2010-11-30T00:00:36 *** Azrathud has joined #aichallenge 2010-11-30T00:05:44 1st place is looking pretty secured.. second is totally up for grabs though 2010-11-30T00:08:20 *** Azrathud has quit IRC (Ping timeout: 260 seconds) 2010-11-30T00:13:53 *** smloh1 has joined #aichallenge 2010-11-30T00:16:12 *** Mathnerd314 has quit IRC (Quit: ChatZilla 0.9.86-rdmsoft [XULRunner 1.9.2.12/20101026210630]) 2010-11-30T00:16:49 *** bss03 has joined #aichallenge 2010-11-30T00:17:38 So, is there an announce/devel list I can join to help out next year? 2010-11-30T00:19:09 While I can't predict my time availability next year, I definitely want to help with the starter packs, visualizers, and stuff like that. 2010-11-30T00:20:21 http://ai-contest.com/visualizer.php?game_id=9230610 2010-11-30T00:20:28 my send everything move paid off again :) 2010-11-30T00:20:41 this time on the second planet take 2010-11-30T00:21:20 medrimonia: http://ai-contest.com/visualizer.php?game_id=9233021 nice graph :) 2010-11-30T00:27:00 *** Azrathud has joined #aichallenge 2010-11-30T00:31:45 and would you have had success with that in the end? 2010-11-30T00:32:20 Did you have more ships than medrimonia could have produced while yours are in flight? 2010-11-30T00:33:14 yes, he was able to successfully defend all the way 2010-11-30T00:34:12 how's your bot doing antimatroid? 2010-11-30T00:34:20 smloh1: would he have captured medrimonia's central planet was my question 2010-11-30T00:34:52 frontier: medrimonia didn't have centre 2010-11-30T00:34:59 i'm pretty sure i would have held it indefinitely 2010-11-30T00:36:42 what i don't understand is the logic behind the attack, is it some kind of endgame code? if behind, attack hoping enemy will defend suboptimally? 2010-11-30T00:36:51 antimatroid: well, the planet closest to the center. you both have the same planets on your respective side, but it looks to me as if in turn 400 or so, you wouldn't have enough extra ships to actually defeat medrimonia. 2010-11-30T00:39:04 no, but i'd probably still be ahead 2010-11-30T00:39:11 smloh1: If you have enough ships you can slowly decimate the enemy forces while keeping each planet secure. And with even more ships you will finally reach the point where you conquer the enemy's front planet even if growth is equal. It looks all pretty sane to me :) 2010-11-30T00:39:13 expanding better and sustaining == win 2010-11-30T00:40:01 next cutoff change will start in about 20 minutes 2010-11-30T00:40:15 Frontier: i guess i don't usually attack unless the enemy has no chance of defending, and in this case it looked pretty clearly defendable 2010-11-30T00:40:16 and end at the 3020 mark at 5am server time 2010-11-30T00:40:30 is that for starter pack bots? 2010-11-30T00:40:39 yes 2010-11-30T00:40:49 janzert: :) 2010-11-30T00:40:55 i started an arimaa bot yesterday 2010-11-30T00:40:59 although i'm still setting it up 2010-11-30T00:41:15 Great 2010-11-30T00:41:22 gnight all. 2010-11-30T00:41:26 *** davidjliu has quit IRC (Quit: Page closed) 2010-11-30T00:41:30 smloh1: I don't quite understand what you mean by defendable? Is that in turn 200, if you send all ships you don't gain at least 1 ship? 2010-11-30T00:42:42 janz, only cutting off 30% for now? 2010-11-30T00:42:56 Frotnier: we had the same production, he can't wear me down without wearing himself down 2010-11-30T00:42:57 Frontier: i mean that you are able to attack with enough ships that the opponent cannot get enough ships to the planet in time to defend 2010-11-30T00:43:01 i was able to defend any attack he sent 2010-11-30T00:43:11 burny: basically just going to just beyond the starter bots for the moment 2010-11-30T00:43:40 antimatroid: that is what i assumed 2010-11-30T00:43:44 antimatroid: actually, since you were 46 ships ahead you could theoretically have scooted all your spare ships up and taken the central planet 2010-11-30T00:43:44 *** Azrathud has quit IRC (Quit: Leaving) 2010-11-30T00:44:05 janzert: Anyone planning for next year, yet? 2010-11-30T00:44:24 janzert: I wanted to add myself to a list of potential volunteers for prep. next year. 2010-11-30T00:44:26 smloh1: not without weakening your own position somewhere I guess 2010-11-30T00:44:31 I know Jeff wants to have something for next term 2010-11-30T00:44:41 bss03: talk to amstan next time you see him 2010-11-30T00:44:53 @seen amstan 2010-11-30T00:44:53 bss03: amstan was last seen in #aichallenge 3 hours, 40 minutes, and 33 seconds ago: but there's a high chance a will beat b 2010-11-30T00:44:57 bss03: yo 2010-11-30T00:45:00 he can point you in the right direction 2010-11-30T00:45:13 bss03: umm... just watch the suggestions/behind the scenes forums 2010-11-30T00:45:22 bss03: and hang out more in here 2010-11-30T00:46:02 smloh1: When I think of attacking a planet with all my force and the enemy defends with all his force, then there isn't generally only one point where I win. The planet may switch sides a few times until the fleets are all worn down. 2010-11-30T00:46:08 amstan: I'm really not much of an IRC'er most of the time. But, I'll use RSS to watch the behind the scenes forums. Thanks. 2010-11-30T00:46:27 bss03: we'll have planning in there too 2010-11-30T00:47:38 Frontier: but you can tell by calculating whether all your forces would be enough to take on all the forces the enemy will have there by the time your fleet arrives 2010-11-30T00:47:39 Hey amstan and janzert, I put up some ideas based on the old RoboCom game in the ideas forum. 2010-11-30T00:48:12 Great 2010-11-30T00:48:26 Top 10 players: bocsimacko(3925), Slin-.-(3733), _iouri_(3732), _Astek_(3727), GreenTea(3726), jimrogerz(3724), Accoun(3703), protocolocon(3697), george(3695), asavis(3656) 2010-11-30T00:48:45 it terible 2010-11-30T00:48:51 dable games 2010-11-30T00:48:55 double games 2010-11-30T00:49:04 wrind winer detection 2010-11-30T00:49:11 wrond winer detection 2010-11-30T00:49:20 all final games mast be remekred! 2010-11-30T00:49:29 all final games mast be remaked! 2010-11-30T00:49:40 i thought janzert fixed the double games and explained the apprent wrong winner? 2010-11-30T00:49:55 all final games mast be remaked!!! 2010-11-30T00:50:05 smloh1: I wouldn't have 'one fleet' either. I can send few ships from a close planet or many ships if I wait some turns and accumulate ships from farther away. Both strategies work differentlywork 2010-11-30T00:50:38 it's not possible to re run the contest 2010-11-30T00:50:43 that would cause the biggest shit storm 2010-11-30T00:50:59 Frontier: yes, i'm not sure we don't actually agree. that's what i do too. 2010-11-30T00:51:07 it would have to severely go wrong for that to be a feasible option 2010-11-30T00:51:17 bigger than yours Accoun, unless you repeat your proposal 100 times more ;) 2010-11-30T00:51:18 there was no wrong winner and a total of 5 games that were played twice (out of over 400k) 2010-11-30T00:51:32 all final games mast be remaked!!! 2010-11-30T00:51:40 ok, 99 to go 2010-11-30T00:51:50 there is something strange about position #11... it might be cursed! 2010-11-30T00:52:02 and yes the doubled games were removed from the rankings 2010-11-30T00:52:09 medrimonia and i both had the largest losing streaks as we got to around #11 2010-11-30T00:52:29 all final games mast be remaked!!! 2010-11-30T00:52:35 that's enough 2010-11-30T00:53:38 all final games mast be remaked!!! 2010-11-30T00:53:51 woohoo 2010-11-30T00:53:55 int he top 50 finally 2010-11-30T00:53:57 #49 2010-11-30T00:54:07 This case reveals the instability of the servers 2010-11-30T00:54:07 shows that McLeopold breaks on different moves 2010-11-30T00:54:18 My robot has now unexplainable loss 2010-11-30T00:54:19 it can only be attributed to a strong decrease in server performance 2010-11-30T00:54:20 grats, bhasker 2010-11-30T00:54:41 покой! 2010-11-30T00:54:53 thanks 2010-11-30T00:54:54 unexplained loss, that's a new one 2010-11-30T00:57:03 goodnight everyone, see you tomorrow 2010-11-30T00:57:07 night 2010-11-30T00:58:50 * burny blames her losses on cpu up-time 2010-11-30T00:59:19 Accoun: which game is this? 2010-11-30T01:00:35 *** jonoff has joined #aichallenge 2010-11-30T01:02:33 how many games he plan to make? 2010-11-30T01:03:06 now _iouri_ vs bocsimacko 2-2 2010-11-30T01:04:18 Accoun vs bocsimacko 0-3 2010-11-30T01:04:34 what is accoun complaining about? 2010-11-30T01:04:40 it luck 2010-11-30T01:04:46 need more games 2010-11-30T01:04:55 you will get more games once the cutoff changes? 2010-11-30T01:05:07 it might not just be luck 2010-11-30T01:05:21 smloh1: rubbish, accoun has the best bot 2010-11-30T01:05:23 anywey need more games 2010-11-30T01:05:23 *** mega1 has joined #aichallenge 2010-11-30T01:05:32 :P 2010-11-30T01:05:48 4 and 3 games not enougth 2010-11-30T01:06:21 antimatroid: lol 2010-11-30T01:07:40 reminds me of fallout 2 and torr 2010-11-30T01:10:01 awesome memory :) 2010-11-30T01:12:02 *** Frontier has quit IRC (Ping timeout: 250 seconds) 2010-11-30T01:17:34 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-11-30T01:17:36 *** bhasker has quit IRC (Quit: bhasker) 2010-11-30T01:22:03 *** Frontier has joined #aichallenge 2010-11-30T01:22:54 *** Utkarsh has joined #aichallenge 2010-11-30T01:27:21 *** superflit_ has joined #aichallenge 2010-11-30T01:28:30 *** superflit has quit IRC (Ping timeout: 250 seconds) 2010-11-30T01:28:30 *** superflit_ is now known as superflit 2010-11-30T01:30:40 *** nullkuhl has quit IRC (Remote host closed the connection) 2010-11-30T01:31:50 holy shit, julian assange is only 39 2010-11-30T01:31:53 he looks about 60 2010-11-30T01:33:10 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-30T01:33:57 dealing with leaks must be a stressful lifestyle 2010-11-30T01:34:02 *** sigh has quit IRC (Ping timeout: 240 seconds) 2010-11-30T01:34:51 *** nullkuhl has joined #aichallenge 2010-11-30T01:40:40 take that rsergio, shangas ad oldman :) 2010-11-30T01:40:43 almost back to top 30 2010-11-30T01:41:57 i only beat oldman cause he crashed 2010-11-30T01:42:01 hardly satisfying 2010-11-30T01:42:56 *** bhasker has joined #aichallenge 2010-11-30T01:46:21 bhasker: this isn't a normal time for you to enter :P 2010-11-30T01:46:27 isn't it like the middle of the night there 2010-11-30T01:46:37 10:45 2010-11-30T01:46:42 ah, i guess not 2010-11-30T01:46:42 got called from work:-\ 2010-11-30T01:46:52 i'm used to you coming on at like 1am my time 2010-11-30T01:46:57 and saying morning 2010-11-30T01:46:57 so working on production time:( 2010-11-30T01:47:01 oops production issues 2010-11-30T01:48:27 Top 10 players: bocsimacko(3930), _iouri_(3735), _Astek_(3732), jimrogerz(3728), Slin-.-(3724), GreenTea(3724), protocolocon(3705), Accoun(3704), george(3703), asavis(3662) 2010-11-30T01:49:11 *** Sylph2 has joined #aichallenge 2010-11-30T01:50:14 *** Sylph has quit IRC (Ping timeout: 265 seconds) 2010-11-30T01:54:16 holding #50 now antimatroid 2010-11-30T01:55:31 *** smloh1 has quit IRC (Ping timeout: 265 seconds) 2010-11-30T01:55:55 *** sanddbox has quit IRC () 2010-11-30T02:01:41 bhasker: nice, i just climbed back up to 30 2010-11-30T02:10:00 just beat shangas and sin_sun^ 2010-11-30T02:10:18 nice 2010-11-30T02:10:49 grats on the 30 2010-11-30T02:14:57 Frontier: i just read your suggestion, hasn't htat already been done? 2010-11-30T02:18:06 a lot of people (including myself) would have been disqualified by now if the rules weren't altered 2010-11-30T02:20:06 *** smloh1 has joined #aichallenge 2010-11-30T02:20:36 ouch, good job bhasker, i got no excuse for that loss. 25 turns! 2010-11-30T02:21:20 oh didnt see that :D 2010-11-30T02:21:30 ? 2010-11-30T02:21:34 oh rage bot defense fail 2010-11-30T02:21:56 i have the same bug in my bot 2010-11-30T02:22:09 though due to a different reason 2010-11-30T02:22:20 *** virdo has quit IRC (Ping timeout: 265 seconds) 2010-11-30T02:22:34 yeah it's a tough one to fix and still be aggressive enough on expansion 2010-11-30T02:22:45 you sent 36 ships out on turn 1 2010-11-30T02:22:51 when you were only 6 turns away 2010-11-30T02:22:58 turn 1 is easy 2010-11-30T02:23:13 you should never send more than distance between starting planet * 5 2010-11-30T02:23:17 at turn 1 2010-11-30T02:23:25 after that its tricky 2010-11-30T02:23:43 since you have to factor in growth from neutrals you will acquire/enemy will acquire etc 2010-11-30T02:24:01 i am sure you lose on rage bot on this map 2010-11-30T02:24:03 i don't have turn 1 specific code, so it's all (supposedly) taken care of by my eval function 2010-11-30T02:24:05 lose to* 2010-11-30T02:24:25 you should test against ragebot, if you lose on this map 2010-11-30T02:25:00 yeah, i forgot about testing against ragebot, that could have been a good basic test to run any version 2010-11-30T02:26:13 bhasker: it's easy to not send that, but it's hard to do it and plan over the next few moves 2010-11-30T02:26:21 cause you are limiting the number of ships you expect to have 2010-11-30T02:26:45 that's the soul reason i ended up not using such a hack 2010-11-30T02:26:55 it completely messed with what my valid options might be 2010-11-30T02:28:14 *** sigh has joined #aichallenge 2010-11-30T02:28:27 i beat accoun :) 2010-11-30T02:28:30 he's going to hate that 2010-11-30T02:28:38 accoun has the same bug btw 2010-11-30T02:28:47 http://ai-contest.com/visualizer.php?game_id=9293578 2010-11-30T02:29:20 nice job 2010-11-30T02:29:45 i've had wins against 2, 4, 5, 7 and 8 2010-11-30T02:30:04 antimatroid on turn 1 there is no reason to send out more than the distance * 5 unless there are planets closer by that you can capture for sure 2010-11-30T02:30:14 bhasker, if distance to enemy is 15, and there's a 5grow that costs 99 2steps away, it's worth taking in turn1 2010-11-30T02:30:29 if enemy is 15 turns away 2010-11-30T02:30:35 bhasker: that is true, but doing so required limiting the ships i would expect to send in a few turns etc. 2010-11-30T02:30:39 which ruins a lot of planning 2010-11-30T02:30:41 true 2010-11-30T02:30:45 or even if enemy is 12units away 2010-11-30T02:30:47 but yea without such planning 2010-11-30T02:30:53 you are susceptible to rage bot attacks 2010-11-30T02:30:59 bhasker: my bot relies on a game tree :P 2010-11-30T02:31:03 it needs correct predictions 2010-11-30T02:31:17 in which case you won't be susceptible to a rage bot attack 2010-11-30T02:31:24 because it would have factored the future growth in 2010-11-30T02:31:25 my bot was always too conservative when i tried to add in rage bot defense 2010-11-30T02:31:31 my bot relies on re-enforcng the front lines.. but it doesn't re-enforce the front lines :( 2010-11-30T02:31:42 *** Appleman1234 has quit IRC (Remote host closed the connection) 2010-11-30T02:31:49 the problem burny is on some maps 2010-11-30T02:31:50 i wont add an expansionary move if it's snipable, but then if i add a second, the first might become snipable 2010-11-30T02:31:54 the front line is not always a nice line 2010-11-30T02:31:59 and bots that rely on that 2010-11-30T02:32:00 flunk 2010-11-30T02:32:04 bhask: I don't reneforce ANYthing :P 2010-11-30T02:32:11 i rely on the front line :P 2010-11-30T02:32:24 although i leave ships planned for future moves 2010-11-30T02:32:30 but don't store moves across turns 2010-11-30T02:32:31 watch nearly any of the games I've lost.. 2010-11-30T02:32:35 I have a lead.. 2010-11-30T02:32:42 then take a planet I can barely secure.. 2010-11-30T02:32:46 and then don't secure it 2010-11-30T02:32:47 dmj beat me :( 2010-11-30T02:32:51 peky american 2010-11-30T02:32:53 pesky* 2010-11-30T02:33:11 burny its a tricky problem to send correct number of ships 2010-11-30T02:33:13 to a neutral 2010-11-30T02:33:16 including a few games from some of the top10 2010-11-30T02:33:24 i make the same mistakes at times 2010-11-30T02:33:33 and so do some of the top bots 2010-11-30T02:33:34 would have had some easy top10 wins, if I had enoguh time to add re-enforcement code 2010-11-30T02:33:43 what do you mean by reinforcement? 2010-11-30T02:33:48 like.. 2010-11-30T02:34:08 your home has 0 ships, so I send 80ships to attack it.. 2010-11-30T02:34:18 then 3 turns later, you can re-capture it unless I send more.. 2010-11-30T02:34:20 and I don't send more 2010-11-30T02:34:38 k 2010-11-30T02:35:15 and during that time, you're expanding, and I'm doing nothing(because I lost a lot of expansion time, by playing a move that should have been 'a win' 2010-11-30T02:35:20 if i send to a neutral, i plan to send more than they possibly could 2010-11-30T02:35:29 not sure i understand 2010-11-30T02:35:29 same here nati 2010-11-30T02:35:33 and then.. I don't :P 2010-11-30T02:35:35 you sent 80 ships to capture it 2010-11-30T02:35:45 and you didnt send more because you saw you will lose it 2010-11-30T02:35:57 in that case you underestimated the # of ships 2010-11-30T02:35:59 ya, cause it can regrow 40 by the time I get there, and it's other planets can grow 23.. 