10-28T12:42:09 i may decide to not take a neutral after sending some ships there 2010-10-28T12:42:12 but it's not really a waste 2010-10-28T12:42:21 when i made them they were expected to be the best move 2010-10-28T12:42:35 that's how it happens to my bot, too... 2010-10-28T12:43:00 but it's really terrible, it's like burning money... 2010-10-28T12:43:03 *** Mjothvitnir has quit IRC (Quit: Page closed) 2010-10-28T12:43:22 antimatroid1, my bot looks at future costs before selecting a planet... so it wouldn't see it as cheaper until the fleet you sent is at least 1 closer to that neutral than one of my planets 2010-10-28T12:44:09 *** McLeopold has joined #aichallenge 2010-10-28T12:45:44 My bot keeps climbing on the tcp server, but I don't know why... I didn't change it... 2010-10-28T12:46:04 lavalamp_: i bet its a cosmic ray that flipped the magic bit in your code 2010-10-28T12:46:40 damn I hoped it was spontaneously learning 2010-10-28T12:47:50 lol 2010-10-28T12:48:28 *** Tobu has joined #aichallenge 2010-10-28T12:49:54 *** amstan has quit IRC (Ping timeout: 240 seconds) 2010-10-28T12:51:49 ah I see, I have a decent record vs FireworkBot 2010-10-28T12:52:58 *** yasith has joined #aichallenge 2010-10-28T12:53:31 *** bhasker has joined #aichallenge 2010-10-28T12:56:13 *** amstan has joined #aichallenge 2010-10-28T12:56:13 *** ChanServ sets mode: +o amstan 2010-10-28T12:56:36 *** irchs has quit IRC (Quit: irchs) 2010-10-28T13:02:06 *** Itkovian has quit IRC (Quit: Itkovian) 2010-10-28T13:02:53 *** Itkovian has joined #aichallenge 2010-10-28T13:04:03 *** Cromulent has quit IRC (Quit: Cromulent) 2010-10-28T13:12:23 Top 10 players: bocsimacko(3920), rsergio(3880), GreenTea(3860), rs5147(3831), dmj111(3789), felixcoto(3749), protocolocon(3734), shangas(3732), Accoun(3713), sin_sun(3710) 2010-10-28T13:14:45 *** Cyndre has quit IRC (Read error: Connection reset by peer) 2010-10-28T13:15:11 *** Cyndre has joined #aichallenge 2010-10-28T13:15:58 *** Arthur_ has quit IRC (Ping timeout: 255 seconds) 2010-10-28T13:17:56 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]) 2010-10-28T13:17:57 *** Florian95 has quit IRC (Remote host closed the connection) 2010-10-28T13:18:13 *** Arthur has joined #aichallenge 2010-10-28T13:21:44 *** irchs has joined #aichallenge 2010-10-28T13:22:33 *** irchs has quit IRC (Client Quit) 2010-10-28T13:32:54 found this book linked from wikipedia: http://www.masfoundations.org/mas.pdf 2010-10-28T13:33:34 reading through it trying to grasp game theory... it's a pretty intense book, a good amount is over my head 2010-10-28T13:35:53 so i screwed up my bot royally 2010-10-28T13:35:59 need to roll back all kinds of changes 2010-10-28T13:36:23 so cfaft question 2010-10-28T13:36:35 do you think its a wise idea to recompute available ships after every launch 2010-10-28T13:36:50 *** fawek has joined #aichallenge 2010-10-28T13:37:08 available ships for you or the enemy? 2010-10-28T13:37:15 for both 2010-10-28T13:37:28 since everytime i launch 2010-10-28T13:37:34 enemy's available ships changes too 2010-10-28T13:38:02 bhasker, you know, i hadn't considered recomputing them... that could be a good idea 2010-10-28T13:38:11 yea i am trying it out 2010-10-28T13:38:15 its just that its taking too long 2010-10-28T13:38:20 i need to optimize it 2010-10-28T13:38:28 *** Sarin has quit IRC (Read error: Connection reset by peer) 2010-10-28T13:38:32 though i am not doing it after every launch right now 2010-10-28T13:38:38 more like after i handle each target planet 2010-10-28T13:38:52 cause doing it after each launch can make me have to cancel launches 2010-10-28T13:39:02 and then recompute with the orders cancelled and rolled back 2010-10-28T13:39:05 which may be ideal 2010-10-28T13:39:08 but i am thinking 2010-10-28T13:39:26 i think computing at start of turn and keeping it static through the launch sequence is not a good idea 2010-10-28T13:40:32 mine is static now, but you're right... once one planet launches ships, the available ships on other planets are affected 2010-10-28T13:40:45 yup 2010-10-28T13:40:55 hm I don't pay attention to their availible ships 2010-10-28T13:40:58 yes! another layer of complication! 2010-10-28T13:41:13 heh 2010-10-28T13:41:20 time to bring the python profiler out 2010-10-28T13:41:25 my turn time shot up 2010-10-28T13:41:34 as i started recomputing per planet 2010-10-28T13:42:48 there's probably an efficient way to redistribute the available ships without doing the whole calculation all over again 2010-10-28T13:43:12 if you know what planets are dependent on the planet that launched ships 2010-10-28T13:43:32 true 2010-10-28T13:43:52 basically having to do after every launch will require me to update my state map 2010-10-28T13:43:55 recalculate 2010-10-28T13:43:58 then try and launch 2010-10-28T13:44:01 if i don't have ships 2010-10-28T13:44:04 all that has to be undone 2010-10-28T13:44:06 undo state 2010-10-28T13:44:12 recalculate ships 2010-10-28T13:44:16 rinse repeat 2010-10-28T13:44:26 my history object keeps track of availible ships along with arrivials and departures; to do it, with orders, I give it a list of "held" ships, ships that the order depends on 2010-10-28T13:45:24 bhasker: you just described my performace problems 2010-10-28T13:45:27 *** Arthur has quit IRC (Disconnected by services) 2010-10-28T13:45:50 *** Arthur_ has joined #aichallenge 2010-10-28T13:45:58 yea doing it in python is probably also killing the bot 2010-10-28T13:46:03 will also kill the bot rather 2010-10-28T13:46:09 anyone familiar with delt0r's tool? 2010-10-28T13:46:14 you guys must do more complicated sending logic than I do... 2010-10-28T13:46:24 mcleopold, what tool? 2010-10-28T13:46:31 Bot Manager 2000 2010-10-28T13:46:47 Yeah I'm using it 2010-10-28T13:46:50 It rocks 2010-10-28T13:46:58 It confuses me... 2010-10-28T13:47:27 maybe not the best interface but a lot better than nothing... what part of it are you talking about? 2010-10-28T13:47:40 mcleopold, link"? 2010-10-28T13:47:57 http://www.mabs.at/ewing/JBotManager.jar 2010-10-28T13:48:30 on the tcp tab, what is reps? 2010-10-28T13:48:30 thanks 2010-10-28T13:48:46 I think it's how many games you want it to play 2010-10-28T13:48:47 number of games to play on tcp? 2010-10-28T13:48:51 ok 2010-10-28T13:49:19 on the combat tab, my map dropdown is pulling from the wrong folder 2010-10-28T13:49:24 yeah although there might be a bug, once it stopped I couldn't get it started again, probably need to restart the program 2010-10-28T13:49:34 yeah mine does that too 2010-10-28T13:50:38 is there a way to set the map folder? 2010-10-28T13:51:39 I think it's just buggy... if you type in a folder sometimes it remembers it 2010-10-28T13:51:54 you can type a path in the box 2010-10-28T13:51:55 press enter 2010-10-28T13:52:00 ah 2010-10-28T13:52:29 I'm trying combat, but I'm sure my bot is broken. Will it display who timed out? 2010-10-28T13:52:32 *** amstan has quit IRC (Ping timeout: 265 seconds) 2010-10-28T13:53:30 yeah their ships go to 0 in the graph 2010-10-28T13:54:51 wow this tool does rock 2010-10-28T13:54:58 can someone link what this tool does? 2010-10-28T13:55:08 let's you play bots locally or on the TCP 2010-10-28T13:55:14 I can't get it to rock yet. Just sissor... 