2010-11-30T02:36:11 ah so the way i fix it burny 2010-11-30T02:36:15 and then 3turns later, they grow another 25, and re-capture it 2010-11-30T02:36:16 is by overestimating 2010-11-30T02:36:29 overestimating isn't a 'fix'.. 2010-11-30T02:36:41 well there are two ways you can look at it 2010-11-30T02:36:48 don't launch at enemy during expansion phase 2010-11-30T02:36:54 just expand rapidly 2010-11-30T02:37:03 and then you can do an exact snipe of each planet 2010-11-30T02:37:10 and eventually the growth loss will cost the enemy 2010-11-30T02:37:11 mmm i love milo 2010-11-30T02:37:11 mathematically, and realistically, it is a highly viable move 2010-11-30T02:37:16 I do take the planet. 2010-11-30T02:37:19 and can secure it 2010-11-30T02:37:28 and end up with more growth and population than them 2010-11-30T02:37:29 etc. 2010-11-30T02:37:30 hmm you can't secure it if you lost it after 3 turns? 2010-11-30T02:37:37 I CAN secure it.. 2010-11-30T02:37:42 but I don't 2010-11-30T02:37:56 hmm how can you secure it by continuing to send your ships? 2010-11-30T02:37:59 from your home planet 2010-11-30T02:38:04 5 at a time? 2010-11-30T02:38:17 I had more then 80.. 2010-11-30T02:38:25 maybe 76 was needed.. 2010-11-30T02:38:29 but you sent 80 at my home planet? 2010-11-30T02:38:31 plus an extra 3 per turn.. 2010-11-30T02:38:36 and I send 80, plus 0 per turn 2010-11-30T02:38:52 but that assumes he has at least 8 growth nearby? 2010-11-30T02:38:57 to your 5? 2010-11-30T02:39:35 leme find an example game 2010-11-30T02:39:44 so burny the other way to solve this is to do funnelling/staging 2010-11-30T02:40:00 basically my ships flow to planets nearest to enemy 2010-11-30T02:40:06 http://ai-contest.com/visualizer.php?game_id=9264378 2010-11-30T02:40:06 so if i see i am going to capture an enemy planet 2010-11-30T02:40:10 then my ships flow to it 2010-11-30T02:40:14 before it turns to me 2010-11-30T02:40:15 about 20% in, I attack at the top 2010-11-30T02:40:15 and not after 2010-11-30T02:40:48 bhasker: are you at the same workplace as a1k0n? or just both at yahoo 2010-11-30T02:40:57 same place same team practically 2010-11-30T02:40:57 *** Rubicon-|-Cross has joined #aichallenge 2010-11-30T02:41:01 we sit 2 cubes away 2010-11-30T02:41:10 ahh, fair enough :) 2010-11-30T02:41:45 20% in can you give ma specific turn number 2010-11-30T02:41:46 if the top was re-enforced: that would have been a win 2010-11-30T02:42:00 turn 34? 2010-11-30T02:42:08 about that, ya 2010-11-30T02:42:22 but you sent like 30 ships at a planet with 90? 2010-11-30T02:42:40 and again much later at the bottom 2010-11-30T02:42:41 in fact 2010-11-30T02:42:48 at turn 33 2010-11-30T02:42:58 you send 28 ships 2010-11-30T02:43:06 when the planet has 39 2010-11-30T02:43:07 on them 2010-11-30T02:43:10 that just seems wrong 2010-11-30T02:43:18 I sent 90? 2010-11-30T02:43:21 smiley is slowly rising 2010-11-30T02:43:21 oooh wait 2010-11-30T02:43:22 doh 2010-11-30T02:43:24 he better not catch me 2010-11-30T02:43:26 you are the blue 2010-11-30T02:43:28 ./bonk 2010-11-30T02:43:30 i don't think he will though 2010-11-30T02:44:03 k your problem is you are waiting till the planet converts before staging to the planet 2010-11-30T02:44:08 which is not required 2010-11-30T02:44:11 no.. 2010-11-30T02:44:13 I don't EVER stage 2010-11-30T02:44:26 did anyone else redirect ships? 2010-11-30T02:44:40 hmm 2010-11-30T02:44:50 so what is causing the routing 2010-11-30T02:45:01 if i planed to send ships addded at time zero which could be redirected closer to the enemy and finish to their destination in time, i redirected them 2010-11-30T02:45:06 something is forwarding the ships right? 2010-11-30T02:45:12 it wants to attack the enemy, but doesn't have enough strength, 2010-11-30T02:45:24 *** bss03 is now known as bss03-AFK 2010-11-30T02:45:29 and so it check's for available planets on the route to it's desired target 2010-11-30T02:45:42 antimatroid: i tried doing that, but it didn't necessarily perform better 2010-11-30T02:45:47 ah you are routing always through your planets? 2010-11-30T02:45:56 smloh1: it seemed to work better than not for me 2010-11-30T02:45:56 where the fail comes in, is when it's next target, is not the enemy 2010-11-30T02:46:05 so it's holding the ships there, waiting to attack a neutral 2010-11-30T02:46:18 why not treat enemies/neutrals the same 2010-11-30T02:46:24 it does 2010-11-30T02:46:28 bhasker: cause they are not the same 2010-11-30T02:46:29 it attacked the enemy.. 2010-11-30T02:46:37 for staging they are 2010-11-30T02:46:39 and knows it's securable, and it's going to capture, etc. etc. 2010-11-30T02:46:43 if you take an enemy for a turn, that's good, if you take a neutral for a turn, that's bad 2010-11-30T02:46:45 antimatroid: i think it kept more ships on the move and not available for me 2010-11-30T02:46:52 but I didn't have enough time to add code to properly re-enforce stuff 2010-11-30T02:46:54 *** tapwater has quit IRC (Quit: tapwater) 2010-11-30T02:47:16 if you take anything for a turn, expecting to take it forever, that's bad 2010-11-30T02:47:18 jimrogerz up to third 2010-11-30T02:47:42 burny: i made moves with the expectation that they would send every ship they had at the planet 2010-11-30T02:47:50 ok, even stronger exmaple.. http://ai-contest.com/visualizer.php?game_id=9293862 2010-11-30T02:48:01 anti, so do I 2010-11-30T02:48:10 but the next turn, when it's figuring out what to do next.. 2010-11-30T02:48:15 it doesn't try to re-enforce 2010-11-30T02:48:28 burny your problem is simple 2010-11-30T02:48:28 Top 10 players: bocsimacko(3927), _iouri_(3727), jimrogerz(3721), Slin-.-(3721), _Astek_(3714), GreenTea(3709), protocolocon(3695), Accoun(3693), george(3693), asavis(3654) 2010-11-30T02:48:31 you are not moving ships to the front 2010-11-30T02:48:34 fast enough 2010-11-30T02:48:39 you have ships sitting around at the back 2010-11-30T02:48:42 doing nothing 2010-11-30T02:48:44 isn't that what he said at the start? :P 2010-11-30T02:48:44 I don't move ships to teh front ATALL :P 2010-11-30T02:48:55 well you have to:-\ 2010-11-30T02:49:01 obviously :P 2010-11-30T02:49:09 bhasker: he wasn't programming for very long 2010-11-30T02:49:20 ah 2010-11-30T02:49:30 i wonder how many female competitors there were in the end 2010-11-30T02:49:36 i am yet to come across one 2010-11-30T02:49:53 you probably have, and just don't realize? 2010-11-30T02:50:01 well, :P 2010-11-30T02:50:05 you know what i mean 2010-11-30T02:52:16 seeing my bot slowly drop from the top 30ish, makes me cry about not having just another couple days :( 2010-11-30T02:52:44 its okay burny there is always next year 2010-11-30T02:52:56 see my bot fail here 2010-11-30T02:52:57 http://ai-contest.com/visualizer.php?game_id=9292955 2010-11-30T02:53:03 stupid shit tried to take the middle planet 2010-11-30T02:53:08 and killed itself 2010-11-30T02:53:22 with how many games I lost, because of the same 2 missing features.. 2010-11-30T02:53:36 I think I would have been top10 2010-11-30T02:53:49 burny don't be so sure lots of bots have those two features 2010-11-30T02:53:53 and don't make it to top 10 2010-11-30T02:53:54 even teh game I had with bosci, I would have won if not for that 2010-11-30T02:54:22 what are the 2 features burny? 2010-11-30T02:54:54 damn my first turn is horrible 2010-11-30T02:54:57 http://ai-contest.com/visualizer.php?game_id=9298325 2010-11-30T02:54:58 re-enforcing the front lines, and making an attack that leaves the source planet vulnerable to counter attacks(basically the same thing) 2010-11-30T02:55:27 yes those are important 2010-11-30T02:55:35 the other thing I would have liked to do, was have counter-attack logic.. so if I'm losing a planet, to attack back, instead of just taking it in the face 2010-11-30T02:55:39 but they're not trivial to implement 2010-11-30T02:55:39 i am still debating about #2 2010-11-30T02:55:49 my current bot has no such consideration 2010-11-30T02:56:06 smloh: trivial no.. but I could have for sure done it in a day 2010-11-30T02:56:07 it considers anything over minimum required to handle attacks alread in progress to be available 2010-11-30T02:56:22 and it actually does reasonably well 2010-11-30T02:56:31 despite losing planets after launching attacks 2010-11-30T02:56:32 ya, same here 2010-11-30T02:56:38 my bot is completely broken tree wise, it's amazing it's doing well at all 2010-11-30T02:56:58 i spent a lot of time trying to find an optimal reinforcement algo which would adequately protect/supply multiple frontlines and accumulate ships for potential attacks on enemy/neutrals 2010-11-30T02:57:05 if I could add JUST re-enforcing the front lines.. I'm sure I'd be nearly top10 2010-11-30T02:57:20 sometimes neutrals are behind the front line and leaving enough ships at the back to capture one tilts the balance 2010-11-30T02:57:33 burny: just storm everyone next comp :) 2010-11-30T02:57:49 there's going to be even more slaves from the start next time 2010-11-30T02:58:13 smloh1: there are too many RPS models 2010-11-30T02:58:29 if you have a strong central and you don't reinforce to it and enemy reinforces to his central you can lose it 2010-11-30T02:58:42 on some maps holding back can win the map 2010-11-30T02:58:48 when enemy funnels to center 2010-11-30T02:58:50 on others it can lose it 2010-11-30T02:58:51 bhasker: true, it's tough 2010-11-30T02:58:57 i like my hack 2010-11-30T02:59:04 just send up to 4* planets growth 2010-11-30T02:59:10 and only redistribute after turn 4 2010-11-30T02:59:18 i tried to fix this with doing like planet density counts 2010-11-30T02:59:28 and deciding to hold back if there are lot of good neutrals nearby 2010-11-30T02:59:37 but it never worked very well:-\ 2010-11-30T02:59:43 here's another.. 2010-11-30T02:59:44 http://ai-contest.com/visualizer.php?game_id=9284924 2010-11-30T02:59:59 about 20% into it again 2010-11-30T03:00:45 as much as taking the 36neutral wasn't very clever.. if the home base was re-enforcing: it would have been fine 2010-11-30T03:01:11 but instead, home sends to teh top, preparing to attack the opponents top planet next 2010-11-30T03:01:38 and then after that.. just lets it's ships sit there, roting, waiting to die 2010-11-30T03:02:24 so burny another thing i do 2010-11-30T03:02:46 is i funnel to the planet at (avgdistfrommy+avgdistfromenemy)/2 2010-11-30T03:02:54 the one closest to that 2010-11-30T03:02:59 so its usually a central planet 2010-11-30T03:03:18 and well i include neutrals under acquisition into the my/enemy planet list 2010-11-30T03:03:28 so that i take into account where enemy will be in the future 2010-11-30T03:03:31 when funnelling 2010-11-30T03:03:36 so that i don't end up funnelling to the top 2010-11-30T03:03:41 when enemy is capturing stuff at the bottom 2010-11-30T03:03:45 ya, I do the future stuff too 2010-11-30T03:04:01 just missing, re-enforcing front lines 2010-11-30T03:04:11 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-11-30T03:04:42 of all the losses I watched of mine: they're all either, attacking leaving the source undefended, or failing to re-enforce 2010-11-30T03:04:47 yup that is critical since holding ships back once neutrals are taken is pointless 2010-11-30T03:04:57 you can always protect your back line from your front line 2010-11-30T03:05:18 they have a shorter distance to travel than the enemy at any time 2010-11-30T03:05:31 uh huh 2010-11-30T03:05:33 like this http://ai-contest.com/visualizer.php?game_id=9300115 2010-11-30T03:05:43 my central planet can protect any of my planets 2010-11-30T03:05:45 from attacks 2010-11-30T03:05:52 what's your analysis of this one? ;) http://ai-contest.com/visualizer.php?game_id=9150531 2010-11-30T03:06:15 mine? 2010-11-30T03:06:28 no, burny 2010-11-30T03:07:47 also another thing, there is very little advantage in launching attacks from back planets 2010-11-30T03:07:56 so you might as well send everything to the front 2010-11-30T03:08:31 it should have re-enforced the top planet 2010-11-30T03:09:03 *** Utkarsh has joined #aichallenge 2010-11-30T03:09:06 i think it was more the decision to attack the lower planet, which left insufficient ships to reinforce/defend the top planet 2010-11-30T03:09:55 * antimatroid has moved onto thinking about arimaa :) 2010-11-30T03:10:50 sending 71 was ok.. I had enough to defend against the 36 +5 per turn 2010-11-30T03:10:51 so when is the arimaa competition 2010-11-30T03:10:58 but I didn't send them there 2010-11-30T03:11:01 I let them rot 2010-11-30T03:11:28 Pairing cutoff number changed? 2010-11-30T03:11:35 it lower 2010-11-30T03:11:42 but yes: it did not take into account how much needed to be kept behind for defense 2010-11-30T03:12:30 no i think it became undefendable once the top planet flipped then it became a defense against 37 +8 per turn 2010-11-30T03:12:43 it wouldn't have flipped 2010-11-30T03:13:08 no, i mean my top planet 2010-11-30T03:13:45 I had 9on teh top we each had 3grow 2010-11-30T03:13:57 *** jaspervdj has joined #aichallenge 2010-11-30T03:14:29 i think my bot has reached its potential 2010-11-30T03:14:37 i think its going to hover around 45-555 2010-11-30T03:14:38 55* 2010-11-30T03:14:44 for the rest of the competition 2010-11-30T03:15:12 bhasker: no, be optimistic! i need you to climb up so the loss against you doesn't count as much against my elo! :P 2010-11-30T03:15:20 lol 2010-11-30T03:15:42 looking at the bots above me 2010-11-30T03:15:49 i know i can't beat most of the consistently on tcp server 2010-11-30T03:16:01 dabino i am probably 50-50 with 2010-11-30T03:16:08 leopold stomps my bot 2010-11-30T03:16:19 oldman is like 35-65 2010-11-30T03:16:34 yes, in fact i'm surprised how far down many familiar names from before are 2010-11-30T03:17:07 so nope don't have hopes of climbing much further 2010-11-30T03:17:12 unless my bot becomes sentient 2010-11-30T03:17:16 and starts thinking for itself 2010-11-30T03:17:40 when on wednesday does it end? 2010-11-30T03:17:53 no idea 2010-11-30T03:18:10 well it's at least another 24 full hours of processing, a lot can happen 2010-11-30T03:18:22 so, good luck! 2010-11-30T03:18:37 hehe thanks lets see 2010-11-30T03:18:47 me rebel have very similar bots 2010-11-30T03:18:51 and our elo's reflect that 2010-11-30T03:18:57 we are like 2-3 ranks from each other 2010-11-30T03:19:04 1 rank now 2010-11-30T03:19:10 3481 vs 3477 2010-11-30T03:19:12 are they similar in logic? 2010-11-30T03:19:37 i think so , rebel ran some tests of my bot vs his 2010-11-30T03:19:44 on the full 150 map set i think we are almost even 2010-11-30T03:19:56 on the first 75 his bot dominates 2010-11-30T03:20:00 on the second set mine dominates 2010-11-30T03:20:32 ah 2010-11-30T03:20:47 i wish there were some way to remove map selection bias 2010-11-30T03:20:50 but it's impossible 2010-11-30T03:21:01 *** Accoun has quit IRC () 2010-11-30T03:21:55 anyway time to hit the bed 2010-11-30T03:21:57 l8r folks 2010-11-30T03:22:14 hopefully in the morning my bot will surprise me by being in the top 20 2010-11-30T03:22:17 sweet dreams 2010-11-30T03:22:26 later man 2010-11-30T03:23:53 and oi.. another major "lets not bother defending it" round.. 2010-11-30T03:23:56 http://ai-contest.com/visualizer.php?game_id=9300881 2010-11-30T03:25:09 how does one handle maps like this 2010-11-30T03:25:09 http://www.ai-contest.com/visualizer.php?game_id=9303168 2010-11-30T03:25:14 was taking the 57 a mistake? 