2010-10-28T13:55:16 http://ai-contest.com/forum/viewtopic.php?f=18&t=751 2010-10-28T13:55:50 it looks like my bot needs to be a full command "python MyBot.py" 2010-10-28T13:55:55 it does 2010-10-28T13:55:57 full path, too 2010-10-28T13:56:03 yeah 2010-10-28T13:56:09 well... full path to the script 2010-10-28T13:56:37 the bot setup is a little annoying because my bot takes half a dozen args depending on what I'm doing... :/ 2010-10-28T13:56:50 I got ragebot working... 2010-10-28T13:57:21 lamo loss to bockismacko by 8 ships 2010-10-28T13:57:22 :-\ 2010-10-28T13:57:22 http://72.44.46.68/canvas?game_id=750817 2010-10-28T13:57:28 wow this is fun 2010-10-28T13:57:40 thanks for bringing this to my attention 2010-10-28T13:57:45 i should read the forums -_- 2010-10-28T14:00:40 lavalamp looks like we have a match 2010-10-28T14:01:27 am I choosing crappy neutrals? 2010-10-28T14:01:42 it would be great if this tool could automatically run matches over a range of maps and report the total win/loss percentage 2010-10-28T14:02:01 let me see 2010-10-28T14:02:08 I think there's another one on the forum somewhere that will do that... or maybe it's just a script 2010-10-28T14:02:42 it's a C# app 2010-10-28T14:03:02 runs yours against other bots of yours and/or the sample bots on whatever map or all maps 2010-10-28T14:03:28 *** p4p4 has joined #aichallenge 2010-10-28T14:03:48 but does it have the slick game rendering that Bot Manager 2000 has? 2010-10-28T14:04:21 bhasker: yeah I wasted 140+ ships on those two +1 neutrals 2010-10-28T14:04:52 nah, it's visualization is kinda crappy 2010-10-28T14:04:57 cfaftw: do you run python? I can't get it to work with BM2000 2010-10-28T14:05:04 yeah, python 2010-10-28T14:05:09 and i'm in windows 2010-10-28T14:05:14 me too 2010-10-28T14:05:23 arg! 2010-10-28T14:05:28 python is in my path env 2010-10-28T14:05:36 same here 2010-10-28T14:05:48 the command runs on the command line, but BM2000 doesn't run it 2010-10-28T14:06:02 so i just have it do: python "c:\dir1\dir2\bot.py" 2010-10-28T14:07:07 first turn crash, my log files aren't even being produced 2010-10-28T14:07:24 syntax error somewhere? 2010-10-28T14:07:44 no, the bot runs outside of the tool just fine 2010-10-28T14:07:54 do dir1 and dir2 have spaces in them? 2010-10-28T14:07:55 did you put the path in quotes? 2010-10-28T14:09:09 ah looks like my bot played its 500 games and is done for the day 2010-10-28T14:09:31 you do you have it wait between games or does it just do em back to back? 2010-10-28T14:11:24 it just does 'em back to back 2010-10-28T14:11:28 Hmm... close and open and it works. 2010-10-28T14:12:08 *** Bobng has joined #aichallenge 2010-10-28T14:12:23 Top 10 players: bocsimacko(3931), rsergio(3889), GreenTea(3869), rs5147(3841), dmj111(3798), felixcoto(3758), protocolocon(3744), shangas(3741), Accoun(3724), sin_sun(3719) 2010-10-28T14:13:40 like magic 2010-10-28T14:14:05 the magic of programs that write other programs 2010-10-28T14:14:32 man this is depressing 2010-10-28T14:14:46 my current devel bot is generally worse than my official bot 2010-10-28T14:14:50 er, s/write/run/ 2010-10-28T14:15:12 not by a lot... but marginally 2010-10-28T14:15:16 Okay, just played a tcp game with it... and lost. 2010-10-28T14:15:24 Now, how do the replays work? 2010-10-28T14:16:21 I thought my bot was worse than my old bot because it only beats it 30% of the time... but running it on tcp I see that it does better against other bots... 2010-10-28T14:16:33 ah, I see a file called tcpGame0 2010-10-28T14:16:38 Haven't used the replay feature 2010-10-28T14:18:12 McLeopold, is this the game you just played? http://72.44.46.68/canvas?game_id=751066 2010-10-28T14:18:45 yep 2010-10-28T14:19:26 pretty pathetic, huh 2010-10-28T14:19:38 that... wasn't the best neutral choice 2010-10-28T14:19:50 I think I broke something... 2010-10-28T14:20:24 yeah, you let him take those central planets without much fight 2010-10-28T14:20:48 you might think about changing the name each time you try a new bot on tcp, so it doesn't carry over the old rating 2010-10-28T14:21:13 I'm not interested in ratings 2010-10-28T14:21:23 Until Dec :) 2010-10-28T14:21:39 contest is over by then :) 2010-10-28T14:21:45 Nov 27? 2010-10-28T14:21:58 that's what i hear 2010-10-28T14:22:15 yea 2010-10-28T14:24:02 *** RainCT has joined #aichallenge 2010-10-28T14:24:46 *** Naktibalda has joined #aichallenge 2010-10-28T14:25:18 I think JBotManager has a bug. 2010-10-28T14:25:51 turn 1 is actually showing the fleet movements of turn 1 and 2 2010-10-28T14:26:28 *** boegel has joined #aichallenge 2010-10-28T14:26:47 I noticed its turn numbers are 0 based 2010-10-28T14:27:06 yeah, that seems off a little. 2010-10-28T14:27:28 It looks like I launched 3 fleets to start, but I only did 1 2010-10-28T14:28:03 huh 2010-10-28T14:28:38 *** narnach has quit IRC (Quit: Leaving.) 2010-10-28T14:29:10 *** cfaftw has quit IRC (Quit: going home...) 2010-10-28T14:32:12 *** boegel has quit IRC (Ping timeout: 245 seconds) 2010-10-28T14:38:06 *** choas has joined #aichallenge 2010-10-28T14:42:14 *** boegel has joined #aichallenge 2010-10-28T14:45:47 *** amstan has joined #aichallenge 2010-10-28T14:45:47 *** ChanServ sets mode: +o amstan 2010-10-28T14:49:54 anyone know of a game engine besides PlayGame.jar? 2010-10-28T14:50:44 maybe one that outputs some good error messages? 2010-10-28T14:51:14 i modified the actual python engine to run on client machines 2010-10-28T14:52:03 http://bitbucket.org/zannick/planet-wars/src 2010-10-28T14:53:11 (war2) 2010-10-28T14:53:18 i don't know how useful you want the error messages 2010-10-28T14:53:37 Zannick: excellent 2010-10-28T14:53:46 great for testing! 2010-10-28T14:53:50 yep 2010-10-28T14:54:01 because i was trying to run the java stuff on my VPS 2010-10-28T14:54:10 and you know how java likes to have memory, memory and more memory 2010-10-28T14:54:10 *** area has quit IRC (Read error: Connection reset by peer) 2010-10-28T14:54:16 the bots just don't run lol 2010-10-28T14:54:37 i did that because i wanted to run against the actual engine 2010-10-28T14:54:46 this was when the java engine had some bugs 2010-10-28T14:54:59 so, if I send an invalid order, will it say that when it ends the game? 2010-10-28T14:55:10 *** area has joined #aichallenge 2010-10-28T14:55:32 yeah 2010-10-28T14:55:45 the engine is writing that to stderr 2010-10-28T14:56:41 *** perror has quit IRC (Quit: Bye all !) 2010-10-28T14:58:02 Zannick: is it the same syntax as PlayGame.jar? 2010-10-28T14:58:07 add some custom error message, "Bot 1 sucks, it can't even do proper orders." 2010-10-28T14:58:14 McLeopold: no 2010-10-28T14:58:35 ./war2 [bot3, ...] [--gui] 2010-10-28T14:58:55 there's a README with the info 2010-10-28T14:59:28 oh, i just realized you might have to alter the #! if your python isn't in /usr/local/bin 2010-10-28T15:02:07 ooo what does --gui do... bring up a browser with visualizer.js? 2010-10-28T15:02:32 *** boegel has quit IRC (Quit: *poof!*) 2010-10-28T15:02:48 no, just ShowGame.jar :P 2010-10-28T15:03:05 noes :S 2010-10-28T15:03:24 you need a standalone python gui ^^ 2010-10-28T15:03:31 i should 2010-10-28T15:03:46 it probably isn't too difficult to hack up 2010-10-28T15:04:24 to write out an html file with visualizer.js, i mean 2010-10-28T15:05:11 it doesn't do so good in windows 2010-10-28T15:05:24 no fnctl, but you don't need it anyway 2010-10-28T15:05:49 ah 2010-10-28T15:05:59 Zannick: I already have the visualizer.js tool 2010-10-28T15:06:35 http://github.com/DanielVF/Planet-Wars-Canvas-Visualizer 2010-10-28T15:07:04 *** Bobng has quit IRC (Read error: No route to host) 2010-10-28T15:07:08 oh, hey, then that's really easy 2010-10-28T15:07:38 you could just change the part of war2 that invokes ShowGame to invoke visualize_localy.py 2010-10-28T15:08:01 *** Bobng has joined #aichallenge 2010-10-28T15:08:35 File "C:\Python27\lib\subprocess.py", line 882, in _execute_child 2010-10-28T15:08:35 startupinfo) 2010-10-28T15:08:35 WindowsError: [Error 193] %1 is not a valid Win32 application 2010-10-28T15:10:29 what's your commandline? 2010-10-28T15:10:46 and, the BDFL would not approved of the 2 space indents... 2010-10-28T15:11:05 i just left those in from the server's version 2010-10-28T15:11:09 don't look at me :P 2010-10-28T15:11:56 python planet-wars\war2 7 ScotBot5.py AntBot.py 2010-10-28T15:12:23 Top 10 players: bocsimacko(3932), rsergio(3879), GreenTea(3858), rs5147(3833), dmj111(3787), felixcoto(3748), shangas(3734), protocolocon(3722), Accoun(3717), sin_sun(3710) 2010-10-28T15:12:46 are the bots executable? 2010-10-28T15:12:53 . ./ScotBot5.py False 2010-10-28T15:12:55 yes 2010-10-28T15:13:12 that's working_directory, shell_command and security_on in the __init__ function 2010-10-28T15:13:40 i'm not sure 2010-10-28T15:13:51 oh, the "./" does not work in windows 2010-10-28T15:13:54 i don't really know what is happening on windows 2010-10-28T15:13:59 ah 2010-10-28T15:14:29 nobody knows whats happening on windows 2010-10-28T15:14:58 you thought "minefields from 'nam were bad... try windows." 2010-10-28T15:18:32 okay, I imported platform and checked platform.system() == "Windows" 2010-10-28T15:18:54 now I get ... 2010-10-28T15:18:55 os.kill(self.command_process.pid, signal.SIGKILL) 2010-10-28T15:18:55 AttributeError: 'module' object has no attribute 'SIGKILL' 2010-10-28T15:19:36 But I did make it 2 turns... 2010-10-28T15:20:54 yeah, windows doesn't have sigkill 2010-10-28T15:21:26 I don't think windows has signals, period 2010-10-28T15:21:29 lol! 2010-10-28T15:21:36 well, it has the signal module 2010-10-28T15:21:48 really? weird 2010-10-28T15:21:50 i think it's just a wrapper around thier calls 2010-10-28T15:22:03 so you might be able to use signal.CTRL_C_EVENT 2010-10-28T15:22:17 i would do dir(signal) to see what SIG* values are there 2010-10-28T15:22:35 http://docs.python.org/library/signal.html 2010-10-28T15:22:51 amstan, is janzert still stuck without power in a 3rd world country surrounded by malaria infested mosquitoes? 2010-10-28T15:23:39 idk 2010-10-28T15:25:30 *** Sarin has joined #aichallenge 2010-10-28T15:27:33 Zannick: CTRL_C_EVENT is 2.7, the server runs 2.5 2010-10-28T15:28:00 and the script i wrote is 2.6 2010-10-28T15:28:21 you can hack around the files i supplied, they won't affect the server :P 2010-10-28T15:28:41 okay, that worked 2010-10-28T15:30:15 cool 2010-10-28T15:30:28 except after player 2 dies, it doesn't return 2010-10-28T15:30:48 I've got the game results, but no command prompt 2010-10-28T15:30:57 *** Accoun has quit IRC () 2010-10-28T15:31:29 it printed the game results and is still running? 2010-10-28T15:31:30 I need a tougher kill signal 2010-10-28T15:31:59 yeah, ctrl-c is sigint, not sigkill 2010-10-28T15:32:12 i'm not sure if ctrl_break is sigkill 2010-10-28T15:32:42 also possible python wants a sys.exit(0) at the end of main 2010-10-28T15:32:55 *** cfaftw has joined #aichallenge 2010-10-28T15:34:43 ctrl_break doesn't work either 2010-10-28T15:34:52 sigterm gets me an access is denied 2010-10-28T15:35:32 so wait, did the bots not exit or did the engine not exit? 2010-10-28T15:36:15 if the latter, ctrl_c is probably fine, just add sys.exit(0) after main in war2 2010-10-28T15:36:24 at the end of main, i mean 2010-10-28T15:36:24 the bot crashes, when I end task it, the engine then prints "end monitor" 2010-10-28T15:36:30 oh 2010-10-28T15:36:50 so the bot is just not getting the signal to terminate 2010-10-28T15:39:33 mcleopold, are you still trying to get bot manager 2000 to work? 2010-10-28T15:39:51 I did get it to work 2010-10-28T15:39:57 except for human mode 2010-10-28T15:40:20 oh wow didn't even notice human mode 2010-10-28T15:42:12 not sure what it's supposed to do... can't get it to do anything 2010-10-28T15:42:13 i think the signal wasn't sent because it didn't reach sadbox.kill() 2010-10-28T15:42:14 no start button 2010-10-28T15:42:27 oh there is 2010-10-28T15:42:30 my window wasn't big enough 2010-10-28T15:42:35 that function would have set is_alive to false, which would have then caused "end monitor" to be printed 2010-10-28T15:42:55 just got beat by rage bot in human mode. :D 2010-10-28T15:43:03 *** Accoun has joined #aichallenge 2010-10-28T15:43:20 maybe kill is blocking? 2010-10-28T15:43:28 nifty 2010-10-28T15:44:24 oh, could be 2010-10-28T15:48:21 not that we need more engine implementations... but I have been using a version for my own testing. Its python, but I haven't tried it in windows. 2010-10-28T15:48:55 I have some fixes to check in, and a tournament script too. Probably will check them in later today. 2010-10-28T15:49:35 subprocess.Popen("taskkill /F /T /PID %i"%self.command_process.pid , shell=True) 2010-10-28T15:49:54 dmj111: I'll check it out 2010-10-28T15:51:51 *** boegel has joined #aichallenge 2010-10-28T15:53:26 *** Todd200 has joined #aichallenge 2010-10-28T15:53:33 anyone here? 2010-10-28T15:53:48 im ticked at the ai contest servers :( 2010-10-28T15:54:32 they are runnign so slow that suddenly my bot cant start within the 1 second limit 2010-10-28T15:55:36 *** amstan has quit IRC (Quit: Konversation terminated!) 2010-10-28T15:55:48 *** amstan has joined #aichallenge 2010-10-28T15:55:48 *** ChanServ sets mode: +o amstan 2010-10-28T15:56:14 amstan = admin? 2010-10-28T15:56:23 who summoned me? 2010-10-28T15:56:50 me 2010-10-28T15:57:04 i don't think theres much to do except wait for cloud servers to come back up 2010-10-28T15:57:22 slow servers? you should wait 2010-10-28T15:57:36 my bot, which is a simple mod to starter package, just timed out 2010-10-28T15:57:39 :( 2010-10-28T15:57:47 what language? 2010-10-28T15:57:50 python 2010-10-28T15:58:01 you sure it's not an exception? 2010-10-28T15:58:06 did you test it locally? 2010-10-28T15:58:15 it won three games last night 2010-10-28T15:58:38 so.. you'll get timeout for any kind of error 2010-10-28T15:58:44 including exception 2010-10-28T15:58:48 test your bot 2010-10-28T15:58:58 there's tons of other people including McLeopold that use python 2010-10-28T15:59:11 http://ai-contest.com/profile.php?user_id=6765 2010-10-28T15:59:13 wait, was it dmj? 2010-10-28T15:59:28 it won 3 times last night 2010-10-28T15:59:28 amstan when will the almighty cloud be back up? :) 2010-10-28T15:59:41 Todd200: dude.. you lost one game, chill down 2010-10-28T15:59:48 Error323: define almighty? 