2010-11-30T03:25:30 and how does one preven such a mistake 2010-11-30T03:25:52 i don't handle that map well, but if my tree worked right it shoul 2010-11-30T03:26:28 my current bot sends 100 at one of the centre planets on the left aha 2010-11-30T03:26:52 heh i guess there is no easy way to solve such maps without a 1-2 step look ahead 2010-11-30T03:27:24 http://www.ai-contest.com/profile.php?user_id=7434 2010-11-30T03:27:30 aha, that bot crashes right as it loses 2010-11-30T03:27:37 always 2010-11-30T03:27:49 *** aerique has joined #aichallenge 2010-11-30T03:27:52 lucky they didn't get that the other way around 2010-11-30T03:28:01 probably has a divide by zero bug 2010-11-30T03:28:09 *** criis has joined #aichallenge 2010-11-30T03:28:15 yep 2010-11-30T03:28:15 probably divides by number of ships or planets some where 2010-11-30T03:28:19 or growth 2010-11-30T03:28:27 yeah, but only for them too 2010-11-30T03:28:35 *** Accoun has joined #aichallenge 2010-11-30T03:28:39 true 2010-11-30T03:28:42 *** criis has left #aichallenge 2010-11-30T03:28:48 why do you refer it as "them" 2010-11-30T03:29:02 that is actually one person's name :) 2010-11-30T03:29:14 dhaval shreyas 2010-11-30T03:29:25 *** yasith has quit IRC (Ping timeout: 264 seconds) 2010-11-30T03:29:40 the s at the end throwing you off? 2010-11-30T03:30:09 what? 2010-11-30T03:30:14 the oppoent is them? 2010-11-30T03:30:20 there is me and them :P 2010-11-30T03:31:03 Accoun vs GreenTea 0-4 , cant belive that 2010-11-30T03:31:53 accoun bhasker vs top 30 bots 0-30 makes no sense 2010-11-30T03:31:54 go figure 2010-11-30T03:32:23 how do you find those stats? 2010-11-30T03:32:31 i am sure if janzert hadn't picked the maps i would have been #1 2010-11-30T03:32:40 lol :P 2010-11-30T03:32:45 i am sure he hand picked it to make me lose 2010-11-30T03:32:57 manyale caclulete 2010-11-30T03:33:04 manyale calculete 2010-11-30T03:33:06 wth does that mean? 2010-11-30T03:33:14 anyone? 2010-11-30T03:33:17 pc load letter? 2010-11-30T03:34:26 *** bhasker has quit IRC (Quit: bhasker) 2010-11-30T03:34:40 somethin wrong with this map set 2010-11-30T03:34:54 you say that about every map set 2010-11-30T03:35:02 *** mega1 has quit IRC (Remote host closed the connection) 2010-11-30T03:35:22 *** mega1 has joined #aichallenge 2010-11-30T03:36:47 mega1: 15 win streak :) 2010-11-30T03:37:46 cool, now I want to pay back iouri 2010-11-30T03:38:55 i beat him twice :P 2010-11-30T03:39:07 i've not once come close to beating you 2010-11-30T03:41:36 just wait for the right map 2010-11-30T03:42:11 *** yasith has joined #aichallenge 2010-11-30T03:46:13 *** phreeza has quit IRC (Ping timeout: 264 seconds) 2010-11-30T03:48:28 Top 10 players: bocsimacko(3930), _iouri_(3730), Slin-.-(3716), _Astek_(3711), GreenTea(3710), jimrogerz(3707), george(3693), protocolocon(3685), Accoun(3684), asavis(3652) 2010-11-30T03:54:25 i'm not sure, my bot is pretty hopeless for this 2010-11-30T03:54:30 in TCP Accoun all time be near _iouri_ 2010-11-30T03:54:38 move selection had me beat 2010-11-30T03:54:56 well, while staying defended 2010-11-30T03:56:10 map set random 2010-11-30T03:56:27 why not 999 maps? 2010-11-30T03:56:52 499 maps on TCP give more right result 2010-11-30T03:57:21 maps set random/preselect 2010-11-30T03:57:59 no any reson to this 2010-11-30T04:00:04 Pairing cutoff number lower and lower 2010-11-30T04:00:22 it mean ELO score? 2010-11-30T04:00:52 *** smloh1 has quit IRC (Quit: Page closed) 2010-11-30T04:01:11 not, strange number 2010-11-30T04:05:08 199 games per bot 2010-11-30T04:08:51 *** narnach has joined #aichallenge 2010-11-30T04:08:51 *** narnach is now known as narnach_mb 2010-11-30T04:10:06 2 games won of four against iouri^^ 2010-11-30T04:10:52 medrimonia: me too 2010-11-30T04:10:58 ^^ 2010-11-30T04:10:58 medrimonia: i think me too :) 2010-11-30T04:11:23 and 1 out of 3 against slin 2010-11-30T04:11:26 nope, i'm 2/3 :) 2010-11-30T04:11:39 2/3 against Accoun 2010-11-30T04:11:46 while locally I won the great majority of games against iouri 2010-11-30T04:11:46 i'm 1/1/1 against accoun 2010-11-30T04:12:07 1/3 against astek 2010-11-30T04:12:18 1/4 against against slin 2010-11-30T04:12:29 i haven't beaten slin 2010-11-30T04:12:53 1/3 against astek 2010-11-30T04:13:38 *** narnach_mb has quit IRC (Client Quit) 2010-11-30T04:13:44 3 wons on four games against jimrogerz^^ 2010-11-30T04:14:04 make that 1/4 against astek 2010-11-30T04:14:20 i'm 1/5 against jim 2010-11-30T04:14:33 i won against astek from a very bold first move 2010-11-30T04:14:43 http://ai-contest.com/visualizer.php?game_id=9194659 2010-11-30T04:14:55 2 games won on four games against GreenTea 2010-11-30T04:15:27 I think that if I wasn't loosing so much against the others, I could be in top ten 2010-11-30T04:15:28 i don't think i've beaten him either 2010-11-30T04:15:41 my bot should be lower than it is now i think 2010-11-30T04:15:48 the tree is oh so broken 2010-11-30T04:16:04 If they stop the pairing cutoff at 25, I think I'll gain some rankings 2010-11-30T04:16:10 arimaa actually looks quite fun to make a bot for 2010-11-30T04:16:23 i don't think pairing cutoff will even get to 100 2010-11-30T04:16:24 1000* 2010-11-30T04:17:01 Yes but I can only hope^^ 2010-11-30T04:17:22 i've beaten wagstaff up a few times 2010-11-30T04:17:28 i expected him to do better tbh 2010-11-30T04:19:01 my arimaa bot is going to have trees inside trees :) 2010-11-30T04:19:03 1 win 1 draw and 2 loss against Darkemi... that's making me loose a lot of rankings 2010-11-30T04:19:15 your bot gets a really long time to pick it's moves too 2010-11-30T04:19:27 darkemi has been an issue for me too :) 2010-11-30T04:19:32 *** narnach has joined #aichallenge 2010-11-30T04:19:32 *** narnach is now known as narnach_mb 2010-11-30T04:21:01 2 whole minutes in fact for each turn 2010-11-30T04:22:33 although i think the run the bot over each turn, which means i can't just keep a tree in memory and update it every move 2010-11-30T04:23:12 Hmmm you've the right to create file nop? 2010-11-30T04:23:28 i can probably do that, but it'd be faster if it was already in memory 2010-11-30T04:23:43 still faster than regenreating though 2010-11-30T04:23:47 ohhhh pairing cutoff at 3500 that's nice 2010-11-30T04:24:20 i'm going to try doing trees inside trees, and just see how it does :) 2010-11-30T04:25:12 bah, 0/3 against mcleopold, despite beating him more than half the time on these maps on his tcp server 2010-11-30T04:25:14 oh well 2010-11-30T04:25:35 i had it too, then gave it away 2010-11-30T04:27:45 are maps in c++ reasonably fast to use once set up? 2010-11-30T04:31:35 *** Default-08-15 has joined #aichallenge 2010-11-30T04:31:45 any admin here? 2010-11-30T04:34:04 doesn't look like it 2010-11-30T04:34:12 what's your question? we may be able to help 2010-11-30T04:35:32 I produces a gallic village ;-) 2010-11-30T04:35:37 produced 2010-11-30T04:35:46 http://ai-contest.com/visualizer.php?game_id=9309789 2010-11-30T04:38:16 without this bug i would have easily won the game ;-) *g* 2010-11-30T04:38:27 what's the bug? 2010-11-30T04:38:28 *** antimatroid has quit IRC (Read error: Connection reset by peer) 2010-11-30T04:38:51 *** antimatroid has joined #aichallenge 2010-11-30T04:40:07 the blue planets almost form a sideways infinity 2010-11-30T04:40:25 that's called an 8 2010-11-30T04:40:30 :P 2010-11-30T04:40:31 hmm, there can really be a growth rate of 0? 2010-11-30T04:40:35 Default-08-15: yes 2010-11-30T04:40:36 i know that, but it looks more like an infinity 2010-11-30T04:40:46 only for the centre planet 2010-11-30T04:40:55 oh, i didnt know that 2010-11-30T04:41:22 thanks 2010-11-30T04:41:55 sigh: looks like i've got australia in the bad :) 2010-11-30T04:42:03 yup :) 2010-11-30T04:42:05 bag* 2010-11-30T04:42:24 Default-08-15> it indian reservation to make prevent genocide 2010-11-30T04:42:34 Default-08-15> it indian reservation to prevent genocide 2010-11-30T04:42:43 Default-08-15> it indian reservation to avoid genocide 2010-11-30T04:42:48 Accoun: once is enough 2010-11-30T04:43:05 they were each different ;) 2010-11-30T04:43:25 and avoid was the best replacement for make prevent 2010-11-30T04:43:43 lol 2010-11-30T04:44:03 *** nullkuhl has left #aichallenge ("Konversation terminated!") 2010-11-30T04:44:10 that wasn't a joke -.- 2010-11-30T04:44:15 *** Default-08-15 has quit IRC (Quit: Page closed) 2010-11-30T04:44:17 why janzert add the new maps set after I won 2nd place? 2010-11-30T04:44:45 yes, a now hate all canadian 2010-11-30T04:45:07 because he doesn't like to see you on second place and changed the maps until you drop to a lower rank 2010-11-30T04:45:29 499 map set on TCP anywey give more correct result than current oficial 2010-11-30T04:45:38 *** Naktibalda has joined #aichallenge 2010-11-30T04:45:48 who decides what is more correct? 2010-11-30T04:45:59 Frontier> yes- it canadian chiting like lase year 2010-11-30T04:46:02 Frontier> yes- it canadian chiting like last year 2010-11-30T04:47:11 i now hate all canadian more and more 2010-11-30T04:47:33 on TCP im be 2nd 2010-11-30T04:48:13 it's funniest that janzert isn't even canadian 2010-11-30T04:48:28 Top 10 players: bocsimacko(3921), _iouri_(3715), Slin-.-(3705), jimrogerz(3702), GreenTea(3700), _Astek_(3698), george(3677), protocolocon(3676), Accoun(3674), asavis(3653) 2010-11-30T04:49:56 I have a funny idea for a contest. The whole game would run in a virtual machine and you can construct and extend factories on the planets. The ships, planets and factories all have their own execution thread. To capture another planet you would have to construct a ship with 'transport' room, move your planet's population into the ship and unload it at its destination. The more population you have on one planet the faster y 2010-11-30T04:49:56 extend your factory there. 2010-11-30T04:51:12 *** janzert has quit IRC (Ping timeout: 265 seconds) 2010-11-30T04:51:42 biggest planet wins :) 2010-11-30T04:51:46 Frontier: so growth is proportional to ships rather than a fixes constant? 2010-11-30T04:52:01 *fixed 2010-11-30T04:52:09 then there is no incentive to expand 2010-11-30T04:52:19 population (humans) and ships would be 2 ressources 2010-11-30T04:52:42 *** smloh1 has joined #aichallenge 2010-11-30T04:53:13 growth would produce more people faster, who can in turn build a factory faster 2010-11-30T04:54:12 what is the incentive to expand? 2010-11-30T04:54:17 aww man, sif c++ wont let me define things like int[2] loc 2010-11-30T04:54:31 then the planet transfers some virtual machine program into the factory, which in turn produces ships proportional to the factory size and gives the ships another program so they can freely decide where to go ^^ 2010-11-30T04:54:41 doh, medrimonia, you beat me 3 times out of 4 2010-11-30T04:54:52 sigh: You get more population and can grow faster 2010-11-30T04:55:09 antimatroid: int loc[2] 2010-11-30T04:55:28 antimatroid: lol, n00b ;) 2010-11-30T04:55:32 sigh: vector pieces[2]; is much less clear than say vector[2] pieces; 2010-11-30T04:55:34 Frontier: but doubling your resources on two planets is the same as doubling your resources on 2 2010-11-30T04:55:57 i know i can do it the other way, i'm one that likes my code to be readable :) 2010-11-30T04:56:02 well, at least for me 2010-11-30T04:56:10 then stop using c++? :P 2010-11-30T04:56:15 :P 2010-11-30T04:56:28 i installed and started playing around with haskell yesterday 2010-11-30T04:56:45 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-30T04:56:52 Frontier: sorry, doubling your resources on two planets is the same as doubling your resources on one 2010-11-30T04:57:03 i'd rather do an entirely new game 2010-11-30T04:57:05 how are you finsing it? 2010-11-30T04:57:10 and if i was to choose, i'd be doing 2 player mario kart 2010-11-30T04:57:10 *finding 2010-11-30T04:57:16 so we've heard :P 2010-11-30T04:57:19 it's alright 2010-11-30T04:57:48 are you making anything with it? 2010-11-30T04:58:01 not at the moment, making an arimaa bot now 2010-11-30T04:58:15 smloh1: I was sleeping, but I agree with you with the loosing streak at #11 2010-11-30T04:58:15 in haskell? 2010-11-30T04:58:24 antimatroid: There was an ai competition with the racing game TORCS 2010-11-30T04:58:26 nope, c++ 2010-11-30T04:58:34 ah, i see 2010-11-30T04:58:36 good luck :) 2010-11-30T04:58:49 smloh1: And yep, I'm quite good against the best bots, but I'm loosing too much against bots out of top 50 2010-11-30T04:59:42 aww man, the demolution derby ai comp sounds awesome 2010-11-30T04:59:52 medrimonia: well, come on back up and we can try assaulting #11 again 2010-11-30T04:59:55 go late bloomers! 2010-11-30T05:00:03 all bots on TCP order expect me 2010-11-30T05:00:17 all bots in TCP order expect me 2010-11-30T05:00:49 not everyone tested on tcp! 2010-11-30T05:01:23 6 loses in a row :\ 2010-11-30T05:02:04 sigh: I'm not sure I get your point. If I have the population growth of two planets I can build up my factories on both planets fast. So two planets always gives me twice the ships in the end 2010-11-30T05:02:48 Frontier: I guess I don't understand the growth function... can you give me an idea how the growth is calculated? 2010-11-30T05:02:54 are the elo for top 2-9 going down? they aren't looking as invulnerable as they did earlier today 2010-11-30T05:03:33 *** medrimonia1 has joined #aichallenge 2010-11-30T05:04:01 *** medrimonia has quit IRC (Quit: Leaving.) 2010-11-30T05:04:21 *** perror has joined #aichallenge 2010-11-30T05:04:28 *** medrimonia1 is now known as medrimonia 2010-11-30T05:04:34 sigh: I'm not so sure myself yet. But I guess if the planet has +5, that affects the population like it does the ships in PlanetWars. The population in turn would be related to the 'growth' of a factory on this planet. 2010-11-30T05:05:19 might be as a result of good bots with bad start moving back up the ranks, the top bots are facing tougher opponents in the "neighbourhood" and getting more losses 2010-11-30T05:05:21 So with 10 population the factory can produce like 1 ship more each turn, with 20 population it is 2 ships more each turn and so on. 2010-11-30T05:05:46 *** eashoka has joined #aichallenge 2010-11-30T05:06:27 ah i see... yeah, that might work 2010-11-30T05:06:43 bah, how is int[2] loc not better? 2010-11-30T05:06:57 sigh: Ships could have different sizes though. Maybe even carriers would be possible ^^. At least fighters and transport ships would be nice. The fighters would reduce the factory on the enemy planet and the transport would bring population to the planet eleminating the enemy population there. 2010-11-30T05:07:01 Yep smloh1, in top 10 they have all something like 20 points ELO less thans yesterday 2010-11-30T05:07:05 sometimes i really hate c++, but most of the time it's love 2010-11-30T05:07:10 antimatroid: that's just how c is 2010-11-30T05:07:17 loc[2] is of type int 2010-11-30T05:07:31 jsut like in int *loc, *loc is of of type int 2010-11-30T05:07:41 i prefer to think of an array of ints as a different type to a single int 2010-11-30T05:07:56 medrimonia: that is good news :) 2010-11-30T05:08:06 but int* a, b <- b is of type int, a is of type int* 2010-11-30T05:08:26 Frontier: there is no such type as int* 2010-11-30T05:08:40 *a is if type int <- this is the corrent way to think of it 2010-11-30T05:08:45 there is no pointer to an integer?? 2010-11-30T05:08:49 smloh1: Yes but I need a really a good winning streak to come back in top 15 already 2010-11-30T05:09:14 there aren't types for pointers 2010-11-30T05:09:29 medrimonia: i was lucky to get a 7-win streak right after my 8-loss streak so i didn't drop too drastically 2010-11-30T05:09:36 just pointers to different (all) types 2010-11-30T05:09:38 is that right? 2010-11-30T05:10:18 *** eashoka has quit IRC (Ping timeout: 265 seconds) 2010-11-30T05:10:44 smloh1: I'm winning something like 2 of three, but I've just lost against #57, that's pretty bad to be upping again 2010-11-30T05:11:10 kind of... what I'm getting at more is that the whole type declaration thing makes more sense if you think of *a as being of type int, rather than int* being it's own type 2010-11-30T05:11:28 medrimonia: yeah i got a bunch of losses against ~50's those are painful 2010-11-30T05:12:11 smloh1: I hope I'll get some wins against top 10 players, that would be really helpgul 2010-11-30T05:12:14 ignoring how the language actually is, isn't it better to think of them differently? 2010-11-30T05:12:15 helpful* 2010-11-30T05:12:15 it's the way that K&R says to think about it :) 2010-11-30T05:12:51 i think of pointers as variables for memory addresses 2010-11-30T05:13:07 that's what they are 2010-11-30T05:13:23 yes, but when it's taught it's like referencing blah blah blah 2010-11-30T05:13:26 medrimonia: yeah, i actually have a pretty good record against some of them which is probably helping to keep my ranking up 2010-11-30T05:13:35 just tell me it's a variable for memory locations godamnit! 2010-11-30T05:13:36 antimatroid: depends who taught you :P 2010-11-30T05:13:44 i went go to utas :P 2010-11-30T05:13:45 sigh: I could give the game a story line and several single player missions, where your task is to code a bot that beats the opponent bot I set up for that mission. 2010-11-30T05:13:53 went/go* 2010-11-30T05:14:15 The competition would just be the multiplayer part of it ^^ 2010-11-30T05:14:18 Frontier: that's what galcon is already, no? :P 2010-11-30T05:14:29 Frontier: why not something more different? :P 2010-11-30T05:14:49 sigh: Galcon is a fast interactive game, not a programming game 2010-11-30T05:15:10 what do you mean by "programming game"? 2010-11-30T05:15:12 antimatroid: Because I am a sigh-fi fan ^^ 2010-11-30T05:15:13 oh, you want programming missions? 2010-11-30T05:15:29 that sounds interesting, maybe you should start a project :) 2010-11-30T05:15:34 * antimatroid is a game theory fan, hence i'm happy with 2 player simultaneous games 2010-11-30T05:16:11 sigh: Exactly that is the idea. You write code for planets, factories and ships and hope your code completes the mission successfully 2010-11-30T05:16:49 i still don't get what the difference is with a programming game? 2010-11-30T05:17:10 antimatroid: a programming game is generally a game where you write code. 2010-11-30T05:17:21 wasn't this a game where we wrote code? 2010-11-30T05:17:38 Galcon is not PlanetWars 2010-11-30T05:18:04 sorry, when people say galcon in here i assume they're talking about the planet wars bot game :) 2010-11-30T05:18:30 i'm still lost though, what is your game replacing galcon for? 2010-11-30T05:18:38 weren't you talking about new games to use? 2010-11-30T05:18:47 *** Naktibalda has left #aichallenge 2010-11-30T05:19:42 I was thinking about an extended PlanetWars (with factories and different ships maybe) where each unit is independent and runs its own code out of a set you - the programmer - have predefined. 2010-11-30T05:20:29 i don't see how it's any different 2010-11-30T05:20:39 other than more processing for friend bots to share their information 2010-11-30T05:22:05 You say that as if it didn't make a difference :) . If we don't know anything about the enemy planets and ship movements that would also require sending scouts sometimes. 2010-11-30T05:22:11 Pairing cutoff is really going down now^^ 2010-11-30T05:22:30 what could be cool is thousands of planets and huge distances 2010-11-30T05:23:13 antimatroid: The virtual machine must handle all the unique ships and planets... remember ^^ 2010-11-30T05:23:53 *** Epona has joined #aichallenge 2010-11-30T05:23:58 *** Sylph2 is now known as Sylph 2010-11-30T05:24:51 I don't think thousands of planets would be interesting to watch. How would it be different from 30 planets? 2010-11-30T05:25:06 More time of expand before battle 2010-11-30T05:26:07 yeah, much longer battles 2010-11-30T05:26:09 Would that give particular bots time to play out their strengths? 2010-11-30T05:26:25 you could also have them starting with more than one home planet 2010-11-30T05:26:28 all over the map 2010-11-30T05:26:44 we can run maps like that on mcleopolds server if enough people are keen 2010-11-30T05:27:01 That would be feasible 2010-11-30T05:27:11 *** nn has joined #aichallenge 2010-11-30T05:27:13 i'm pretty keen to do 3-5 planet maps too 2010-11-30T05:27:25 see who can do the best there 2010-11-30T05:27:46 http://ai-contest.com/visualizer.php?game_id=9320803 2010-11-30T05:27:48 I am awesome on 1 planet maps 2010-11-30T05:27:51 second cock and balls map i've seen :) 2010-11-30T05:27:54 my generator is awesome 2010-11-30T05:27:59 it's a small cock 2010-11-30T05:28:19 or are they just really big balls? :P 2010-11-30T05:28:23 I think you need to go to a doctor 2010-11-30T05:28:27 :P 2010-11-30T05:28:45 cock and balls? with a LOT of imagination ... sigh: 2nd that 2010-11-30T05:28:50 has anyone released their code yet? 2010-11-30T05:28:57 Epona: plenty of people 2010-11-30T05:28:59 it's in the forums 2010-11-30T05:29:06 going to the forums* 2010-11-30T05:29:13 Epona: if you mean EVERYONE - no, I'm rank 500 or so. I don't bother 2010-11-30T05:29:23 you'll need to wait a few days for bocsimackos code, but it'll be out then 2010-11-30T05:29:34 Epona: http://ai-contest.com/forum/viewtopic.php?f=19&t=1168 2010-11-30T05:29:43 my code should not be looked at, it's not doing anything like what it appears to be with the tree 2010-11-30T05:29:53 o_O 2010-11-30T05:30:00 Woot! Won against Astek! 2010-11-30T05:30:00 * sigh is off to look at antimatroid's code 2010-11-30T05:30:08 @google "unleash your bots code" 2010-11-30T05:30:09 Frontier: Comprehensive guide to .htaccess- Blocking bad bots and site ...: ; Virtual Sales Bot - Simple Interactive Exit Traffic Sales Agents: ; How To Stay Motivated And Unleash Your Greatness: it never works as expected 2010-11-30T05:30:34 sigh: https://github.com/amstan/antimatroid/blob/master/Tree.cc 2010-11-30T05:30:48 inside make children, there's stuff inside the for loop for player 1 that should be fixed and outside the for loop 2010-11-30T05:30:55 and it shoudl also then be deleting unused made children 2010-11-30T05:31:03 haha, why is it in amstan's github? 2010-11-30T05:31:25 cause i have nfi how to use it :P 2010-11-30T05:32:18 aw, it's not that hard... even I figured out how to use it :P 2010-11-30T05:32:20 and lots of comments are outdated :P i couldn't be bothered cleaning before uploading 2010-11-30T05:32:34 i've never actually tried, i will at some point 2010-11-30T05:32:56 yeah, that's the main reason I haven't released my code yet... that and the fact that my bot sucks 2010-11-30T05:33:11 most of my stuff is in gamestate.cc if anyone is actually looking at my code 2010-11-30T05:35:02 medrimonia: i think we should be a tag team ha! 2010-11-30T05:35:35 smloh1: The eleventh streaking loosers? 2010-11-30T05:36:08 lol! no i just beat Astek right after you did 2010-11-30T05:36:21 Oh right I've just seen that^^ 2010-11-30T05:36:44 *** nn has quit IRC (Quit: Page closed) 2010-11-30T05:37:09 but i would guess that our rankings are kind of clustered since we climbed up pretty much at the same time 2010-11-30T05:39:12 I hope that this time will be climbing without falling =P 2010-11-30T05:39:26 arf loss 2010-11-30T05:40:00 well you're not at 11 yet so it shouldn't be a streak :P 2010-11-30T05:41:27 *** medrimonia1 has joined #aichallenge 2010-11-30T05:44:10 *** medrimonia has quit IRC (Ping timeout: 250 seconds) 2010-11-30T05:45:55 *** medrimonia1 is now known as medrimonia 2010-11-30T05:48:29 Top 10 players: bocsimacko(3913), _iouri_(3711), jimrogerz(3701), Slin-.-(3700), GreenTea(3692), _Astek_(3691), protocolocon(3673), Accoun(3671), george(3669), asavis(3643) 2010-11-30T05:53:11 *** smloh1 has quit IRC (Quit: Page closed) 2010-11-30T05:56:06 Does someone know if the contest when the contest will stop exactly? (0h00 or 23h59 the first December?) 2010-11-30T05:56:35 hmmm, i have an idea for how you might better approximate real time planet wars 2010-11-30T05:56:57 just make the turn time tiny, and have people maintain information across turns 2010-11-30T05:57:09 and much larger distances 2010-11-30T05:57:37 you could do distances in units of 4 each, to minimise used memory 2010-11-30T05:58:24 baned on forum 2010-11-30T05:58:40 you need to look up the definition of banned 2010-11-30T05:58:43 unlacky map set 2010-11-30T05:59:03 hate all canadian 2010-11-30T05:59:05 unlacky? 2010-11-30T05:59:38 rabidus_> http://ai-contest.com/forum/viewtopic.php?f=18&t=1188&p=7958#p7958 2010-11-30T06:00:05 "and please replace/add maps 2010-11-30T06:00:06 current map set unlucky for me" 2010-11-30T06:00:23 lol :D 2010-11-30T06:00:28 why janzert add the new maps set after I won 2nd place? 2010-11-30T06:01:07 why? 2010-11-30T06:01:45 Hmm Accoun, I heard you saying that they must add map because it'll be more fair. 2010-11-30T06:02:25 Now, they have added 75 maps and you're complaining... it doesn't seem right to me. 2010-11-30T06:02:46 there are now 150 maps? 2010-11-30T06:02:48 medrimonia> http://ai-contest.com/forum/viewtopic.php?f=18&t=1188&p=7958#p7958 2010-11-30T06:02:49 clearly the maps were chosen specifically to disadvantage Accoun 2010-11-30T06:02:55 Except if the only thing that matters for you is that you win... 2010-11-30T06:03:02 yes 2010-11-30T06:03:06 sigh: true! 2010-11-30T06:03:17 that's so obvious! 2010-11-30T06:03:24 there is no other reasonable answer 2010-11-30T06:03:38 Yes it's a conspiracy Accoun! 2010-11-30T06:03:58 yep 2010-11-30T06:04:14 last year he do it 2010-11-30T06:04:19 i'm not sure we're helping :) 2010-11-30T06:04:32 No but that's so... incredible 2010-11-30T06:05:12 what is incredible? 2010-11-30T06:05:24 it's a threat that you wouldn't follow through on optimally 2010-11-30T06:05:24 You, I've never seen someone likes you 2010-11-30T06:05:34 technically* 2010-11-30T06:05:38 bocsimacko's bot is pretty incredible 2010-11-30T06:06:09 =) win against asavis 2010-11-30T06:06:15 likes tom me not my target 2010-11-30T06:06:18 likes to me not my target 2010-11-30T06:06:39 im be stabile 2-3 on TCP 2010-11-30T06:07:03 near _iouri_ 2010-11-30T06:07:40 Accoun... in TCP, not everyone was playing at the same time , not everyone was playing even and some people had a lot of timeouts due to connections issues 2010-11-30T06:07:43 are astek, slin, jimrogerz and green tea on there? 2010-11-30T06:08:21 _iouri_ george bocsimacko is played 2010-11-30T06:08:34 _Astek_ if it as* bots 2010-11-30T06:08:50 i'm pretty sure as* bots are asavis 2010-11-30T06:09:16 anywey to lower on TCP than me 2010-11-30T06:09:21 anywey it lower on TCP than me 2010-11-30T06:09:25 you just showed nothing 2010-11-30T06:09:37 Accoun, if you loose you loose, if you had never lost a game on TCP, there it would be strange 2010-11-30T06:09:47 But your bot is not unbeatable at all 2010-11-30T06:09:58 no, i have a decent record against it on tcp 2010-11-30T06:10:02 no any bot unbeatable at all 2010-11-30T06:10:15 bocsis is a clear winner 2010-11-30T06:10:16 No but bocsimacko is clearly the best bot 2010-11-30T06:10:17 always has been 2010-11-30T06:10:21 You're not 2010-11-30T06:10:24 That's all 2010-11-30T06:10:33 im 2nd ! 2010-11-30T06:11:03 I see you on seventh place but maybe my eyes are wrong... 2010-11-30T06:12:41 *** dmj111 has joined #aichallenge 2010-11-30T06:13:07 go play minecraft.... o.o 2010-11-30T06:13:39 Epona: it's a trap 2010-11-30T06:13:50 although so are ai contests 2010-11-30T06:14:21 i'll just take a poke at thi... 3 months later.. oh god can't go to bed, one more game pleaaase!! 2010-11-30T06:14:35 lol 2010-11-30T06:14:39 antimatroid: So right... 2010-11-30T06:14:53 I didn't really play it that often 2010-11-30T06:15:04 just saw the minecraft computer 2010-11-30T06:15:07 I can't turn my computer off and I'm looking at it two or three times during the night^^ 2010-11-30T06:15:09 antimatroid: but at least didn't have anything better to do after your exams :P 2010-11-30T06:15:20 GreenTea_For_TCP played on TCP 2010-11-30T06:15:35 and be lower george 2010-11-30T06:15:37 http://ai-contest.com/visualizer.php?game_id=9328307 finish him 2010-11-30T06:16:04 does anyone have bocsimackos code 2010-11-30T06:16:17 Epona: a couple of days, he's cleaning 2010-11-30T06:16:26 Epona: I do. I'm cleaning it up. 2010-11-30T06:16:27 ok 2010-11-30T06:16:47 my old bot version make 8 games agnyst GreenTea_For_TCP 4 win 2010-11-30T06:16:55 mega1: you moving onto another ai contest at all after this? 2010-11-30T06:16:59 it seems there's actually quite a few 2010-11-30T06:17:04 no way 2010-11-30T06:17:08 :P 2010-11-30T06:17:11 I'm so much behind on work. 2010-11-30T06:17:54 GreenTea_For_TCP never be over me on TCP 2010-11-30T06:20:00 their last submission was november 27 2010-11-30T06:20:06 and i beat you on tcp numerous times 2010-11-30T06:20:27 antimatroid: best just to ignore him? 2010-11-30T06:20:50 if one can actually ignore people on irc, he's not seeing a thing i say :P 2010-11-30T06:20:57 i just respond anyway 2010-11-30T06:21:09 I'm hesitating, because sometimes it's hilarious to see what he writes 2010-11-30T06:21:15 last week TCP top - bocsimacko _iouri_ Accoun 2010-11-30T06:21:21 medrimonia: exactly :) 2010-11-30T06:23:01 finally, sin sun has been beating me all contest 2010-11-30T06:23:06 Everybody is losing Elo points now right? 2010-11-30T06:23:25 *** Utkarsh has quit IRC (Ping timeout: 264 seconds) 2010-11-30T06:23:50 *** wh1teside_ has joined #aichallenge 2010-11-30T06:23:50 maybe because the bottom bots are knocked out? 2010-11-30T06:23:54 Because I'm winning games, climbing up in ranking, but loosing elo points^^ 2010-11-30T06:24:00 rufes band almost have 4 people in the top 100 2010-11-30T06:24:02 * sigh is just making up random guesses 2010-11-30T06:24:05 sigh: Yep that was what I was thinking 2010-11-30T06:24:15 actually yeah, there was meant to be noise from that 2010-11-30T06:25:41 http://ai-contest.com/profile.php?user_id=10001 2010-11-30T06:25:46 well they were confident from the start 2010-11-30T06:26:00 on TCP all time Accon vs GreenTea_For_TCP 19 - 14 2010-11-30T06:27:06 antimatroid: woah, he timeout at the same time his opponent crashed: http://ai-contest.com/visualizer.php?game_id=9255020 2010-11-30T06:27:11 *** wh1teside has quit IRC (Ping timeout: 265 seconds) 2010-11-30T06:27:21 decent effort 2010-11-30T06:27:57 on TCP all time Accon14-16versions vs GreenTea_For_TCP 7 - 4 2010-11-30T06:28:57 hehe, mcleopold and i are dead even on elo 2010-11-30T06:29:51 *** Utkarsh has joined #aichallenge 2010-11-30T06:30:04 woah, this map didn't look like it would lead to an interesting game but I was wrong: http://ai-contest.com/visualizer.php?game_id=9321407 2010-11-30T06:30:29 action starts n move 70 :) 2010-11-30T06:30:57 lead to an uninteresting game? 2010-11-30T06:31:32 Waouw such a funny game^^ 2010-11-30T06:31:33 oh, didn'tt*, nm 2010-11-30T06:31:53 the game speed is certainly picking up 2010-11-30T06:32:02 now i just need to not flop 2010-11-30T06:37:50 hmm there's a big difference in elo between 16 and 14 2010-11-30T06:38:51 a new game every 11 min =) 2010-11-30T06:45:40 i haven't had one that recently D: 2010-11-30T06:45:54 give me my damn fix@! 2010-11-30T06:47:51 haha 2010-11-30T06:48:02 well ur bot is slower 2010-11-30T06:48:10 ur games last 11 min on itself :P 2010-11-30T06:48:13 the server doesn't keep track of that :P 2010-11-30T06:48:21 11 minutes isn't possible? 2010-11-30T06:48:25 is it 2010-11-30T06:48:30 Top 10 players: bocsimacko(3912), _iouri_(3708), _Astek_(3691), Slin-.-(3689), jimrogerz(3688), GreenTea(3685), Accoun(3673), protocolocon(3665), george(3652), asavis(3642) 2010-11-30T06:48:36 nah 2010-11-30T06:48:41 200 seconds is 3 minutes something 2010-11-30T06:48:59 3.3 2010-11-30T06:49:11 3.33333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333... 2010-11-30T06:49:16 yep 2010-11-30T06:49:25 i don't think i quite got enough 3's 2010-11-30T06:49:40 take that xrillo 2010-11-30T06:49:49 antimatroid: I'm assuming Error323 just divided 1500/135 to get 11 min 2010-11-30T06:50:02 indeed i am 2010-11-30T06:50:17 well it was 1510/128 :P 2010-11-30T06:50:20 but yeah sth like that 2010-11-30T06:50:50 Good afternoon/morning/evening/night all, I'm going to my english class^^ 2010-11-30T06:50:59 english class? 2010-11-30T06:51:04 Yep 2010-11-30T06:51:13 what grade are you in? 2010-11-30T06:51:21 Second year of Licence 2010-11-30T06:51:28 is that uni? 2010-11-30T06:51:34 what's the english class for? :P 2010-11-30T06:51:35 yep 2010-11-30T06:51:38 ./ 2010-11-30T06:51:41 ./s* 2010-11-30T06:51:57 Because they know that with only french, you can't go far in informatic 2010-11-30T06:52:00 ./s? 2010-11-30T06:52:16 sarcasm without getting unknown command 2010-11-30T06:52:21 haha 2010-11-30T06:52:33 /s 2010-11-30T06:52:38 spaces suck 2010-11-30T06:52:44 haha 2010-11-30T06:52:48 :P 2010-11-30T06:52:49 today is the day 2010-11-30T06:52:50 But we've only two hours a week, so we aren't learning a lot 2010-11-30T06:52:55 that i'm gonna clean my room 2010-11-30T06:53:01 its gonna be awesome 2010-11-30T06:53:07 my room is rather messy 2010-11-30T06:53:09 i should do that 2010-11-30T06:53:27 my room is enormously messy 2010-11-30T06:54:02 my room is exactly as messy as it needs to be 2010-11-30T06:54:30 haha 2010-11-30T06:55:47 *** sigh has quit IRC (Remote host closed the connection) 2010-11-30T06:56:10 *** sigh has joined #aichallenge 2010-11-30T06:59:16 bah, i crashed when i was winning 2010-11-30T06:59:43 that's like my third or fourth crash 2010-11-30T06:59:50 probably to be expected with the brokenness 2010-11-30T07:00:07 *** yozi has joined #aichallenge 2010-11-30T07:00:09 I just can't seem to get through the mountain of #)$(% that's in front of me 2010-11-30T07:00:20 on the leaderboard btw 2010-11-30T07:00:23 not in my room :P 2010-11-30T07:00:37 so you're room isn't filled with feces? :P 2010-11-30T07:00:51 *** yozi has quit IRC (Client Quit) 2010-11-30T07:00:53 only a little 2010-11-30T07:01:14 http://ai-contest.com/organization_profile.php?org_id=157 2010-11-30T07:01:17 i think i won that one 2010-11-30T07:02:24 how come smiley ain't there 2010-11-30T07:02:40 but yeh gratz =) 2010-11-30T07:02:53 that was for uni not australia 2010-11-30T07:02:59 http://ai-contest.com/country_profile.php?country_id=25 2010-11-30T07:03:02 got smiley anyway though 2010-11-30T07:14:59 2 crashes in a row :| 2010-11-30T07:15:42 *** rhinobot has joined #aichallenge 2010-11-30T07:18:03 ouch anti 2010-11-30T07:18:08 two timeouts probably 2010-11-30T07:20:26 i hope it's not a new trend :) 2010-11-30T07:20:28 i just won again 2010-11-30T07:21:23 my main goal now is to finish above mcleopold :P 2010-11-30T07:21:38 i'm 5 elo points above him atm 2010-11-30T07:23:51 *** narnach_mb has quit IRC (Quit: Leaving.) 