2010-10-28T16:00:01 *** Gorovkhaia2 has quit IRC (Ping timeout: 272 seconds) 2010-10-28T16:00:16 gpm >= 60 2010-10-28T16:00:20 :P 2010-10-28T16:00:38 *** Gorovkhaia2 has joined #aichallenge 2010-10-28T16:00:45 what he said 2010-10-28T16:01:21 i found a cool glitch with thecoring yesterday, it put me in 22 place 2010-10-28T16:01:25 the scoring 2010-10-28T16:01:34 *** Sarin has quit IRC (Quit: Verlassend) 2010-10-28T16:04:18 *** amstan has quit IRC (Ping timeout: 252 seconds) 2010-10-28T16:05:02 well thats rude 2010-10-28T16:05:04 ^^ 2010-10-28T16:07:52 *** eburnette_ has quit IRC (Ping timeout: 265 seconds) 2010-10-28T16:11:07 lol 2010-10-28T16:11:10 *** McLeopold has quit IRC (Ping timeout: 255 seconds) 2010-10-28T16:12:24 Top 10 players: bocsimacko(3943), rsergio(3891), GreenTea(3869), rs5147(3839), dmj111(3799), felixcoto(3777), shangas(3736), sin_sun(3733), protocolocon(3724), Accoun(3713) 2010-10-28T16:17:24 *** narnach has joined #aichallenge 2010-10-28T16:22:00 noob question, how do i run a sim with a python bot? i only know how to for java 2010-10-28T16:22:23 I wish I could see more moves in http://72.44.46.68/canvas?game_id=750817 , it looks like something interesting was about to happen 2010-10-28T16:23:48 how do you use the tcp with python bot? 2010-10-28T16:27:39 *** davidjliu has joined #aichallenge 2010-10-28T16:29:41 Todd200: The same way you would with any other bot 2010-10-28T16:36:57 goddamned 2010-10-28T16:37:03 infinite loop in my code 2010-10-28T16:37:07 and I can't find it 2010-10-28T16:37:08 urgh 2010-10-28T16:37:25 just wait until it finishes 2010-10-28T16:37:41 boegel: put a unique logging statement in each loop :P 2010-10-28T16:38:18 you can't kill it and get a stack trace? 2010-10-28T16:38:40 if I ctrl+4 a go executable it dies and prints a stack trace 2010-10-28T16:39:13 unless you're on windows in which case nvm and gl 2010-10-28T16:39:20 *** narnach has quit IRC (Quit: Leaving.) 2010-10-28T16:39:41 *** narnach has joined #aichallenge 2010-10-28T16:42:18 tcp wont work 2010-10-28T16:42:34 you have to make the python script an executable 2010-10-28T16:42:47 ? 2010-10-28T16:42:55 it says waiting for connection 2010-10-28T16:43:04 chmod +x MyBot.py and add a line to the top of the file #! 2010-10-28T16:43:19 maybe you're not connecting to the server, then 2010-10-28T16:43:25 with that stop "connection reset by peer" 2010-10-28T16:43:28 will 2010-10-28T16:43:41 what is the server listed in tcp.c? 2010-10-28T16:43:47 tcp.exe 213.3.30.106 995 Alberto -p password C:\Python22\python.exe MyBot.py 2010-10-28T16:44:16 that doesn't look like the right ip 2010-10-28T16:44:25 @tcp 2010-10-28T16:44:26 JamesMG: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/ 2010-10-28T16:44:29 what ip is it? 2010-10-28T16:44:32 that noe 2010-10-28T16:44:33 also the arguments are bad, yeah. 2010-10-28T16:44:33 one* 2010-10-28T16:45:11 tcp only likes one-word arguments for your bot 2010-10-28T16:45:24 lavalamp_: it's Haskell 2010-10-28T16:45:27 ? 2010-10-28T16:45:28 so you should make a MyBot.bat file that contains 'C:\Python22\python.exe Mybot.py' 2010-10-28T16:45:30 wrap the execution of your bot in a script or do what Zannick said 2010-10-28T16:45:37 yeah, that 2010-10-28T16:45:46 in fact, I'm not sure the loop is *infinite* (can you ever be?), but it's taking way too long anyways 2010-10-28T16:45:50 the thing i said before is for unix, it probably won't work 2010-10-28T16:45:50 im a noob at thsi 2010-10-28T16:46:14 what should i do 2010-10-28T16:46:17 im using the exe 2010-10-28T16:46:47 right, and so tcp will think that exe is your bot 2010-10-28T16:47:07 actually it seems to pass on args for me 2010-10-28T16:47:13 hey its working 2010-10-28T16:47:17 that command line should work 2010-10-28T16:47:17 really? then i guess nevermind 2010-10-28T16:47:18 changed the ip 2010-10-28T16:47:43 must be a newer version, I had to put mine in a bash script that just says "mono bot.exe" 2010-10-28T16:47:52 how do i view it? 2010-10-28T16:48:47 go to the link contest bot gave and find yourself on the ratings page after you play a game 2010-10-28T16:48:53 http://72.44.46.68/getplayer?player=Alberto 2010-10-28T16:49:07 wow, it says synatx error in the code that sends go 2010-10-28T16:49:09 if you are using Alberto as the username 2010-10-28T16:51:04 *** Azrathud has joined #aichallenge 2010-10-28T16:51:07 it keeps failing 2010-10-28T16:51:13 EOF while parsing 2010-10-28T16:51:34 eroor in main 2010-10-28T16:51:41 that's in your code 2010-10-28T16:51:43 error in main() 2010-10-28T16:52:02 currentline = raw_input() 2010-10-28T16:52:23 thats not my code, its provided 2010-10-28T16:52:47 sorry, I don't know python, you'll have to catch one of those guys 2010-10-28T16:53:16 def main(): 2010-10-28T16:53:16 map_data = '' 2010-10-28T16:53:16 while(True): 2010-10-28T16:53:16 current_line = raw_input() 2010-10-28T16:53:16 if len(current_line) >= 2 and current_line.startswith("go"): 2010-10-28T16:53:16 pw = PlanetWars(map_data) 2010-10-28T16:53:16 DoTurn(pw) 2010-10-28T16:53:17 pw.FinishTurn() 2010-10-28T16:53:17 map_data = '' 2010-10-28T16:53:18 else: 2010-10-28T16:53:18 map_data += current_line + '\n' 2010-10-28T16:53:21 don't. do that. 2010-10-28T16:53:32 whoops 2010-10-28T16:53:41 use pastebin 2010-10-28T16:53:50 so, it's probably your tcp connection breaking 2010-10-28T16:53:52 Todd200: 2010-10-28T16:53:56 wrap raw_input() 2010-10-28T16:54:01 in a try/except then 2010-10-28T16:54:10 and catch the exception and exit I guess. 2010-10-28T16:54:20 and the pipe is getting eof while you're trying to read the input from it 2010-10-28T16:54:44 not much point in capturing the exception though. 2010-10-28T16:54:55 you have to quit, might as well let the exception terminate it. 2010-10-28T16:55:07 *** yasith has quit IRC (Ping timeout: 272 seconds) 2010-10-28T16:55:20 tell me in 14 year old language how to fix it 2010-10-28T16:55:35 im probally the youngest person in thsi comp who can wrie a bot 2010-10-28T16:55:47 comp, write 2010-10-28T16:56:18 Todd200: you will learn much quicker if you search the python docs/google yourself for the answer 2010-10-28T16:56:32 *** boegel has quit IRC (Quit: Leaving) 2010-10-28T16:56:48 takes a bit more time, but the payoff is way higher 2010-10-28T16:56:58 *** RainCT has quit IRC (Read error: Connection reset by peer) 2010-10-28T16:56:58 i know, python (well enough to write an ok bot) 2010-10-28T16:57:09 *** p4p4 has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2010-10-28T16:58:03 search for "python exception handling" or sth like that. 2010-10-28T16:59:13 i just turned off my firewall for a test 2010-10-28T16:59:49 no good 2010-10-28T17:00:30 well the part of the code thats crashing is the part that submits my move, whih i did not write 2010-10-28T17:02:47 * Todd200 trys example 2010-10-28T17:03:56 wow im a failer 2010-10-28T17:04:03 wrong verison of python 2010-10-28T17:04:12 it happens, don't worry 2010-10-28T17:04:18 heh, yeah, python 2.2 2010-10-28T17:05:26 i had 2.2 and 2.7 on my computer 2010-10-28T17:05:47 might want to get 2.5, I hear that's what the server runs 2010-10-28T17:05:52 *** Azrathud has quit IRC (Ping timeout: 265 seconds) 2010-10-28T17:06:44 2.7 same error 2010-10-28T17:06:47 trying 2.5 2010-10-28T17:07:26 *** choas has quit IRC (Quit: leaving) 2010-10-28T17:08:11 *** cfaftw has quit IRC (Quit: Leaving) 2010-10-28T17:09:31 *** yasith has joined #aichallenge 2010-10-28T17:10:17 * Todd200 crosses fingers 2010-10-28T17:11:35 fail 2010-10-28T17:11:46 same error! 2010-10-28T17:11:55 on the starter package 2010-10-28T17:12:03 which i havent touched 2010-10-28T17:12:22 *** eburnette_ has joined #aichallenge 2010-10-28T17:12:26 Top 10 players: bocsimacko(3959), GreenTea(3904), rsergio(3902), rs5147(3838), dmj111(3796), felixcoto(3779), shangas(3737), protocolocon(3736), sin_sun(3729), Accoun(3703) 2010-10-28T17:12:42 *** Todd200 has quit IRC (Quit: ('i give up!')) 2010-10-28T17:13:12 my advice: don't use windows 2010-10-28T17:13:47 *** delt0r___ has quit IRC (Read error: Operation timed out) 2010-10-28T17:15:05 *** delt0r___ has joined #aichallenge 2010-10-28T17:18:00 *** McLeopold has joined #aichallenge 2010-10-28T17:20:17 *** nibalize1 has joined #aichallenge 2010-10-28T17:26:53 *** jaspervdj has quit IRC (*.net *.split) 2010-10-28T17:26:53 *** javagamer_cpp has quit IRC (*.net *.split) 2010-10-28T17:26:53 *** j3camero has quit IRC (*.net *.split) 2010-10-28T17:26:54 *** nibalizer has quit IRC (*.net *.split) 2010-10-28T17:26:54 *** jmcarthur has quit IRC (*.net *.split) 2010-10-28T17:26:54 *** lovebot_ has quit IRC (*.net *.split) 2010-10-28T17:26:55 *** raedwulf has quit IRC (*.net *.split) 2010-10-28T17:26:55 *** samulihs has quit IRC (*.net *.split) 2010-10-28T17:27:35 *** jaspervdj has joined #aichallenge 2010-10-28T17:27:35 *** javagamer_cpp has joined #aichallenge 2010-10-28T17:27:35 *** j3camero has joined #aichallenge 2010-10-28T17:27:36 *** jmcarthur has joined #aichallenge 2010-10-28T17:27:36 *** lovebot_ has joined #aichallenge 2010-10-28T17:27:36 *** raedwulf has joined #aichallenge 2010-10-28T17:27:36 *** samulihs has joined #aichallenge 2010-10-28T17:29:02 whoa what just happened? 2010-10-28T17:29:17 IRC netsplit 2010-10-28T17:32:02 *** jaspervdj has quit IRC (Quit: NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN) 2010-10-28T17:33:44 http://en.wikipedia.org/wiki/Netsplit 2010-10-28T17:34:24 it looks like contestbot was attempting to take over the channel... 2010-10-28T17:35:10 it's learning ambition 2010-10-28T17:35:41 we're teaching it well 2010-10-28T17:36:55 having the contest bot there makes me want to have it remember random things... 2010-10-28T17:37:28 it's my theory that it's ingesting all the gameplay visualizations we paste, and will have it's own submission on Nov 27 2010-10-28T17:37:29 Or get everyone to do a "later tell j3camero something" so he gets 100 messages when he does show up again 2010-10-28T17:40:25 *** Palmik has quit IRC (Remote host closed the connection) 2010-10-28T17:50:52 *** Bobng has quit IRC (Quit: Leaving) 2010-10-28T17:53:15 @eightball will golang ever get installed on the server? 2010-10-28T17:53:15 lavalamp_: The outlook is good. 2010-10-28T17:53:18 yay 2010-10-28T17:54:10 @eightball were you lying to lavalamp_? 2010-10-28T17:54:10 JamesMG: I doubt it very much. 2010-10-28T17:54:14 wow. 2010-10-28T17:54:14 "ever" != "before the contest ends" 2010-10-28T17:54:38 ;) 2010-10-28T17:54:52 @eightball before the contest ends? 2010-10-28T17:54:52 lavalamp_: The outlook is good. 2010-10-28T17:55:00 wow... 2010-10-28T17:55:04 *** sigh has joined #aichallenge 2010-10-28T17:55:13 it almost seems like it knows... things... 2010-10-28T17:55:17 @eightball Will the answer to this question be equivalent to no? 2010-10-28T17:55:17 Zannick: No clue. 2010-10-28T17:55:26 @eightball Will the answer to this question be equivalent to yes? 2010-10-28T17:55:26 Zannick: It is possible. 2010-10-28T17:55:30 @eightball Will the answer to this question be equivalent to no? 2010-10-28T17:55:30 Zannick: The answer is def-- oooh! shiny thing! 2010-10-28T17:55:36 hahaha 2010-10-28T17:55:44 i have a suspicion he is alternating 2010-10-28T17:55:47 haha that was a pretty convincing sequence 2010-10-28T17:56:17 oh, no clue =~ i don't know 2010-10-28T17:56:19 okay 2010-10-28T17:56:21 :P 2010-10-28T17:56:39 you confused it 2010-10-28T17:57:02 @eightball P=NP? 2010-10-28T17:57:02 davidjliu: Come again? 2010-10-28T17:59:18 lmao 2010-10-28T18:00:48 *** Azrathud has joined #aichallenge 2010-10-28T18:01:03 JamesMG: You're up to -400's I see :) 2010-10-28T18:01:12 oh yeah? 2010-10-28T18:01:40 -415 right now... 2010-10-28T18:02:09 that's my 3 day old code 2010-10-28T18:02:14 I reverted a bunch of times 2010-10-28T18:02:16 oh, lol 2010-10-28T18:02:17 haha 2010-10-28T18:02:32 I discovered my bounded knapsack solution was flawed 2010-10-28T18:02:39 so I've been reworking it today 2010-10-28T18:02:56 *** davidjliu_ has joined #aichallenge 2010-10-28T18:03:09 I haven't figured out how to work the knapsack into my code 2010-10-28T18:03:45 I haven't figured out how to code a bounded knapsack solution I can actually recall the planets from :) 2010-10-28T18:04:21 *** davidjliu has quit IRC (Ping timeout: 265 seconds) 2010-10-28T18:04:45 I'm not convinced it's important... my bot has mid-game planet choice issues and I don't think knapsack will help that 2010-10-28T18:04:57 mine is having opening move planet choice issues 2010-10-28T18:05:09 for the opening my heuristic seems to work fine 2010-10-28T18:07:08 *** CIA-100 has quit IRC (Read error: Connection reset by peer) 2010-10-28T18:07:15 tweaking this code is changing the number of turns it takes me to win against the sample bots by 50-70 turns 2010-10-28T18:07:32 if I could just get it perfect I think it'd help both my bots a bunch 2010-10-28T18:08:03 my heuristic is pretty simple: 2010-10-28T18:08:34 1 calculate value of the planet (growth * turns left after capturing it) 2010-10-28T18:08:59 2 calculate payoff time (ships required / growth) 2010-10-28T18:09:24 priority = value / payoff time 2010-10-28T18:09:51 so, growth squared * turns / ships ? 2010-10-28T18:09:57 sort planets by priority and try to take them in that order 2010-10-28T18:10:13 I'm not saying it makes sense ^^ 2010-10-28T18:10:31 that's basically my midgame neutral prioritization 2010-10-28T18:10:31 oh, that's not an equation 2010-10-28T18:10:34 ships / growth 2010-10-28T18:10:40 *** Appleman1234 has joined #aichallenge 2010-10-28T18:11:20 hm actually that's not quite what I do exactly... I might have a bug there 2010-10-28T18:11:46 i need to work on my bot :/ 2010-10-28T18:11:55 but i never really want to 2010-10-28T18:12:25 Top 10 players: bocsimacko(3946), GreenTea(3891), rsergio(3889), rs5147(3825), dmj111(3784), felixcoto(3766), protocolocon(3723), shangas(3723), sin_sun(3714), Accoun(3689) 2010-10-28T18:13:13 http://pastebin.com/9BgC0whU 2010-10-28T18:14:23 huh and I just realized why 0 growth neutrals were so attractive before I added that condition... 2010-10-28T18:14:55 *** Olathe has quit IRC (Read error: Connection reset by peer) 2010-10-28T18:15:16 glad I could help :) 2010-10-28T18:15:51 oh well, that whole bit is going to change 2010-10-28T18:16:16 but it's good enough to be ~ +100 2010-10-28T18:19:00 I did the same thing with 0 growth neutrals at first 2010-10-28T18:19:17 now it never considers taking them 2010-10-28T18:19:25 but it will stomp on an enemy hiding on one 2010-10-28T18:19:37 *** d3c1ph3r has joined #aichallenge 2010-10-28T18:19:41 yeah that code only runs for neutrals 2010-10-28T18:20:06 yes, my code also stomps on enemies hiding on neutral planets 2010-10-28T18:20:09 :P 2010-10-28T18:20:31 picking apart my english syntax :( 2010-10-28T18:22:01 *** JamesMG has quit IRC (Quit: Leaving.) 