2010-11-30T07:26:21 *** narnach has joined #aichallenge 2010-11-30T07:26:21 *** narnach is now known as narnach_mb 2010-11-30T07:29:50 *** narnach_mb has quit IRC (Client Quit) 2010-11-30T07:29:56 *** narnach has joined #aichallenge 2010-11-30T07:29:56 *** narnach is now known as narnach_mb 2010-11-30T07:39:18 jimrogerz _Astek_ Slin-.- GreenTea 3681 +-2 score, never seen on top on oficial and TCP 2010-11-30T07:40:06 *** janzert1 has joined #aichallenge 2010-11-30T07:46:34 *** GreenTea has joined #aichallenge 2010-11-30T07:48:29 Top 10 players: bocsimacko(3909), _iouri_(3703), jimrogerz(3682), GreenTea(3680), _Astek_(3680), Slin-.-(3679), Accoun(3673), protocolocon(3657), george(3639), asavis(3636) 2010-11-30T07:48:54 *** sigh has quit IRC (Remote host closed the connection) 2010-11-30T07:51:52 mega1: been reeading ur blog a bit, do you have any more pointers to UCT methods? 2010-11-30T07:51:53 Hello world 2010-11-30T07:52:01 hey GreenTea, janzert1 2010-11-30T07:52:29 "Wednesday Dec 1, 2010: official results and final rankings will be released. " in what time, who knows? 2010-11-30T07:52:58 good question 2010-11-30T07:53:31 stick around here and the forums 2010-11-30T07:53:55 Jeff and I are supposed to talk today to set a specific time he wants to do it 2010-11-30T07:54:39 ill be here for a little while making sure its spread to everyone. theres probably other people doing it as well 2010-11-30T07:58:25 janzert1: could i ask you a couple of questions on how to interact with arimaa engine? 2010-11-30T07:58:43 i find it hard to find information and read other peoples bots :P, they arne't hard 2010-11-30T07:59:00 I may or may not be not be around for long but go ahead 2010-11-30T07:59:53 1. my bot gets run from scratch for each turn unlike this yes? can i write to files at least?| 2010-11-30T07:59:54 2. what is the string inputed to my engine? is it their turn as specified, or do i get my turn and comments etc. passed to me as well at different points? 2010-11-30T08:00:20 antimatroid: with the default interface it is restarted every turn yes, but... 2010-11-30T08:00:56 you probably want to use AEI instead of the default http://arimaa.janzert.com/aei/ 2010-11-30T08:01:09 yeah i've got that :) 2010-11-30T08:01:26 ahh, ok. under AEI your bot stays running for the whole game 2010-11-30T08:01:29 i'm just trying to set my c++ files up and working out how to handle input 2010-11-30T08:01:42 so i can store and update my tree between turns? cool 2010-11-30T08:01:46 that makes things easier 2010-11-30T08:01:57 *** mceier has joined #aichallenge 2010-11-30T08:02:02 input output is over stdin/out similar to planetwars 2010-11-30T08:02:32 2g Ra2n Ra3e Rb3n Rb4e 2010-11-30T08:02:32 2s ra7s ra6s ra5e rb5e 2010-11-30T08:02:32 would i only ever see their lines of that coming from input? 2010-11-30T08:02:38 or do they send my moves back to me? 2010-11-30T08:02:50 they == the engine :) 2010-11-30T08:03:22 no, the protocol description at http://arimaa.janzert.com/aei/aei-protocol.html or in the interface download as aei-protocol.txt(?) and then also the very simple engine in the download are probably the best starting spots 2010-11-30T08:03:45 i hadn't found that page, thanks, i'll read :) 2010-11-30T08:04:06 basically each line of input is a command from the controller 2010-11-30T08:04:32 of the form "command 2010-11-30T08:05:31 the protocol is fairly similar in form to the UCI protocol for chess engines if you are familiar with that 2010-11-30T08:05:43 Error323: I searched for UCT and read everything. 2010-11-30T08:05:56 then searched for uct and go and read everything again. 2010-11-30T08:06:07 mega1: yeah i'm reading some papers now 2010-11-30T08:06:46 cool stuff, too bad it didn't work out at trno ;) 2010-11-30T08:06:47 tron 2010-11-30T08:07:31 *** Epona has quit IRC (Ping timeout: 255 seconds) 2010-11-30T08:07:39 i want to do something that isn't minimax 2010-11-30T08:07:42 i'll also have to learn more lisp i see :/ 2010-11-30T08:07:45 i don't think arimaa is a good choice for that 2010-11-30T08:07:48 but arimaa seems fun too 2010-11-30T08:07:50 because otherwise i won't understand ur code. 2010-11-30T08:07:59 iouri is 2/0/2 against me and 12/5/133 in local testing 2010-11-30T08:08:20 either he's lucky or I optimized for my laptop. 2010-11-30T08:08:27 mega1> what maps set used? 2010-11-30T08:08:27 luck of the draw? 2010-11-30T08:08:29 heh, i'm 49/1/100 against iouri 2010-11-30T08:08:42 Accoun: finals1 + finals2 2010-11-30T08:08:49 but stuck at 80ish :( 2010-11-30T08:08:51 i have lost every game to mcleopold in the final tournament, but seemed to have him on tcp with those two exact bots 2010-11-30T08:09:01 mega1> want to test my bot? 2010-11-30T08:09:13 Accoun: sure 2010-11-30T08:09:41 mega1> i can sent MyBot.exe 2010-11-30T08:10:03 Accoun: does it take much cpu? 2010-11-30T08:10:19 mega1: do you use your full second much? 2010-11-30T08:10:24 mega1> not 2010-11-30T08:10:41 mega1> i can run 16 copies on my dual core 2010-11-30T08:10:55 antimatroid: only in the opening phase 2010-11-30T08:11:41 for the naish equilibrium? 2010-11-30T08:12:15 no, for 4 ply alpha beta 2010-11-30T08:12:20 i didn't think of this till too late, i wonder if you tried it? make the game tree with each player making one move per node, but only actually add each of player ones children where the payoff for that turn is greater than doing nothing (with them responding to nothing) 2010-11-30T08:12:22 oh i see 2010-11-30T08:12:32 that probably would have got me a few more moves out each turn 2010-11-30T08:13:06 greater than or equal* 2010-11-30T08:13:12 I tried and it suck for some reason. didn't have the time to investigate. 2010-11-30T08:13:24 yep fair enough 2010-11-30T08:14:10 that's the best idea i have for a minimax tree with this, i really don't trust alpha beta pruning while still extending 2010-11-30T08:14:30 not trust? 2010-11-30T08:14:37 i should come up with an example to show my reasoning and see if i can show what i mean 2010-11-30T08:14:39 i'll do that now 2010-11-30T08:14:51 *** aerique has quit IRC (Quit: ...) 2010-11-30T08:14:53 yes, i'm curiuos 2010-11-30T08:15:20 either you or me doesn't understand its concept fully :P 2010-11-30T08:17:49 i got another win against bix0r 2010-11-30T08:18:01 either he screwed up at the end, or he played resbumit too much 2010-11-30T08:18:11 heh 2010-11-30T08:19:15 awww, my online alpha beta applet thing is down 2010-11-30T08:20:22 *** Sylph2 has joined #aichallenge 2010-11-30T08:20:46 *** fmeyer has joined #aichallenge 2010-11-30T08:20:48 just try to explain it 2010-11-30T08:22:11 yeah i was just about to, consider this situation 2010-11-30T08:22:52 *** Sylph has quit IRC (Ping timeout: 240 seconds) 2010-11-30T08:23:00 you made a move, to which they consider attacking, but you could then own their source planets of that attack if you made another move, but if you use alpha beta before extending, you might cut that branch out before extending to see that you would have come good again 2010-11-30T08:23:05 *** fmeyer has quit IRC (Client Quit) 2010-11-30T08:23:27 *** Palmik has joined #aichallenge 2010-11-30T08:23:42 one could argue that's a fault of the evaluation function/move selection 2010-11-30T08:23:50 but wherever you want to place the problem, there's an issue there 2010-11-30T08:24:10 *** Appleman1234 has joined #aichallenge 2010-11-30T08:24:32 does that make sense at all? 2010-11-30T08:25:12 you could make an example with 3 planets and a triangle 2010-11-30T08:25:30 you could take a neutral closer to you that's snipable, but not without you then sniping their home planet 2010-11-30T08:25:41 which is easy to account for with just one source, but what about when you have more? 2010-11-30T08:28:03 Error323: you're not allowed to make me explain and not respond :P 2010-11-30T08:28:07 throw me a bone here :) 2010-11-30T08:29:19 lol, i'm thinking. I don't really understand your example, but alphabeta will provably always perform the same as minimax 2010-11-30T08:29:54 it would if you use it while evaluating the entire tree you plan to use 2010-11-30T08:30:00 but i don't trust it when cropping a tree i still plan to extend, not unless the payoff is certain 2010-11-30T08:30:17 you could do some pruning with that information and get exactly the same results (if not better) than minimax guaranteed 2010-11-30T08:30:32 mega1: ^^ am i just an idiot? 2010-11-30T08:31:20 yes, extensions and reductions/pruning will break the alpha-beta minimax equivalency 2010-11-30T08:31:46 but are still very frequently better in practice because of the extra depth that can be obtained 2010-11-30T08:31:56 yeah, i had been thinking that lately :) 2010-11-30T08:32:12 hence my last idea that didn't get implemented properly 2010-11-30T08:32:31 i figured it's also so messy with this it's probably better to consider more moves 2010-11-30T08:32:44 especialyl as the depth converted to number of moves i made per turn here 2010-11-30T08:32:50 my move selection is heavy (basically a 1 ply search), but it's still not good enough to make alpha-beta work reliably 2010-11-30T08:34:15 my move selection is probably the most expensive item, second is calculating usable ships at a planet, as each time it needs updating i have to do two whole sweeps of that planets future states, no matter when it was changed 2010-11-30T08:35:55 I should've kept pursuing my alphabeta branch :P 2010-11-30T08:36:12 and change my action generation 2010-11-30T08:37:56 hmmm, i just hope i do a little better with the next game 2010-11-30T08:37:59 this one stumped me 2010-11-30T08:38:30 yeah me too, I have to delve into the domain more 2010-11-30T08:38:41 which i didn't this time 2010-11-30T08:38:58 i know some basic game theory 2010-11-30T08:39:05 it's very mathsy after a little bit though 2010-11-30T08:39:17 i haven't done enough analysis or algebra to understand most of it 2010-11-30T08:39:28 in a few years :) 2010-11-30T08:39:36 lol "mathsy" xD 2010-11-30T08:39:55 john von neumann is a god 2010-11-30T08:40:19 i am astounded at how much that man did 2010-11-30T08:40:24 and he died fairly young 2010-11-30T08:40:41 he and turing 2010-11-30T08:40:46 turing is a really sad story 2010-11-30T08:40:58 yep 2010-11-30T08:41:03 got banned and castrated for being gay -.- 2010-11-30T08:41:23 although the actual castraction is a lot less bad than what i first thought when reading that 2010-11-30T08:41:26 or no he got castrated 2010-11-30T08:41:29 instead of banned iirc 2010-11-30T08:41:34 he could chose 2010-11-30T08:41:36 not that it isn't terrible, just not what i pictured :) 2010-11-30T08:41:58 if you use a computer you use a gay invention 2010-11-30T08:42:04 lol 2010-11-30T08:42:13 my mum is gay, i'm a gay invention :) 2010-11-30T08:42:24 haha 2010-11-30T08:42:40 http://www.markallencam.com/?p=144 2010-11-30T08:42:46 oh man mistmanov and me are sooo close now :S 2010-11-30T08:42:56 ehehe 2010-11-30T08:43:06 i can't tell if she's serious or being sarcastic :\ 2010-11-30T08:43:16 i'm going with sarcastic 2010-11-30T08:43:43 omg is that real 2010-11-30T08:43:57 she has a rainbow flag 2010-11-30T08:44:06 @rainbow yay it's a fucking rainbow! 2010-11-30T08:44:06 antimatroid: yay it's a fucking rainbow! 2010-11-30T08:44:06 it is sarcasm 2010-11-30T08:46:54 davidd: unfortunatly there are enough places where ppl agree to this -.- 2010-11-30T08:47:35 http://quotenil.com/git/?p=planet-wars.git;a=summary 2010-11-30T08:47:38 i've been quite surprised at how racist some people in australia have been the last few years, at first i thought the media was being retarded, but there's been some pretty bad attacks 2010-11-30T08:48:01 some guy i have actually personally met and have known about for a while raped and killed a chinese uni chick last year :\ 2010-11-30T08:48:06 the piece of scum is now behind bars 2010-11-30T08:48:22 mega1: <3 2010-11-30T08:48:23 mega1: who00000000t 2010-11-30T08:48:29 Top 10 players: bocsimacko(3905), _iouri_(3700), jimrogerz(3679), GreenTea(3672), Slin-.-(3671), _Astek_(3669), Accoun(3665), protocolocon(3647), asavis(3628), george(3627) 2010-11-30T08:49:31 so many maps with close start planets.. 2010-11-30T08:49:37 ;;;; This file must be here because the contest server compile ;;;; environment relies on it to recognize that it is a Common Lisp ;;;; submission. :D 2010-11-30T08:49:50 GreenTea: sorry :( 2010-11-30T08:50:04 No, this is good for me) 2010-11-30T08:50:13 i was in favour of possibly increasing the min starting angle distance 2010-11-30T08:50:21 ah okay, then you're welcome :P 2010-11-30T08:50:22 the new mapset has been horror for me btw 2010-11-30T08:50:22 hehe 2010-11-30T08:50:32 i don't think it did me any favours :P 2010-11-30T08:50:40 but i think it was better 2010-11-30T08:50:45 should have been earlier 2010-11-30T08:51:10 ye 2010-11-30T08:51:14 oh well 2010-11-30T08:51:40 actually, i wouldn't have a clue how to do symmetric maps for 2 player mario kart 2010-11-30T08:52:01 except maybe have them start in the same place, then once separated you can't go through each other anymore 2010-11-30T08:52:40 imo: RTS game, or something like transport tycoon would be cool aswell 2010-11-30T08:53:52 with mario, think controls could just be, set acceleration/deacceration (max jump size specified), max speed, set rate of angle turn (max rate specified) and fire weapon 2010-11-30T08:54:19 and possibly all flat maps 2010-11-30T08:55:14 then weapons could be bots? that do their own specified thing 2010-11-30T08:55:40 Error323 the anti-computer/anti-internet crowd is a group i dont take that seriously, although they may really mean what they say 2010-11-30T08:55:43 mega1: wasn't maxplanets > 23? 2010-11-30T08:55:52 and i think it is safe to post that link here 2010-11-30T08:56:00 haha ofc 2010-11-30T08:56:02 Error323: yes, I forgot to change that 2010-11-30T08:56:03 *** Appleman1234 has quit IRC (Read error: Connection reset by peer) 2010-11-30T08:56:39 Error323: as long as the ship count is representable by 16 bits it's ok 2010-11-30T08:56:59 does anyone else like the idea of mario kart ai? or am i by myself? :) 2010-11-30T08:57:01 ah ok, well i think its safe to assume that :P 2010-11-30T09:01:19 yea i need to delve into lisp more 2010-11-30T09:01:32 ` and ' and ((())))((((())(((())))))))((((((((((((((( 2010-11-30T09:01:36 xD 2010-11-30T09:08:13 http://www.ai-contest.com/visualizer.php?game_id=9347535 2010-11-30T09:08:29 *** ZogStriP has joined #aichallenge 2010-11-30T09:09:59 wow mega1 2010-11-30T09:10:06 thats... awesome 2010-11-30T09:10:46 looks good, but unfortunately it just timed out earlier/later when thinking 2010-11-30T09:11:12 mega1: what made you decide to make that move? 2010-11-30T09:11:14 it's not that it was playing with the nerves of the opponent by waiting 60 turns 2010-11-30T09:11:30 antimatroid: I think a timeout 2010-11-30T09:11:31 ah i see 2010-11-30T09:11:34 *** davidjliu has joined #aichallenge 2010-11-30T09:11:54 *** exaide has joined #aichallenge 2010-11-30T09:12:14 well it paid off :P 2010-11-30T09:12:18 mega1: what linux distro do you run 2010-11-30T09:12:26 debian testing 2010-11-30T09:12:30 mkej 2010-11-30T09:13:03 i had wrong lisp hehe 2010-11-30T09:13:10 sbcl is what i need 2010-11-30T09:13:22 * Error323 gentoo 2010-11-30T09:16:12 mega1: you had to make ur own heap etc? 2010-11-30T09:16:29 no, that's a patch for sbcl 2010-11-30T09:16:39 oh 2010-11-30T09:16:40 backported from current sources 2010-11-30T09:17:06 it's a 3 year old sbcl with bugs in timeout handling 2010-11-30T09:17:38 one of which I personally put there to waste a day on years later 2010-11-30T09:18:08 timer.lisp is thus not really part of the codebase 2010-11-30T09:19:00 I see 2010-11-30T09:22:54 \o/ it compiles 2010-11-30T09:29:22 hmmm i think i might take my time with arimaa and do the following years contest 2010-11-30T09:29:30 one month really isn't enough time to give that a good go 2010-11-30T09:31:45 *** davidjliu has quit IRC (Quit: Page closed) 2010-11-30T09:35:10 http://i.imgur.com/9vxH5.jpg :D 2010-11-30T09:35:20 anyone plaing borderlends or Titan Quest Immortal Throne ? 2010-11-30T09:35:21 that's the douchebag minister in australia doing all the censorship crap 2010-11-30T09:35:38 it's got the ring and everything 2010-11-30T09:38:04 @janzert1 2010-11-30T09:38:05 medrimonia: You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2010-11-30T09:38:24 :'( 2010-11-30T09:38:36 @seen janzert1 2010-11-30T09:38:36 medrimonia: janzert1 was last seen in #aichallenge 1 hour, 6 minutes, and 50 seconds ago: but are still very frequently better in practice because of the extra depth that can be obtained 2010-11-30T09:39:27 janzert1: Do you already know when and how the cutoff will go now? 2010-11-30T09:39:41 finally, i beat smloh1 2010-11-30T09:40:32 *** rebelxt_ has joined #aichallenge 2010-11-30T09:44:13 another crash under my best 2010-11-30T09:44:20 i must be the highest crash bot aha 2010-11-30T09:46:30 do you know why you crash? 2010-11-30T09:48:12 yeah, my tree is completely broken 2010-11-30T09:48:18 i'm surprised it doesn't always crash 2010-11-30T09:48:31 Top 10 players: bocsimacko(3906), _iouri_(3697), jimrogerz(3678), _Astek_(3666), GreenTea(3666), Slin-.-(3661), Accoun(3655), protocolocon(3640), asavis(3624), george(3623) 2010-11-30T09:48:33 there is memory leaks i'm pretty sure 2010-11-30T09:48:51 due to the tree not doing what i meant 2010-11-30T09:49:12 ha, i beat george 2010-11-30T09:49:25 only cause he crashed 2010-11-30T09:51:25 ^^ 2010-11-30T09:55:49 *** amstan has quit IRC (Remote host closed the connection) 2010-11-30T09:55:49 i can't even win the best crash bot award :P 2010-11-30T09:55:59 at least i think i have australia in the bag 2010-11-30T09:56:55 *** Accoun has quit IRC () 2010-11-30T09:58:00 well its a pretty exciting match for #1 and #2 of holland 2010-11-30T09:58:05 unfortunatly :P 2010-11-30T09:58:32 ouch, 8/2/140 w/d/l vs mega1's submission 2010-11-30T09:58:51 from which 7 in final-1 set 2010-11-30T09:58:55 CANADIAN CHITERS 2010-11-30T09:59:10 iouri's 12/5/133 2010-11-30T09:59:27 And me mega1: =P 2010-11-30T09:59:33 yes final-1 was very bad for me 2010-11-30T09:59:49 9/3/63 on final-1 but I don't know on final-2 2010-11-30T09:59:58 final is bad for me, against multiple players 2010-11-30T10:00:04 final2 is 2010-11-30T10:00:10 medrimonia: you can run it yourself now 2010-11-30T10:00:40 I'll try 2010-11-30T10:06:37 how big a map set was benzendrine running? 2010-11-30T10:07:13 http://ai-contest.com/profile_games.php?user_id=8565 2010-11-30T10:07:18 look at that damn page of games 2010-11-30T10:07:54 I'm on my longest winning streak in the finals 2010-11-30T10:08:09 best time to shine :) 2010-11-30T10:08:15 http://ai-contest.com/visualizer.php?game_id=9341258 2010-11-30T10:08:23 ^^^ incredibly lucky 2010-11-30T10:08:46 yep 2010-11-30T10:09:20 wow 2010-11-30T10:09:22 i think rankings aren't far off where they're going to stay near the top 2010-11-30T10:09:36 *** p4p4 has joined #aichallenge 2010-11-30T10:10:04 i think ur right 2010-11-30T10:10:16 it's hard to know 2010-11-30T10:10:21 been hovering at 80-90 for 24 hrs now 2010-11-30T10:10:26 actually, i think davidjliu will be above george 2010-11-30T10:10:31 poor bugger had a very rough start 2010-11-30T10:10:39 and asavis just made it into the top 10? 2010-11-30T10:10:41 I checked likelihood of of superiority but now I think it's not right 2010-11-30T10:10:59 how so 2010-11-30T10:11:02 mega1: did you end up getting accouns bot? 2010-11-30T10:11:20 no, I told him to run it himself because it's a windows exe 2010-11-30T10:11:31 hehe 2010-11-30T10:11:34 :P 2010-11-30T10:11:35 I have a linux jail set up 2010-11-30T10:11:43 jail 2010-11-30T10:11:44 ? 2010-11-30T10:11:47 I have a windows virtual machine too, but that's for work 2010-11-30T10:12:10 Error323: hey, that's where your binaries ran too :-) 2010-11-30T10:12:21 technically australia spells it gaol, but nobody follows that, it's a terrible spelling 2010-11-30T10:12:23 *** medrimonia has quit IRC (Quit: Leaving.) 2010-11-30T10:12:27 *** medrimonia1 has joined #aichallenge 2010-11-30T10:12:32 lol, I don't know what jail means in that context. 2010-11-30T10:13:04 *** JamesMG has joined #aichallenge 2010-11-30T10:13:06 Error323: a sandbox, a safe restricted env to run untrusted binaries in 2010-11-30T10:13:18 no offense intended 2010-11-30T10:13:23 *** rebelxt_ has quit IRC (Quit: Page closed) 2010-11-30T10:13:27 yeah my binaries i send, explored your system scanning your *.lisp and uploading to my server :P 2010-11-30T10:13:37 mega1: none taken ;) 2010-11-30T10:14:09 except there were no sources in the jail and host only networking 2010-11-30T10:14:10 you've been more than helpful for someone trying to stay on top imo 2010-11-30T10:14:21 agree 2010-11-30T10:14:47 Argh: http://ai-contest.com/visualizer.php?game_id=9349924 2010-11-30T10:15:01 should've spent the last week porting it to something not C# :( 2010-11-30T10:15:01 JamesMG: i've had like at least 6 crashses :) 2010-11-30T10:15:12 oh, everyone is getting them? 2010-11-30T10:15:18 not everybody 2010-11-30T10:15:20 *** Utkarsh has quit IRC () 2010-11-30T10:15:22 i'm not :P 2010-11-30T10:15:28 my bot just sux balls ^_^ 2010-11-30T10:15:32 my tree is very screwed 2010-11-30T10:15:34 I'm not too 2010-11-30T10:15:47 this is the first one I have found 2010-11-30T10:15:52 george has had some, mcleopold too 2010-11-30T10:15:53 but I haven't been watching all my games 2010-11-30T10:16:03 i have watched a good portion of mine, not all 2010-11-30T10:16:19 if i go to sleep i usually can't be bothered watching all the new ones when i wake up 2010-11-30T10:16:27 note the if :P 2010-11-30T10:16:33 C# is still crashing a lot? 2010-11-30T10:16:44 jmrogerz is c#, so is astek 2010-11-30T10:16:47 how many of them were turn1 crashes antimatroid? 2010-11-30T10:16:49 so i'd say probably not 2010-11-30T10:16:57 a couple, not all 2010-11-30T10:17:05 those are the ones that scare me 2010-11-30T10:17:13 i'm in 33rd though, so it hasn't hurt me 2010-11-30T10:17:23 i don't think it'd be much higher without them 2010-11-30T10:17:43 mine isn't going to do particularly well either way 2010-11-30T10:18:11 i'll just keep grumbling to myself "next time" :) 2010-11-30T10:18:25 i wanna know when the next contest will be :P 2010-11-30T10:18:38 i'd say around jan-march next year it will start 2010-11-30T10:18:58 but that's a guess based off of tron and this 2010-11-30T10:18:59 oh that's pretty quick 2010-11-30T10:19:09 fron ended in february :P 2010-11-30T10:19:12 tron* 2010-11-30T10:19:14 *** ulope_ has quit IRC (Quit: ZNC - http://znc.sourceforge.net) 2010-11-30T10:19:51 I'm really looking forward to the next one 2010-11-30T10:20:07 this one was my intro to C#(mistake) and game theory 2010-11-30T10:21:00 *** aerique has joined #aichallenge 2010-11-30T10:22:20 *** Accoun has joined #aichallenge 2010-11-30T10:24:27 i really like the idea of group contests as well next time 2010-11-30T10:24:50 where websites/rufes band etc. make a group, and separate to the final contest a contest and leaderboard is generated souly for that group 2010-11-30T10:25:00 and irc group could even be made :D 2010-11-30T10:25:24 but obviously games for those can't be counted towards the official contest 2010-11-30T10:25:57 90% of all users would be in the reddit group ehe 2010-11-30T10:26:21 I don't even know how to compare countries in this contest. 2010-11-30T10:26:52 i think last tiem they just did number of bots in top 200, then tie broke on highest ranked bot 2010-11-30T10:27:13 seems reasonable enough 2010-11-30T10:27:30 but don't even need to do a ranking across groups, just inside each one 2010-11-30T10:28:03 it feels weird to say groups in the normal sense after two years of algebra classes 2010-11-30T10:30:11 I'd probably try to fit a suitable distribution to the elo points or the ranks 2010-11-30T10:31:17 or just lump together their games in the top 200 and calc bayeselo 2010-11-30T10:36:22 that's a good idea =) 2010-11-30T10:36:24 *** phreeza has joined #aichallenge 2010-11-30T10:36:49 maybe the pentagon has a job for mega1 http://www.economist.com/node/17572232?story_id=17572232?fsrc=spblg/251110 2010-11-30T10:37:51 i wouldn't be interested in creating war robots personally 2010-11-30T10:38:25 * antimatroid should keep reading before opening his mouth 2010-11-30T10:39:18 i can just imagine first attempts at sending robots into war zones, they'd be really stupid and the enemy would just laugh as they took them out 2010-11-30T10:43:48 think flying, not walking... 2010-11-30T10:48:31 like so 2010-11-30T10:48:31 Top 10 players: bocsimacko(3901), _iouri_(3694), jimrogerz(3672), _Astek_(3661), GreenTea(3654), Slin-.-(3645), Accoun(3644), protocolocon(3630), asavis(3618), george(3610) 2010-11-30T10:48:33 http://xkcd.com/652/ 2010-11-30T10:55:25 bummer, another turn 1 crash just now 2010-11-30T10:55:44 grr, I should start to win or Accoun outstrip me.. 2010-11-30T11:12:40 *** superflit has quit IRC (Quit: superflit) 2010-11-30T11:13:38 whow to wath p2p internet tv? 2010-11-30T11:13:57 sopcast? 2010-11-30T11:14:00 this is curious. lost 3 times to iouri but on my laptop I can find only one of those losses among those played locally 2010-11-30T11:14:46 mega1> low server perfomance this time 2010-11-30T11:15:00 *** amstan has joined #aichallenge 2010-11-30T11:15:00 *** ChanServ sets mode: +o amstan 2010-11-30T11:15:10 my perl bot crashed all time bul lokali play well 2010-11-30T11:15:16 *** aerique has quit IRC (Quit: ...) 2010-11-30T11:15:26 my perl bot crashed all time but localy play well 2010-11-30T11:16:08 I think servers have two 2.5Ghz intel cpus 2010-11-30T11:16:24 but with two games running at the same time ... 2010-11-30T11:16:30 who knows? 2010-11-30T11:17:33 25 games per minut 2010-11-30T11:17:51 it many bots run at same time 2010-11-30T11:20:30 *** phreeza has quit IRC (Quit: kthxbai) 2010-11-30T11:21:03 *** amstan has quit IRC (Ping timeout: 276 seconds) 2010-11-30T11:24:55 damn, I found the last loss and it's a win here 2010-11-30T11:25:08 hmmm bugger 2010-11-30T11:25:14 the servers aren't very good 2010-11-30T11:25:24 *** wh1teside has joined #aichallenge 2010-11-30T11:25:33 next time i would live for players to be given separate seconds to make their moves 2010-11-30T11:25:37 like* 2010-11-30T11:25:58 hmmm 2010-11-30T11:26:08 running multiple games side by side is not a good decision 2010-11-30T11:26:27 it can very well be that in the end I'm benefitting from that 2010-11-30T11:26:30 mistmanov wins 48 out of 150 against mega1 but also loses 50 out of 150 against a 500 ranked player 2010-11-30T11:26:55 Error323: is mistmanov kind of kamikaze? 2010-11-30T11:27:01 i think so yes 2010-11-30T11:27:08 but that explains his current rank 2010-11-30T11:27:11 mega1: i don't think anyone (except maybe accoun) coiuld argue that's why you're winning 2010-11-30T11:27:54 yes, not likely. But I hate losing to iouri and slin. 2010-11-30T11:28:14 *** wh1teside_ has quit IRC (Ping timeout: 250 seconds) 2010-11-30T11:28:27 :P perfectionist 2010-11-30T11:28:38 haha 2010-11-30T11:28:40 it can also be that my laptop is _slower_ 2010-11-30T11:28:50 and that's somehow good for my bot. 2010-11-30T11:29:00 Error323: what kind of machine do you have? 2010-11-30T11:29:05 Hostname: jarvis - OS: Linux 2.6.35-gentoo-r5/x86_64 - CPU: 4 x Intel(R) Core(TM)2 Quad (2666.906 MHz) - Processes: 184 - Uptime: 1d 1h 25m - Load Average: 0.21 - Memory Usage: 813.32MB/2008.39MB (40.50%) - Disk Usage: 654.01GB/1768.99GB (36.97%) 2010-11-30T11:29:28 i could run iouri vs you here 2010-11-30T11:29:33 could you please try running ... 2010-11-30T11:29:34 yes 2010-11-30T11:30:01 *** rebelxt has joined #aichallenge 2010-11-30T11:31:17 *** amstan has joined #aichallenge 2010-11-30T11:31:17 *** ChanServ sets mode: +o amstan 2010-11-30T11:32:48 cool iouri wrote a nice article: http://iouri-khramtsov.blogspot.com/2010/11/google-ai-challenge-planet-wars-entry.html 2010-11-30T11:33:14 mega1: its running btw 11 games so far 2010-11-30T11:35:38 Error323: great post by iouri 2010-11-30T11:35:46 very good post 2010-11-30T11:35:52 *** amstan has quit IRC (Ping timeout: 240 seconds) 2010-11-30T11:35:55 wut, you guys read fast 2010-11-30T11:36:09 i couldn't be bothered reading it all :) 2010-11-30T11:36:12 it's 3 40am 2010-11-30T11:37:20 *** yasith has quit IRC (Read error: Connection reset by peer) 2010-11-30T11:37:32 *** yasith_ has joined #aichallenge 2010-11-30T11:37:33 http://ai-contest.com/visualizer.php?game_id=9371555 lol 2010-11-30T11:37:43 cool map though 2010-11-30T11:38:59 mega1: ouch another loss to iouri :P 2010-11-30T11:39:27 yes, that's what prompted me to look into it 2010-11-30T11:40:07 Error323: i read his algorithm description on the forums earlier. and it was just like a union of my code and my @todo list ;-) awesome he took time to write even more detailed version 2010-11-30T11:41:34 so, iouri skipped unit tests as well. did anyone have any automated unit tests? ;) 2010-11-30T11:42:57 no 2010-11-30T11:46:05 mega1: intermediate results: 2010-11-30T11:46:06 Number of player 1 wins: 70 2010-11-30T11:46:06 Number of player 2 wins: 5 2010-11-30T11:46:06 Number of draws: 4 2010-11-30T11:46:12 you begin player 1 2010-11-30T11:46:35 that's more or less what I see here 2010-11-30T11:46:58 i'm glad we threw out the draw maps we did 2010-11-30T11:47:02 there's already enough of them 2010-11-30T11:47:02 it seems that I still overestimated server power 2010-11-30T11:47:03 "At last, I am free of it! Google AI Challenge (Planet Wars) is over. No more spending all my free time thinking about how to eek out extra few wins, no more rushing home after work to start testing new bots on the TCP server." oh, how i understand him :) 2010-11-30T11:47:15 GreenTea: we all feel that pain 2010-11-30T11:47:17 hahaha 2010-11-30T11:47:33 ) 2010-11-30T11:47:43 *** superflit has joined #aichallenge 2010-11-30T11:48:17 i don't think we could remove all maps without a non-snipable neutral without potentially removing interesting maps 2010-11-30T11:48:30 GreenTea: same here, lots of suffering during the contest 2010-11-30T11:48:31 Top 10 players: bocsimacko(3891), _iouri_(3689), jimrogerz(3658), _Astek_(3650), GreenTea(3645), Accoun(3637), Slin-.-(3637), protocolocon(3625), asavis(3614), george(3610) 2010-11-30T11:48:48 take that oldman 2010-11-30T11:50:14 if pairing cutoff was 100 game counts for us would be through the roof 2010-11-30T11:50:15 i want 2010-11-30T11:51:56 *** narnach_mb has quit IRC (Quit: Leaving.) 2010-11-30T11:52:52 at the end cutoff must be 10 !! 2010-11-30T11:53:11 not going to happen i don't think 2010-11-30T11:53:19 i don't think it's even going to make it to 1000 2010-11-30T11:53:31 why not 2010-11-30T11:53:33 maybe for last 30 minutes.. 2010-11-30T11:53:44 mega1: ouch :p 2010-11-30T11:53:57 antimatroid: ? 2010-11-30T11:54:12 you lost to fglider 2010-11-30T11:54:14 60th place 2010-11-30T11:54:32 that's ok, I'm more annoyed by losing to iouri 2010-11-30T11:54:44 heheh 2010-11-30T11:54:47 its almost finishd 2010-11-30T11:55:14 *** Mistmanov has joined #aichallenge 2010-11-30T11:55:39 Mistmanov: still holding #1 >:) 2010-11-30T11:56:07 -_- 2010-11-30T11:56:11 we were equal for w hile :p 2010-11-30T11:56:17 yeah 2010-11-30T11:56:22 i did some tests with your bot 2010-11-30T11:56:34 and whilest performing better against good bots 2010-11-30T11:56:42 its also performing much worse against bad bots 2010-11-30T11:56:47 yeah ^^ 2010-11-30T11:57:00 this explains a lot :P 2010-11-30T11:57:36 it's unfortunate, but what can you do.. 2010-11-30T11:57:58 mega1: 2010-11-30T11:57:58 Number of player 1 wins: 135 2010-11-30T11:57:59 Number of player 2 wins: 9 2010-11-30T11:57:59 Number of draws: 6 2010-11-30T11:58:07 bocsimacko vs iouri 2010-11-30T11:58:11 go and murder mega1 and steal his brain power 2010-11-30T11:58:12 *** nn has joined #aichallenge 2010-11-30T11:58:14 beg janzert to let me play 100 games against the top 10 2010-11-30T11:58:29 Error323: thanks, it seems my bot benefits from cpu power 2010-11-30T11:59:39 yep 2010-11-30T11:59:57 just a comment if you make the pairing cutoff too narrow the nature of the field of players changes too much as well, any narrowing is of course a trade off on this 2010-11-30T12:00:04 *** janzert1 is now known as janzert 2010-11-30T12:00:20 janzert: yeah e.g. mistmanov would go through the roof 2010-11-30T12:00:24 janzert: i like the idea of 100 bots running for a day 2010-11-30T12:00:25 whilest i would drop :P 2010-11-30T12:00:52 what happens atm if a bot drops below the cutoff? 