2010-10-28T18:22:19 *** lavalamp_ has quit IRC (Quit: Page closed) 2010-10-28T18:22:22 *** Olathe has joined #aichallenge 2010-10-28T18:22:24 question: why does netbeans highlight a method call yellow saying accessing static field ... -> example -> botEval[b].attackMove botEval highlight yellow which is a local array of objects MyBot which holds 4 variables evaluation score, all moves made, moves counter, and game state 2010-10-28T18:25:21 *** fawek has quit IRC (Read error: Connection reset by peer) 2010-10-28T18:30:16 *** JensTi has joined #aichallenge 2010-10-28T18:34:02 question: why does netbeans highlight a method call yellow saying accessing static field ... -> example -> botEval[b].attackMove botEval highlight yellow which is a local array of objects MyBot which holds 4 variables evaluation score, all moves made, moves counter, and game state 2010-10-28T18:35:23 i think if someone here were able to answer you, they would have 2010-10-28T18:35:35 lol 2010-10-28T18:35:42 im sorry i figured it out anyways 2010-10-28T18:35:47 or you could wait more than 12 minutes 2010-10-28T18:35:51 ah, okay 2010-10-28T18:35:51 i double posted because some new people came in 2010-10-28T18:36:19 it was pretty self explanatory 2010-10-28T18:36:44 i uploaded my bot last night nothing showed up said it was a success .. i uploaded just now again hopefully it changes 2010-10-28T18:36:46 just please don't spam the channel 2010-10-28T18:37:02 what defines spam 2010-10-28T18:37:19 2 msgs within 5 minutes.. or a ddos attack on a irc room 2010-10-28T18:37:37 your message showed up as 5 lines on my client 2010-10-28T18:37:45 lol 2010-10-28T18:37:56 and you asked it twice in 15 minutes 2010-10-28T18:38:11 you using a net book just kidding i realise your screen probably minimized to teh corner 2010-10-28T18:38:11 i'm not calling that spam, i just don't want you to paste it again every time someone new enters 2010-10-28T18:39:42 are you an op or admin? 2010-10-28T18:39:53 op 2010-10-28T18:41:15 thats cool man i see you on here all the time 2010-10-28T18:41:20 i figured you were either one 2010-10-28T18:41:25 what country ? 2010-10-28T18:41:29 i am here all the time :P 2010-10-28T18:41:42 i should be working... >.> 2010-10-28T18:41:46 US, west coast 2010-10-28T18:42:08 me too i had two midterms tomorrow and i rather program my ver 2 of my bot 2010-10-28T18:43:50 *** CIA-38 has joined #aichallenge 2010-10-28T18:44:17 *** pgpaskar_ has quit IRC (Quit: leaving) 2010-10-28T18:44:50 *** pgpaskar_ has joined #aichallenge 2010-10-28T18:45:02 *** mceier has quit IRC (Quit: leaving) 2010-10-28T18:47:41 *** amstan has joined #aichallenge 2010-10-28T18:47:42 *** ChanServ sets mode: +o amstan 2010-10-28T18:55:08 *** yasith has quit IRC (Ping timeout: 252 seconds) 2010-10-28T18:58:04 *** phreeza has quit IRC (Quit: Computer has gone to sleep) 2010-10-28T18:58:37 *** phreeza has joined #aichallenge 2010-10-28T19:08:38 *** d3c1ph3r has quit IRC (Ping timeout: 265 seconds) 2010-10-28T19:08:48 *** antimatroid has joined #aichallenge 2010-10-28T19:09:50 *** yasith has joined #aichallenge 2010-10-28T19:10:55 *** antimatroid1 has quit IRC (Ping timeout: 240 seconds) 2010-10-28T19:12:26 Top 10 players: bocsimacko(3955), GreenTea(3900), rsergio(3898), rs5147(3834), dmj111(3793), felixcoto(3775), protocolocon(3732), shangas(3732), sin_sun(3724), Accoun(3699) 2010-10-28T19:12:48 *** TravisD has joined #aichallenge 2010-10-28T19:13:09 Does anyone know if Groovy is installed on the competition server yet? 2010-10-28T19:13:32 TravisD: what's a groovy command? 2010-10-28T19:13:44 amstan: "groovyc" is the groovy compiler 2010-10-28T19:14:01 nope 2010-10-28T19:14:16 oh well :P 2010-10-28T19:14:33 oh man.. 2010-10-28T19:14:37 zombie mysql process 2010-10-28T19:15:13 there.. 2010-10-28T19:15:16 server much better 2010-10-28T19:16:42 *** TravisD has quit IRC (Read error: Connection reset by peer) 2010-10-28T19:17:01 *** TravisD has joined #aichallenge 2010-10-28T19:18:16 *** TravisD has left #aichallenge 2010-10-28T19:18:22 *** TravisD has joined #aichallenge 2010-10-28T19:26:02 *** amstan_ has joined #aichallenge 2010-10-28T19:26:02 *** ChanServ sets mode: +o amstan_ 2010-10-28T19:26:18 *** amstan has quit IRC (Quit: Konversation terminated!) 2010-10-28T19:33:46 *** amstan_ has quit IRC (Ping timeout: 276 seconds) 2010-10-28T19:53:11 @roulette 2010-10-28T19:53:11 *** McLeopold was kicked by contestbot (BANG!) 2010-10-28T19:53:12 * contestbot reloads and spins the chambers. 2010-10-28T19:59:56 heh he didn't bother rejoining 2010-10-28T20:00:14 @roulette 2010-10-28T20:00:19 @roulette 2010-10-28T20:00:19 antimatroid: *click* 2010-10-28T20:00:20 @roulette 2010-10-28T20:00:21 phreeza: *click* 2010-10-28T20:00:25 @roulette 2010-10-28T20:00:25 phreeza: *click* 2010-10-28T20:00:31 @roulette 2010-10-28T20:00:31 antimatroid: *click* 2010-10-28T20:00:33 @roulette 2010-10-28T20:00:33 *** antimatroid was kicked by contestbot (BANG!) 2010-10-28T20:00:34 * contestbot reloads and spins the chambers. 2010-10-28T20:00:39 *** antimatroid has joined #aichallenge 2010-10-28T20:00:41 owned 2010-10-28T20:00:52 @roulette 2010-10-28T20:00:52 TravisD: *click* 2010-10-28T20:00:56 @roulette 2010-10-28T20:00:56 *** phreeza was kicked by contestbot (BANG!) 2010-10-28T20:00:57 * contestbot reloads and spins the chambers. 2010-10-28T20:09:23 *** yasith has quit IRC (Ping timeout: 252 seconds) 2010-10-28T20:12:26 Top 10 players: bocsimacko(3959), GreenTea(3903), rsergio(3902), rs5147(3838), dmj111(3797), felixcoto(3779), protocolocon(3736), shangas(3735), sin_sun(3727), Accoun(3703) 2010-10-28T20:12:38 @roulette 2010-10-28T20:12:38 TravisD: *click* 2010-10-28T20:12:40 :D 2010-10-28T20:19:57 FireworkBot is awesome 2010-10-28T20:20:35 who created that 2010-10-28T20:22:07 *** yasith has joined #aichallenge 2010-10-28T20:28:23 *** yasith has quit IRC (Ping timeout: 265 seconds) 2010-10-28T20:33:00 *** amstan has joined #aichallenge 2010-10-28T20:33:00 *** ChanServ sets mode: +o amstan 2010-10-28T20:34:20 *** marceloam has joined #aichallenge 2010-10-28T20:34:40 Error323: I don't know, but do you have a link? 2010-10-28T20:35:04 *** cfaftw has joined #aichallenge 2010-10-28T20:35:05 *** hellman has quit IRC (Remote host closed the connection) 2010-10-28T20:35:14 bhasker, what happened to your bot?! 2010-10-28T20:35:17 TravisD: http://72.44.46.68/canvas?game_id=755468 2010-10-28T20:35:26 just looked at it on TCP 2010-10-28T20:35:31 cfaftw: do you know who made the fireworks bot? 2010-10-28T20:35:36 its awesome 2010-10-28T20:35:42 *** Appleman1234 has quit IRC (Read error: Connection reset by peer) 2010-10-28T20:35:42 when i run the command to execute a match on command-line, i'd like to pipe the output ("turn 1 \n turn 2 ... player 1 wins!") onto a file. if i use '> outfile' at the end, it just gets some numbers on it. how do i do that pipe to work correctly? 2010-10-28T20:35:43 i forget 2010-10-28T20:35:53 thanks 2010-10-28T20:36:01 rofl 2010-10-28T20:36:03 hilarious 2010-10-28T20:36:11 Error323, what's awesome about it? 2010-10-28T20:36:44 the fireworks 2010-10-28T20:36:47 :D 2010-10-28T20:36:54 when its winning it shoots fireworks 2010-10-28T20:37:00 ohhhhhhhh 2010-10-28T20:37:00 lol 2010-10-28T20:37:06 holy crap i never noticed 2010-10-28T20:37:10 me neither 2010-10-28T20:37:12 until just now 2010-10-28T20:37:13 i always stopped watching before the end 2010-10-28T20:37:22 but I just love that lol 2010-10-28T20:37:40 *** Zannick has quit IRC (Ping timeout: 265 seconds) 2010-10-28T20:37:43 it screws up at lot in google-chrome on kubuntu 10.04 2010-10-28T20:37:50 toward the end... too much to process 2010-10-28T20:37:54 not sure if it's the code or the browser 2010-10-28T20:38:02 mmm 2010-10-28T20:38:11 same here a bit in the visualizer 2010-10-28T20:38:17 unless you run the sim from the start 2010-10-28T20:38:36 Error323, it just stops at a certain point 2010-10-28T20:38:48 *** Appleman1234 has joined #aichallenge 2010-10-28T20:38:49 Error323: haha pretty awesome 2010-10-28T20:38:53 mmm not here, gentoo with chrome 2010-10-28T20:38:55 fireworkbot is awesome 2010-10-28T20:38:57 it stops at 104 2010-10-28T20:39:02 mega1 should implement that in his bot 2010-10-28T20:39:02 out of 132 2010-10-28T20:39:04 or something like that 2010-10-28T20:39:08 once it's about to win, fireworks! 