2010-11-30T12:01:00 does the bot coming above get more games? 2010-11-30T12:01:40 that's a bit iffy i think, maybe it shouldn't be made so low 2010-11-30T12:02:01 also even if we narrow it top 100 it still won't be enough to get confidence in the ordering, that would require thousands of games per bot 2010-11-30T12:02:05 imo cutoff is quite dangerous 2010-11-30T12:02:16 since there are such small differences in strength 2010-11-30T12:02:21 ye 2010-11-30T12:02:23 ep 2010-11-30T12:02:25 sure, you still wouldn't be 100% certain, but it would certainly be better right 2010-11-30T12:02:35 although, atm i think the top 10 is sorting itself out reasonably well 2010-11-30T12:02:57 btw, I made a post more or less related to this just now in this thread 2010-11-30T12:02:58 http://ai-contest.com/forum/viewtopic.php?f=19&t=1192&view=unread#unread 2010-11-30T12:03:02 4-7 is undecided 2010-11-30T12:04:30 janzert: any chance to run only a single game at a time? 2010-11-30T12:04:30 i am personally predicting top 3 stays as is 2010-11-30T12:04:41 I think narrowing cutoff should be done really really carefully now :S 2010-11-30T12:04:45 mega1: can't fairly change that now 2010-11-30T12:04:47 next contest 2010-11-30T12:05:04 mega1: not for this contest no 2010-11-30T12:05:12 okay 2010-11-30T12:05:58 janzert: any idea yet on when the contest will end? 2010-11-30T12:06:07 (exactly end) 2010-11-30T12:06:15 I've never been really happy with the multiple games myself so one per server is my recommendation for the next one though 2010-11-30T12:06:16 *** mightybyte has joined #aichallenge 2010-11-30T12:06:36 No, haven't seen Jeff yet today 2010-11-30T12:06:45 ok 2010-11-30T12:07:04 I do believe it will probably be early evening (eastern time) though 2010-11-30T12:07:16 but that's as specific as I know right now 2010-11-30T12:09:33 *** nn has quit IRC (Quit: Page closed) 2010-11-30T12:09:43 ok 2010-11-30T12:10:04 janzert: i'm not very good with time zones, how many hour left? 2010-11-30T12:10:25 about 30 I think 2010-11-30T12:10:37 @predict 2010-11-30T12:10:37 I predict that politik will win! 2010-11-30T12:10:45 thanks 2010-11-30T12:10:58 politics ... win? 2010-11-30T12:11:03 yes, somewhere around 30 2010-11-30T12:11:14 hours left that is 2010-11-30T12:11:23 that's quite a lot still 2010-11-30T12:11:52 Mistmanov: 30 hrs for you left to go up 4 places :P 2010-11-30T12:11:58 or 5 2010-11-30T12:12:13 heh i was hoping all of us could relax a little sooner than in 30 hours ;-) 2010-11-30T12:12:55 I actually think it's not very much :p 2010-11-30T12:12:56 i don't think anyone is going to shift too much at this point 2010-11-30T12:13:00 I use my Gnome calender 'Evolution' for the time zone conversion including daylight saving time. It works well with the Vancouver time zone. Hopefully that was the right one ^^ 2010-11-30T12:13:06 I've played like.. 5 games total against top50 bots? 2010-11-30T12:13:13 Should count them 2010-11-30T12:13:36 that's because you were stalling in the 200 area quite some time 2010-11-30T12:13:47 (ok, so maybe I don't desrve to be there.. but I was there like the day before the contest closed :( ) 2010-11-30T12:13:57 Mistmanov: i don't mind if we run it for another week ;) i just have to stop checking my bot's status during the day when it normally sucks 2010-11-30T12:14:10 Mistmanov: rankings then weren't very representative of bots abilities :) 2010-11-30T12:14:16 no offense :) 2010-11-30T12:14:17 yeah I know :p 2010-11-30T12:14:27 my smileys broke 2010-11-30T12:14:50 ugh, I'm turning into a grumpy old man 2010-11-30T12:14:54 *** amstan has joined #aichallenge 2010-11-30T12:14:54 *** ChanServ sets mode: +o amstan 2010-11-30T12:15:01 you say that like it's a bad thing 2010-11-30T12:15:01 that's due to a lot of people submitting strong bots at the last minute... antimatroid, mcleopold, and many others ;-) 2010-11-30T12:15:17 that's how we roll 2010-11-30T12:15:18 yea 2010-11-30T12:15:21 curse on you all 2010-11-30T12:15:48 i want to see someone submit an unknown winning bot in the last few days 2010-11-30T12:15:49 i'm facing a concrete wall reinforced with steel atm 2010-11-30T12:15:51 that'd be cool 2010-11-30T12:16:01 I know that if someone else was whining about how he should be ranked higher, I'd find it extremely annoying <<--- I'm probably being really annoying myself -_- 2010-11-30T12:16:13 Mistmanov: not at all 2010-11-30T12:16:15 * Mistmanov tries to shut up 2010-11-30T12:16:20 we all find things to bitch about :) 2010-11-30T12:16:26 lol, idd 2010-11-30T12:16:26 You could create a new account with the same email address and upload your bot. 2010-11-30T12:16:39 second mapset is horror for my bot chiting bitches 2010-11-30T12:16:57 antimatroid: that would look like an unknown competitor, but is really you in disguise 2010-11-30T12:17:00 * Error323 learned a new word here 2010-11-30T12:17:25 what does chiting mean? isn't that one of accouns inventions? 2010-11-30T12:17:32 Frontier: you'd get disqualified 2010-11-30T12:17:54 Frontier: combination of 2 words, i'll let you figure it out, lol 2010-11-30T12:18:06 :P 2010-11-30T12:18:42 Error323: I know from a reliable source that a new account created with the same email address just deactivates the old account 2010-11-30T12:18:56 oh 2010-11-30T12:18:59 Frontier: that's because the admins were nice 2010-11-30T12:19:05 it was done manuelly 2010-11-30T12:19:27 at least i'm pretty sure that'd how it went down 2010-11-30T12:19:37 that's** 2010-11-30T12:19:44 if this restriction wasn't met, i think the amount of bots would be tripled :P 2010-11-30T12:19:54 atleast 2010-11-30T12:20:01 i don't 2010-11-30T12:20:05 most people saw the rules 2010-11-30T12:20:16 without that rule i mean 2010-11-30T12:20:17 or maybe the database just overwrote the old entry for that email address and now it is totally messed up with submissions belonging to a NULL entry... 2010-11-30T12:20:26 ahhh, yeah not a fair contest then 2010-11-30T12:20:44 you would be uplaoding thousands of bots with a ranking system to try and make one do better 2010-11-30T12:21:17 the duplicate accounts were manually deactivated 2010-11-30T12:21:52 Yeah one bot could always yell "Itsa me Mario" and the other bot, Luigi will lose intenttionally so Mario gets the ELO benefit. 2010-11-30T12:22:14 yep, although you'd wait till well into the game before signalling each other 2010-11-30T12:22:19 janzert: do you think its possible for the next contest to have some i/o space? so stuff like reinforcement learning may be possible 2010-11-30T12:22:20 *** Migi32 has joined #aichallenge 2010-11-30T12:22:24 the start is too important to ensure you can win if it's not a friend 2010-11-30T12:22:38 Error323: amstan made a post about that 2010-11-30T12:22:40 janzert: That's nice. I have to tell that person that he caused some real work on the admin side. 2010-11-30T12:22:51 *** Migi32 has left #aichallenge 2010-11-30T12:22:51 Error323: I think it would be great, I'm not the one that will be coding it though so... 2010-11-30T12:23:01 Frontier:get the message across with a whack to the head :) 2010-11-30T12:23:07 bummer, I'm up to 4 turn 1 crashes today now 2010-11-30T12:23:15 well.. i'm not sure about storage space, there's not really any benefit to using it 2010-11-30T12:23:25 i have had an increase in crashes today too 2010-11-30T12:23:31 amstan: imo there is 2010-11-30T12:23:33 janzert: have there been a spike in crashes? 2010-11-30T12:23:37 the only way to write that io space is from games.. but the devs have logs of all the games already 2010-11-30T12:23:39 amstan: sure there is, map information and such 2010-11-30T12:23:41 he can just fill tehm offline 2010-11-30T12:23:45 learning bots is more awesome 2010-11-30T12:23:54 Error323: you can already do this.. 2010-11-30T12:24:00 you just have to reupload 2010-11-30T12:24:03 janzert: you can already read from files yeah? 2010-11-30T12:24:06 with new maps generated after final submission there isn't anyway to add the precalculated data 2010-11-30T12:24:15 janzert: i guess 2010-11-30T12:24:17 antimatroid: yes, you can 2010-11-30T12:24:32 online learning is still not possible 2010-11-30T12:24:36 hmmm, i like people not knowing a map before starting 2010-11-30T12:24:38 which is quite important 2010-11-30T12:24:42 antimatroid: no, not that I'm seeing 2010-11-30T12:24:48 i guess you could do some map caching 2010-11-30T12:24:49 i wanted to go to a new map for every game 2010-11-30T12:25:01 but if we have i/o we'll have to give devs access to that space.. 2010-11-30T12:25:08 the problem with online learning is it fucks up rankings 2010-11-30T12:25:15 and if there's a security problem or something and they cat /etc/passwd then they'll have access to it 2010-11-30T12:25:20 the bot one person plays at the start of the contest is not the bot someone else plays at the end 2010-11-30T12:25:21 amstan: yes, it's not a simple thing to do for sure 2010-11-30T12:25:28 you can train like that before the final contest 2010-11-30T12:25:32 what's wrong with doing it then? 2010-11-30T12:25:32 right now, the only want to extract info from your bot is to encode it in your games 2010-11-30T12:26:02 maybe allow the bots to do FTP I/O to the dev's own server? 2010-11-30T12:26:08 antimatroid: oh i don't know... a mapchange 20 days before deadline perhaps :P 2010-11-30T12:26:09 could be abused though 2010-11-30T12:26:20 Error323: maps were never set 2010-11-30T12:26:28 I do think we really should give players more access to the bot output just so debugging can be easier. 2010-11-30T12:26:41 antimatroid: not the point :P 2010-11-30T12:26:54 but yes any additional access to feedback from the bots makes any security problems worse 2010-11-30T12:26:58 is there a way to see how many of my games I've crashed out on turn 1 without watching them all? 2010-11-30T12:26:59 changes like that require online learning 2010-11-30T12:27:15 amstan: there are no passwords in "cat /etc/passwd", they hopefully are in /etc/shadow 2010-11-30T12:27:15 *** bhasker has joined #aichallenge 2010-11-30T12:27:25 Frontier: w/e, you know what i mean 2010-11-30T12:27:28 you see the problem with a bot that learns during the contest though right? 2010-11-30T12:27:37 i think that's a pretty big deal breaker 2010-11-30T12:28:04 my bot has not been having a great day today:-( 2010-11-30T12:28:17 people can use the new official tcp server to train their bot leading up to the contest 2010-11-30T12:28:34 Hey amstan, I saw your forum post about persistence. 2010-11-30T12:28:57 I think you could avoid the need to allow persistence by changing the engine protocol. 2010-11-30T12:29:42 All you have to do is play multiple games in one execution of the bot. 2010-11-30T12:30:18 oh mega1 released his sourcecode 2010-11-30T12:30:24 yeep 2010-11-30T12:30:25 omg omg omg 2010-11-30T12:30:43 With symmetric maps and near-optimal play, planetwars essentially becomes a ro-sham-bo-type problem. 2010-11-30T12:31:17 oh freaking hell how the hell does one browse a git repository 2010-11-30T12:31:23 there are no file names nothing 2010-11-30T12:31:25 :-\ 2010-11-30T12:31:29 yeah its rlly obnoxious imo 2010-11-30T12:31:31 have to check it out? 2010-11-30T12:31:34 bhasker: Click "tree"? 2010-11-30T12:31:45 just clone it 2010-11-30T12:31:46 xD 2010-11-30T12:31:49 there is none http://quotenil.com/git/planet-wars.git/ 2010-11-30T12:32:04 clone http://quotenil.com/git/planet-wars.git/ 2010-11-30T12:32:06 mega1's code is very well commented so relatively ez to see what it's doing without any lisp knowledge 2010-11-30T12:32:07 its like inside objects with stupid SHA256 names 2010-11-30T12:32:16 bhasker: That's not the browsable URL. 2010-11-30T12:32:31 oh doh 2010-11-30T12:32:36 he has a separate gitweb link 2010-11-30T12:32:37 doh 2010-11-30T12:32:42 rebelxt: I reorganized it a bit to make it readable top-down 2010-11-30T12:35:18 yep saw that, cool. lisp still hurts my brains a little bit due to inexperience 2010-11-30T12:35:35 my git says: fatal: Unable to find remote helper for 'http' 2010-11-30T12:35:35 :( 2010-11-30T12:35:55 same here ye 2010-11-30T12:35:56 Accoun vs bocsimacko 0 - 5 2010-11-30T12:36:16 mega1: do i need webdav support compiled in? 2010-11-30T12:36:32 git clone http://quotenil.com/git/planet-wars.git 2010-11-30T12:36:41 should work 2010-11-30T12:36:42 k i don't think i can understand this code without learning a bit of lisp 2010-11-30T12:36:45 what Error323 said. works fine 2010-11-30T12:36:50 though comments help 2010-11-30T12:38:32 *** amstan has quit IRC (Ping timeout: 240 seconds) 2010-11-30T12:38:43 looks like i have to recompile git... i can only fetch git://... urls currently :) 2010-11-30T12:38:51 *** amstan has joined #aichallenge 2010-11-30T12:38:51 *** ChanServ sets mode: +o amstan 2010-11-30T12:39:18 Frontier: is that gentoo? 2010-11-30T12:39:52 mega1: Yes, it is and now it works. curl use-Flag was missing. 2010-11-30T12:40:12 * mega1 is a mentalist 2010-11-30T12:41:22 so you didn't win but through reading the other contestants minds and strategies. 2010-11-30T12:41:49 lol 2010-11-30T12:43:16 bhasker: local tests match current official results pretty well so far 2010-11-30T12:43:25 yea 2010-11-30T12:43:31 our bots are dancing around each other 2010-11-30T12:43:44 ;-) 2010-11-30T12:43:52 1 elo difference again 2010-11-30T12:44:23 my bot is a night creature. it performs well at night and then loses during the day ;-) 2010-11-30T12:44:27 the games are getting harder though all my matchups now seem to be against bots in the top 50 2010-11-30T12:45:00 *** Vod has joined #aichallenge 2010-11-30T12:45:00 same here, but i managed to lose to someone at 200.. 2010-11-30T12:45:03 rebelxt> u dont have windows to test my bot 2010-11-30T12:45:22 mega1: so.. you have a blog? 2010-11-30T12:45:40 http://quotenil.com, no recent posts though 2010-11-30T12:45:41 Accoun: no, virtual machine is not fast enough.. 2010-11-30T12:46:39 Next cutoff change is now scheduled, start in 15 minutes and go to rank 1000 over the next 12 hours 2010-11-30T12:46:47 is mcleopolds tcp server still up? 2010-11-30T12:47:13 nice 2010-11-30T12:47:21 still up - no players though 2010-11-30T12:47:26 iouris blog post is almost similar to all the stuff i went through 2010-11-30T12:47:26 * janzert is afk again for a while 2010-11-30T12:47:31 mega1: I need someone to convert lisp code into c++ or pascal for me ^^ 2010-11-30T12:47:40 he seems to have run into the same issues almost 2010-11-30T12:47:45 when i tried to do the reinforcemnt 2010-11-30T12:47:48 i had the same problem 2010-11-30T12:47:51 with ships flying back and forth 2010-11-30T12:47:58 all the time 2010-11-30T12:48:24 *** Vod has quit IRC (Client Quit) 2010-11-30T12:48:33 Top 10 players: bocsimacko(3879), _iouri_(3680), jimrogerz(3649), GreenTea(3638), _Astek_(3631), Accoun(3627), Slin-.