2010-10-28T20:39:13 yeah pretty nice 2010-10-28T20:39:20 it does a splashback sort of effect 2010-10-28T20:39:34 well, stealing an idea isn't ofc, but hell yeah this is epic xD 2010-10-28T20:39:39 after the first wave of fireworks hit the other planets 2010-10-28T20:39:44 then those go off too 2010-10-28T20:39:44 jep 2010-10-28T20:39:47 it propagates 2010-10-28T20:39:50 yeah 2010-10-28T20:40:35 very artistic 2010-10-28T20:40:36 =) 2010-10-28T20:41:33 it doesn't seem to always do it, which is a shame 2010-10-28T20:41:58 too many changes 2010-10-28T20:42:00 and it got dumber 2010-10-28T20:42:05 i am rolling back the changes 2010-10-28T20:42:08 its doing dumb crap 2010-10-28T20:42:20 Error323: you could probably even do some sort of demoscene sort of graphics 2010-10-28T20:42:20 if I understand it correctly it only works when the last enemy planet still needs to be captured 2010-10-28T20:42:26 yeah 2010-10-28T20:42:28 you know how they do awesome graphics out of just 256 bytes of code 2010-10-28T20:42:29 heh 2010-10-28T20:42:37 i was thinking about computing "GG" in text or sth :P 2010-10-28T20:42:43 heheh 2010-10-28T20:42:45 through the planets 2010-10-28T20:42:47 but well 2010-10-28T20:42:47 that'd be ridiculous :D 2010-10-28T20:42:55 lets first build a proper bot :P 2010-10-28T20:43:11 pssh... how do you know the graphics won't win the game? ;) 2010-10-28T20:43:28 if it can figure out how to display GG all the steps leading up to that will require winning! 2010-10-28T20:43:43 haha true 2010-10-28T20:43:53 *** yasith has joined #aichallenge 2010-10-28T20:44:01 Error323: what lang are you doing your bot in 2010-10-28T20:44:18 before there is confusion fireworks bot isn't mine 2010-10-28T20:44:22 c++ 2010-10-28T20:45:04 yeah i know 2010-10-28T20:45:18 but my bot is still bugged >_> 2010-10-28T20:45:31 *** Zannick has joined #aichallenge 2010-10-28T20:46:18 hahaha 2010-10-28T20:46:30 this is the result i got from watching a fireworkbot game: http://imgur.com/AkQyL.png 2010-10-28T20:46:48 LOL 2010-10-28T20:47:54 he wins at breaking visualizers 2010-10-28T20:50:27 cfaftw, hahaha 2010-10-28T20:57:13 *** yasith has quit IRC (Ping timeout: 255 seconds) 2010-10-28T20:59:30 I don't think it was mentioned, but there is a forum post about fireworks bot: http://ai-contest.com/forum/viewtopic.php?f=19&t=1016 2010-10-28T21:00:16 ah its made by voidptr 2010-10-28T21:03:56 is the creator of the visualizer in the room? 2010-10-28T21:04:20 it would be nice/convenient if a bar is put over the graph indicating the current turn 2010-10-28T21:06:38 *** TwoPixelGrid has joined #aichallenge 2010-10-28T21:10:39 *** yasith has joined #aichallenge 2010-10-28T21:12:27 Top 10 players: bocsimacko(3939), GreenTea(3885), rsergio(3884), rs5147(3815), dmj111(3774), felixcoto(3762), shangas(3724), protocolocon(3719), sin_sun(3711), Accoun(3686) 2010-10-28T21:18:47 when i run the command to execute a match on command-line, i'd like to pipe the output ("turn 1 \n turn 2 ... player 1 wins!") onto a file. if i use '> outfile' at the end, it just gets some numbers on it. how do i do that pipe to work correctly? 2010-10-28T21:20:41 &> 2010-10-28T21:20:47 or 2> 2010-10-28T21:20:48 i forget 2010-10-28T21:20:55 which one is stderr 2010-10-28T21:21:56 marceloam, 2> 2010-10-28T21:22:17 cfaftw: do you know why those helping servers are so unreliable 2010-10-28T21:22:25 7gpm since a day now 2010-10-28T21:22:27 ffs 2010-10-28T21:22:28 :/ 2010-10-28T21:22:45 apparently they only get servers allocated when they are cheapest 2010-10-28T21:23:05 not quite sure on the details, but it depends on the cost of the server and how much funds are available 2010-10-28T21:24:08 I hate the fact that google isn't sponsering anything :( 2010-10-28T21:30:32 it is a bit misleading 2010-10-28T21:31:52 that, and so stupid 2010-10-28T21:32:01 i mean its no money at all for them 2010-10-28T21:32:45 cfaftw, thank you! 2010-10-28T21:34:00 i just found it useful today actually :P 2010-10-28T21:34:15 i just put a scholarship application for next year and my uni is the kind to eat that shit up 2010-10-28T21:34:48 nice 2010-10-28T21:34:54 *** Naktibalda has quit IRC (Remote host closed the connection) 2010-10-28T21:37:32 *** hellman has joined #aichallenge 2010-10-28T21:38:57 *** bhasker has quit IRC (Read error: Operation timed out) 2010-10-28T21:39:19 *** seisatsu has joined #aichallenge 2010-10-28T21:41:35 dudes.. it's not servers 2010-10-28T21:41:36 it's admins 2010-10-28T21:41:46 and communication between admins 2010-10-28T21:43:24 *** narnach_mb has quit IRC (Ping timeout: 265 seconds) 2010-10-28T21:43:37 *** narnach has quit IRC (Ping timeout: 276 seconds) 2010-10-28T21:44:23 anyone have any good links for using mixed strategy in practice for computing nash equilibrium? 2010-10-28T21:45:58 *** JensTi has quit IRC (Quit: Leaving.) 2010-10-28T21:47:10 cfaftw: i don't have any good instructional links, but you might like this: http://code.activestate.com/recipes/496825-game-theory-payoff-matrix-solver/ 2010-10-28T21:47:33 Using python to iteratively find the mixed strategy of a payoff matrix. 2010-10-28T21:47:39 dmj111: zero sum game ? 2010-10-28T21:47:51 I believe so. 2010-10-28T21:48:15 Or, yes, I think that is the only kind of game this will solve. 2010-10-28T21:48:18 am not sure how do you see galcon as a zero sum game, could you elaborate more ? 2010-10-28T21:48:18 *** narnach has joined #aichallenge 2010-10-28T21:48:18 *** narnach is now known as narnach_mb 2010-10-28T21:48:31 *** narnach has joined #aichallenge 2010-10-28T21:48:34 well, I mean the solver is for a zero sum game. 2010-10-28T21:49:51 *** hellman has quit IRC (Remote host closed the connection) 2010-10-28T21:50:14 I think galcon is zero sum, with an exponentially large number of strategies. 2010-10-28T21:50:17 *** hellman has joined #aichallenge 2010-10-28T21:51:14 dmj111: honestly am not really experience in game solving/ai, am used to solve classical puzzles with game search trees that are computable or other games where only heuristics is the only possible way of solving it 2010-10-28T21:51:49 Yeah, me either. 2010-10-28T21:53:20 neither am i... but i realize now that a certain bit of probability and prediction are required to make good moves on neutrals 2010-10-28T21:53:37 well... have realized for a while, but i haven't been wanting to tackle it 2010-10-28T21:53:57 I see zero sum (an oversimplification, for sure), as "something good for me is equally bad for my opponent." Galcon fits that bill, I think. 2010-10-28T21:54:40 cfaftw: yes, I think the game tree search assumes to much power for the enemy (knowing my moves), but the game theory approach is , um, harder. 2010-10-28T21:55:30 dmj111, cfaftw: what is your current rank 2010-10-28T21:56:09 on official server, 23 and tied with Simplex 2010-10-28T21:57:20 cfaftw: what is the non official server ? is there more than 1 ? 2010-10-28T21:57:41 @tcp 2010-10-28T21:57:42 dmj111: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/ 2010-10-28T21:58:01 dmj111, i'm just interested in using game theory to pick neutrals, which is a much smaller problem (thanks for the link, btw) 2010-10-28T21:58:01 cfaftw: what are you looking to do 2010-10-28T21:58:04 nullkuhl: there is an unofficial server you can play on over tcp. I'll get the forum link. 2010-10-28T21:58:14 i'm currently setting things up to throw them into an actual zero sum game once i've picked moves 2010-10-28T21:58:35 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-10-28T21:58:41 cfaftw: my neutral selection has lost me several of the stalemate games, so I should do something too.. 2010-10-28T21:58:51 me move selection sucks. 2010-10-28T21:58:58 my*** 2010-10-28T21:59:04 nullkuhl: http://ai-contest.com/forum/viewtopic.php?f=18&t=424 2010-10-28T21:59:52 by "much smaller problem" i mean that it is a much smaller subset of "solving" the entire match 2010-10-28T22:01:19 the only way i think of galcon being a zero-sum game, though, is if you consider it a 3 player game and the points are planet growth rates 2010-10-28T22:01:20 dmj: the person who wrote that code is a troll, cnt and cum aha 2010-10-28T22:01:40 how is it not zero sum? :P 2010-10-28T22:02:42 antimatroid, because you can make moves without affecting the other player (strictly speaking) 2010-10-28T22:03:34 which moves? 2010-10-28T22:03:48 it needs to affect you to count 2010-10-28T22:03:59 otherwise it's a move with payoff 0 to each, still zero sum 2010-10-28T22:04:19 if you redistribute to a better position, that's good for you and bad for them etc. 2010-10-28T22:06:02 well, according to wikipedia's explanation of zero sum, galcon doesn't seem zero sum 2010-10-28T22:06:24 i'm don't know that much about game theory, though 2010-10-28T22:06:30 i* 2010-10-28T22:07:29 * cfaftw goes afk for a second 2010-10-28T22:07:34 heh, it seems like while waiting for the official server to take up his server offer, bogusbrainpower is perpetually playing seven bots on TCP 2010-10-28T22:10:59 what part of the wiki page? 2010-10-28T22:11:24 *** _W_ <_W_!~w@irc.w-wins.com> has quit IRC (Read error: Operation timed out) 2010-10-28T22:12:27 Top 10 players: bocsimacko(3938), GreenTea(3884), rsergio(3882), rs5147(3813), dmj111(3772), felixcoto(3761), shangas(3722), protocolocon(3719), sin_sun(3706), Accoun(3685) 2010-10-28T22:12:36 cfaftw: I think if you consider a "strategy" as the set of all moves you would make over the course of the entire game, given a set of game states, then it looks more zero sum. If you win by N ships, the opponent loses by N ships. 2010-10-28T22:13:22 Non zero sum, to me, means, that sometimes we both may win, or I may lose, but you don't win by as much as I lose by. 2010-10-28T22:13:30 woo.. too much in that sentence. 2010-10-28T22:13:52 thats wrong anyway, now that I think about it. 2010-10-28T22:14:07 i'm treating strategies as vectors of moves, then the payoff is a function of expected states 2010-10-28T22:14:21 it stays zero sum, i'm pretty sure it's fine to do it as such too 2010-10-28T22:15:30 for the moment i'm going to just use minimax to find ne and pick my favourite, if no saddle points exist i have some other hack in there to pick one 2010-10-28T22:15:31 *** marceloam has quit IRC (Quit: Leaving) 2010-10-28T22:15:38 (i haven't finished this bot yet, it's still being made) 2010-10-28T22:15:48 i'm considering trying to set up some kind of minimax regret thing 2010-10-28T22:15:49 * cfaftw is back 2010-10-28T22:15:55 i should go back to working on mine... 2010-10-28T22:16:05 or something that tries to converge to mixed strategy nash equilibrium 2010-10-28T22:16:15 i am about to ride down to the uni, i'll get working again later 2010-10-28T22:18:12 well i gotta figure something out 2010-10-28T22:18:18 if you did minimax regret, i think you want to run iterated strict dominance (no way you'll have time to do the mixed variant) then do it 2010-10-28T22:18:56 although when we did mixed iterated removal we used fourier motzkin elimination, think simplex would be faster 2010-10-28T22:19:00 but still have huge doubts 2010-10-28T22:19:11 and that wasn't for this, just playing around some other time 2010-10-28T22:19:31 "just playing around", haha 2010-10-28T22:19:56 anyway, back later 2010-10-28T22:20:01 c'ya 2010-10-28T22:20:30 an ai comp on randomly generated large games would be cool 2010-10-28T22:20:42 *** _W_ <_W_!~w@irc.w-wins.com> has joined #aichallenge 2010-10-28T22:22:41 *** _W_ <_W_!~w@irc.w-wins.com> has joined #aichallenge 2010-10-28T22:25:10 *** Mjothvitnir has joined #aichallenge 2010-10-28T22:26:52 *** Mjothvitnir has quit IRC (Client Quit) 2010-10-28T22:27:51 *** seisatsu has quit IRC (Quit: Leaving) 2010-10-28T22:29:25 *** hornairs has joined #aichallenge 2010-10-28T22:29:56 *** hellman has quit IRC (Remote host closed the connection) 2010-10-28T22:32:41 *** contestbot has joined #aichallenge 2010-10-28T22:32:41 *** ChanServ sets mode: +o contestbot 2010-10-28T22:37:57 *** Accoun has quit IRC (Ping timeout: 240 seconds) 2010-10-28T22:40:13 *** Accoun has joined #aichallenge 2010-10-28T22:40:24 *** hellman has joined #aichallenge 2010-10-28T22:40:28 *** dddd has joined #aichallenge 2010-10-28T22:45:50 *** n9thbit has left #aichallenge 2010-10-28T22:55:35 *** perestrelka has quit IRC (Ping timeout: 240 seconds) 2010-10-28T22:58:27 *** cfaftw has quit IRC (Quit: sleep) 2010-10-28T23:01:09 *** virdo has joined #aichallenge 2010-10-28T23:01:54 *** McLeopold has joined #aichallenge 2010-10-28T23:03:31 dmj111: any engine updates? 2010-10-28T23:07:11 *** McLeopold has quit IRC (Quit: McLeopold) 2010-10-28T23:08:52 *** Mathnerd314 has joined #aichallenge 2010-10-28T23:09:33 wonder if mega1 has just caused the possibility of a nash 2010-10-28T23:09:44 all my advantage thus far has been superior state calc but now i gotta actually make a strategy 2010-10-28T23:11:06 *** Azrathud has quit IRC (Ping timeout: 252 seconds) 2010-10-28T23:12:28 Top 10 players: bocsimacko(3960), rsergio(3902), GreenTea(3902), rs5147(3815), dmj111(3793), felixcoto(3787), shangas(3749), protocolocon(3738), Hazard(3709), sin_sun(3706) 2010-10-28T23:20:55 *** Mathnerd314 has quit IRC (Ping timeout: 240 seconds) 2010-10-28T23:22:23 *** davidjliu_ has quit IRC (Ping timeout: 265 seconds) 2010-10-28T23:23:50 *** dddd has quit IRC (Ping timeout: 265 seconds) 2010-10-28T23:24:20 *** bhasker has joined #aichallenge 2010-10-28T23:25:40 *** Appleman1234 has quit IRC (Ping timeout: 240 seconds) 2010-10-28T23:27:35 *** iFire has quit IRC (Read error: Connection reset by peer) 2010-10-28T23:27:54 *** iFire has joined #aichallenge 2010-10-28T23:34:48 *** Azrathud has joined #aichallenge 2010-10-28T23:38:23 *** Appleman1234 has joined #aichallenge 2010-10-28T23:45:55 *** rebelxt_ has joined #aichallenge 2010-10-28T23:48:52 *** Gorovkhaia2 has quit IRC (Ping timeout: 245 seconds) 2010-10-28T23:50:02 *** yasith has joined #aichallenge 2010-10-28T23:54:54 *** matburt has quit IRC (Quit: Lost terminal)