-(3626), protocolocon(3613), asavis(3612), george(3598) 2010-11-30T12:48:37 Frontier: I'll blog about it, but the comments should be quite informative until then. 2010-11-30T12:48:40 bhasker: same here 2010-11-30T12:49:03 such a huge distance between the top two bots... 2010-11-30T12:49:11 i took it out i just decided to add a rate of flow parameter into the mix 2010-11-30T12:49:13 at end of turn 2010-11-30T12:49:19 i calculate how many ships are incoming to my planets 2010-11-30T12:49:26 and do a crude rate of flow calculation 2010-11-30T12:49:32 and try to send ships preemptively 2010-11-30T12:49:35 to my planets 2010-11-30T12:49:49 though that tends to kind of mess up with large attacks 2010-11-30T12:50:30 mega1: "The official server is running the ancient sbcl 1.0.11 [...]" ...so you had the same troubles as the C# guys 2010-11-30T12:50:39 mega1: pretty cool 2010-11-30T12:51:13 Frontier: what is the exact trouble the C# guys are having? old mono? 2010-11-30T12:51:42 yeah, congratulations! - yes the old mono version that wouldn't compile their code. 2010-11-30T12:52:29 rebel do you think our bots will climb much further? i really have my doubts 2010-11-30T12:52:34 i think we are stuck in 45-55 range 2010-11-30T12:52:36 i like how accoun demands that we just manually change the rankings so he's second 2010-11-30T12:52:36 for good:) 2010-11-30T12:53:43 bhasker: yeah, i don't see a reason why my bot should climb. it's roughly where it deserves to be ;) 2010-11-30T12:53:44 mega1: You might want to take a look at http://clojure.org/. A lisp dialect compiling to Java Byte-Code. Perhaps it is worth a look, I don't know 2010-11-30T12:53:52 yea same here 2010-11-30T12:54:15 i never understood the need to compile everything to run on jvm 2010-11-30T12:54:20 jvm is so freaking slow 2010-11-30T12:55:06 bhasker: cross platform? 2010-11-30T12:55:26 frankly unless its a client app cross platform is a very weak argument 2010-11-30T12:55:47 in most cases moving platforms is such a high barrier that language is a very small part of it 2010-11-30T12:56:29 and if you are writing in c# why compile it to jvm or lisp to jvm 2010-11-30T12:56:36 c# in most cases you have mono 2010-11-30T12:56:45 lisp has its own implementations on most platforms 2010-11-30T12:56:58 so where is the benefit in compiling down to jvm unless you want to use some specific java feature? 2010-11-30T12:58:13 bhasker: one argument in favor of java is amount of open source code. it's pretty easy to build web sites and business software with it using 100% open source 2010-11-30T12:58:34 i am not so sure, java as a middle tier sure 2010-11-30T12:58:47 *** Apophis has joined #aichallenge 2010-11-30T12:58:48 i hate websites written in JSP 2010-11-30T12:58:53 the are routinely slow as hell 2010-11-30T12:59:18 and i don't know of any major web company that runs stuff written in jsp etc 2010-11-30T12:59:40 google does most of its stuff in python, yahoo in php + c/c++ modules 2010-11-30T12:59:40 bhasker: Java is actually quite fast. 2010-11-30T12:59:51 JSP was bad.. i stopped using it many years ago 2010-11-30T13:00:12 mighty: my experience has not been the same, jvm is such a memory hog 2010-11-30T13:00:36 java has other advantages, but i don't think execution speed is necessarily one of them 2010-11-30T13:00:41 the start up time of jvm is horrible 2010-11-30T13:00:44 bhasker: I've had a Java matrix math library perform faster than the C++ library it was ported from. 2010-11-30T13:00:56 the c++ library was probably poorly written 2010-11-30T13:01:02 start up time is frequently not an issue 2010-11-30T13:01:13 bhasker: It was a direct translation 2010-11-30T13:01:47 i really doubt java interpreted byte code can beat a similarly implemented c++ code unless it was JIT compiled 2010-11-30T13:01:59 even then it will probably be a close match up 2010-11-30T13:02:18 and it will depend on what kind of optimizations are turned on in c++ build 2010-11-30T13:02:18 Well, I have numbers that prove otherwise. 2010-11-30T13:02:34 here is a programming language performance comparison which matched my own tests a couple of years ago (c++ vs java vs python) 2010-11-30T13:02:36 http://shootout.alioth.debian.org/u32/which-programming-languages-are-fastest.php 2010-11-30T13:03:48 It's not always faster, but in some cases it is, and it's usually at least competitive. 2010-11-30T13:05:49 mightybyte: kinda surprised java beat c++ on math problem. c++ normally beats java -server unless c++ code is not optimized at all 2010-11-30T13:06:08 I was surprised too. 2010-11-30T13:06:12 also java doesn't make great use of multicore processors 2010-11-30T13:06:24 bhasker: And C++ does? 2010-11-30T13:06:25 its kind of retarded that so many jvms cant make good use of multicore 2010-11-30T13:06:25 LISP SBCL is pretty fast 2010-11-30T13:06:34 yes c++ does 2010-11-30T13:06:43 fork 4tw 2010-11-30T13:06:46 or native threads 2010-11-30T13:06:49 bhasker: Oh please 2010-11-30T13:07:07 err loading up 8jvm instances to make use of 8 processors seems kind of wasteful? 2010-11-30T13:07:14 Java's threading is vastly nicer than C++ fork 2010-11-30T13:07:25 yes but the jvm itself is tied to a single core 2010-11-30T13:07:34 so threads != multicore usage 2010-11-30T13:07:55 bhasker = ignorant 2010-11-30T13:08:00 from what i can read java7 probably has some new stuff to make multicore usage 2010-11-30T13:08:51 please explain how the jvm you have on your box makes use of more than one core 2010-11-30T13:09:16 at least the jvm i have on my mac or bsd doesn't do it at all 2010-11-30T13:09:27 i haven't looked at the linux jvms very closely 2010-11-30T13:09:33 so i may be wrong 2010-11-30T13:10:48 rebelext and mightybyte: when I graph the languages like that: http://shootout.alioth.debian.org/u32/which-programming-languages-are-fastest.php?gpp=on&java=on&fpascal=on&sbcl=on&python3=on&calc=chart C++ has the smalles time share by far 2010-11-30T13:11:32 fork is a unix system call. it can be done with other techniques by now, like posix threads ;) 2010-11-30T13:12:03 as i said native threads:) which are usually posix api based 2010-11-30T13:12:08 fork has its benefits 2010-11-30T13:12:24 especially in server environments where segregating user data is criticial 2010-11-30T13:12:41 Frontier: well, i look at the raw numbers (median) 2010-11-30T13:12:46 chances of stomping over wrong users data due to programming errors are less 2010-11-30T13:13:01 I'm not arguing that C++ doesn't give you the power to do effective low-level optimizations that can't be done on the JVM. I'm saying that I took a reasonably written (but not super-optimized) C++ matrix library, did a direct translation to Java, and got comparable (and sometimes slightly) better performance from Java. 2010-11-30T13:13:27 mighty its possible, in most cases though if you turn on the right optimizations c++ will usually beat java 2010-11-30T13:13:46 unless you use a JIT compiler 2010-11-30T13:13:49 rebelext: smaller numbers are better :p 2010-11-30T13:13:56 in which case java will be close to c++ 2010-11-30T13:14:29 if you start two threads in java on a dual core cpu, they make use of both cores just fine. of course when you sort a collection it only uses single core.. java 7 should have fork-join support i heard 2010-11-30T13:14:50 Frontier: that's not what she said 2010-11-30T13:14:50 *** nullkuhl has joined #aichallenge 2010-11-30T13:16:07 ok, then maybe she said, that the top of the list is the winner like in planetwars? 2010-11-30T13:17:08 breaking news: i lose to bocsimacko ;) 2010-11-30T13:17:15 mightybyte: what classes did you use for the matrix in C++ and Java? 2010-11-30T13:17:33 We wrote our own. 2010-11-30T13:17:40 That's precisely my point. 2010-11-30T13:19:12 i should open source my c++ bot too.. just so that people can do c++ vs python performance comparo if interested. it was not pretty ;-) 2010-11-30T13:19:36 *** Naktibalda has joined #aichallenge 2010-11-30T13:19:50 tcp 2010-11-30T13:20:01 @tcp 2010-11-30T13:20:02 rebelxt: any chance you tried you python bot on the new pypy release? 2010-11-30T13:20:02 Naktibalda: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) 98.247.248.39:995 (zeroviz.us), or (#4) http://tcp.zeroviz.us/, or (#5) http://tcp2.zeroviz.us/ (crazy!) 2010-11-30T13:20:05 how does it work? 2010-11-30T13:20:18 I'm curious what sort of speed up it might give 2010-11-30T13:20:40 janzert: i built latest pypy and tried but could not get it to work 2010-11-30T13:20:46 ahh :( 2010-11-30T13:22:01 translation ran for almost an hour and failed hard.. i just had no time to figure out what exactly was going on 2010-11-30T13:22:02 mightybyte: Yes, but did you use vector> or did you use float or long double? And what did you use in Java? fixed arrays or lists? 2010-11-30T13:22:48 I know last time I tried it (although it's been a few years) it took huge amounts of memory for translation to complete 2010-11-30T13:23:24 Frontier: I used fixed arrays. It wasn't a best of breed comparison; it was a comparison of exactly analogous programs. 2010-11-30T13:23:27 *** andrei92 has joined #aichallenge 2010-11-30T13:23:40 *** Accoun has quit IRC () 2010-11-30T13:24:05 that is strange in deed 2010-11-30T13:24:16 interesting first time i am reading about pypy 2010-11-30T13:24:17 In most cases, C++ was slightly faster. 2010-11-30T13:24:18 hmm 2010-11-30T13:24:40 But there was at least one case where Java was faster. I was quite surprised initially. 2010-11-30T13:25:10 I assumed we had done something wrong, and maybe the computations were getting completetly optimized away, but that was not the case. 2010-11-30T13:25:30 some perf stats on pypy: http://morepypy.blogspot.com 2010-11-30T13:26:27 interesting let me test my bot 2010-11-30T13:27:15 hmm rebel you wanna start up your bot on zeroviz? 2010-11-30T13:28:01 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]) 2010-11-30T13:29:49 *** yasith_ has quit IRC (Ping timeout: 265 seconds) 2010-11-30T13:30:17 ahhh , this is annoying i'm loosing all games because ' issued an invalid order.. ' every time and don;t know why , ahhh 2010-11-30T13:30:23 bhasker: sure. give me a minute 2010-11-30T13:31:18 my bot deserve a higher rank ... http://ai-contest.com/profile.php?user_id=13335 2010-11-30T13:32:23 *** andrei92 has quit IRC (Quit: Page closed) 2010-11-30T13:33:27 bhasker: running 2010-11-30T13:35:48 hmm i can't get pypy to run my bot 2010-11-30T13:35:49 weird 2010-11-30T13:36:53 bbl 2010-11-30T13:36:58 *** rebelxt has left #aichallenge 2010-11-30T13:37:21 *** mightybyte has left #aichallenge 2010-11-30T13:37:28 *** Accoun has joined #aichallenge 2010-11-30T13:38:19 k got it to run 2010-11-30T13:38:24 lets see how much faster is pypy 2010-11-30T13:38:52 hmm uses a lot more cpu 2010-11-30T13:40:38 hmm 300ms per turn will ahve to run with python to compare 2010-11-30T13:40:41 or so 2010-11-30T13:42:37 hah interesting game 2010-11-30T13:42:38 http://www.ai-contest.com/visualizer.php?game_id=9383214 2010-11-30T13:42:56 *** yasith_ has joined #aichallenge 2010-11-30T13:43:26 weird that my bot won that though 2010-11-30T13:44:01 why> 2010-11-30T13:45:25 i am not sure why it decided to attack 2010-11-30T13:45:31 we were even 2010-11-30T13:45:40 normally my bot should have just sat doing nothing 2010-11-30T13:46:00 *** GreenTea_ has joined #aichallenge 2010-11-30T13:46:04 oh wait 2010-11-30T13:46:06 i was 1 turn ahead 2010-11-30T13:46:09 yeah 2010-11-30T13:46:10 so i had more ships than him 2010-11-30T13:46:12 ah 2010-11-30T13:46:14 makes sense now 2010-11-30T13:46:18 and he had more ships in the bottom right planet 2010-11-30T13:46:46 yea 2010-11-30T13:46:51 he sent like 16 away 2010-11-30T13:46:53 to capture that one 2010-11-30T13:46:55 weird 2010-11-30T13:47:05 and 8 to capture the 2 2010-11-30T13:47:06 at the top 2010-11-30T13:47:15 i guess he estimated how many he would need to prevent a snipe 2010-11-30T13:48:33 Top 10 players: bocsimacko(3864), _iouri_(3671), jimrogerz(3635), GreenTea(3631), _Astek_(3624), Accoun(3618), protocolocon(3614), Slin-.-(3612), asavis(3595), george(3593) 2010-11-30T13:48:46 aww mcleopolds server is not updating ratings anymore 2010-11-30T13:48:49 or the games 2010-11-30T13:48:55 so can't see the games against rebel on the visualizer:-\ 2010-11-30T13:50:57 *** amstan has quit IRC (Remote host closed the connection) 2010-11-30T13:51:24 *** amstan has joined #aichallenge 2010-11-30T13:51:24 *** ChanServ sets mode: +o amstan 2010-11-30T14:02:17 *** dmj111` has joined #aichallenge 2010-11-30T14:02:22 *** dmj111 has quit IRC (Write error: Broken pipe) 2010-11-30T14:03:34 hard battle http://ai-contest.com/visualizer.php?game_id=9390849 2010-11-30T14:09:58 *** exaide has quit IRC (Quit: Page closed) 2010-11-30T14:14:09 *** Utkarsh has joined #aichallenge 2010-11-30T14:17:17 medri around? 2010-11-30T14:21:23 *** jaspervdj has quit IRC (Quit: NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN) 2010-11-30T14:24:08 bhasker: yep 2010-11-30T14:24:14 *** medrimonia1 is now known as medrimonia 2010-11-30T14:26:15 bhasker: I see that you've mean my terribly bad attack =P 2010-11-30T14:26:19 meet* 2010-11-30T14:31:53 *** rebelxt has joined #aichallenge 2010-11-30T14:33:22 bhasker: yeah, you have to manually iterate over game ids to watch games :( here is the last game we've just played http://zeroviz.us:8080/canvas.cgi?game_id=35669 2010-11-30T14:37:04 :d 2010-11-30T14:37:15 that's a lot of ppl on the forums 2010-11-30T14:38:06 ah quite a few interesting games rebel 2010-11-30T14:38:50 medri i like the way my bot did a double attack on both sides 2010-11-30T14:38:56 quite interesting to watch 2010-11-30T14:41:36 *** Mistmanov has quit IRC (Ping timeout: 265 seconds) 2010-11-30T14:41:41 *** Mistmanov has joined #aichallenge 2010-11-30T14:44:25 Grmbl I hate c 2010-11-30T14:45:00 bhasker: are you running modified version or final submission? 2010-11-30T14:47:26 bhasker: we are having more exciting games on zeroviz than on official server, that's for sure 2010-11-30T14:47:45 side swaps ftw 2010-11-30T14:48:28 i really should not be scheduling fleets 20+ turns into the future ;-) too much uncertainty 2010-11-30T14:48:33 Top 10 players: bocsimacko(3859), _iouri_(3665), jimrogerz(3633), GreenTea(3624), _Astek_(3622), Slin-.-(3613), protocolocon(3610), Accoun(3610), george(3593), asavis(3592) 2010-11-30T14:51:55 *** McLeopold has joined #aichallenge 2010-11-30T14:52:32 same version as submission 2010-11-30T14:52:35 no change 2010-11-30T14:56:53 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-11-30T15:00:44 *** bss03-AFK is now known as bss03 2010-11-30T15:09:19 *** tapwater has joined #aichallenge 2010-11-30T15:09:25 *** amstan has quit IRC (Ping timeout: 245 seconds) 2010-11-30T15:09:44 *** bhasker has joined #aichallenge 2010-11-30T15:15:01 *** mega1 has quit IRC (Read error: Operation timed out) 2010-11-30T15:15:58 *** perror has quit IRC (Quit: Bye all !) 2010-11-30T15:18:04 very quiet in here 2010-11-30T15:20:21 Pairing cutoff going down.. 2010-11-30T15:21:12 more games for us 2010-11-30T15:21:24 hmm though games per minue is dropping:-\ 2010-11-30T15:21:27 games taking longeR? 2010-11-30T15:21:31 yeah 2010-11-30T15:21:40 all the low bots take like 1 ms per turn 2010-11-30T15:21:49 (I hope so at least :p) 2010-11-30T15:22:02 heh 2010-11-30T15:22:15 oh you and e323 are now two ranks apart 2010-11-30T15:22:16 and per second too.. 2010-11-30T15:22:59 * games per second 2010-11-30T15:23:46 * per minute %) 2010-11-30T15:24:00 Git is so cool 2010-11-30T15:25:10 I'd never even heard of it before... turns out my cygwin installation already had it installed so downloading bocsi's bot was a breeze 2010-11-30T15:25:22 (just downloading a .zip would've been easier.. but eh..) 2010-11-30T15:25:26 *** sigh has joined #aichallenge 2010-11-30T15:25:26 i am still trying to wrap my head around git concepts 2010-11-30T15:25:31 have the pdf that e323 sent open 2010-11-30T15:25:36 but been lazy to read it 2010-11-30T15:27:40 almost had a hard attack there -_- 2010-11-30T15:27:49 was afraid an unknown dutch guy took the #2 spo 2010-11-30T15:27:51 spot 2010-11-30T15:28:05 turns out it was the lisp-specific ranking.... 2010-11-30T15:29:10 *** yasith_ has quit IRC (Ping timeout: 250 seconds) 2010-11-30T15:31:48 *** bss03 is now known as bss03-AFK 2010-11-30T15:32:18 heh 2010-11-30T15:34:03 *** rhinobot has quit IRC (Quit: Page closed) 2010-11-30T15:38:35 #7 bot in python 2010-11-30T15:38:45 well 6 or 7 2010-11-30T15:38:48 since rebel and me are tied 2010-11-30T15:38:49 right now 2010-11-30T15:42:29 *** yasith_ has joined #aichallenge 2010-11-30T15:46:51 *** meneky has quit IRC (Ping timeout: 265 seconds) 2010-11-30T15:48:33 Top 10 players: bocsimacko(3857), _iouri_(3662), jimrogerz(3630), _Astek_(3616), GreenTea(3614), Slin-.-(3605), protocolocon(3604), Accoun(3601), asavis(3586), george(3586) 2010-11-30T15:52:31 *** p4p4 has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2010-11-30T15:53:01 *** sigh has quit IRC (Remote host closed the connection) 2010-11-30T15:57:18 *** Frontier has quit IRC (Ping timeout: 276 seconds) 2010-11-30T15:57:46 *** antimatroid has quit IRC (Ping timeout: 240 seconds) 2010-11-30T16:03:29 Mistmanov: lol 2010-11-30T16:03:33 aerique? 2010-11-30T16:03:41 yeah 2010-11-30T16:04:34 heh 2010-11-30T16:04:50 Mistmanov: lose moar =) 2010-11-30T16:04:58 -_- 2010-11-30T16:05:48 holly shit, wat goin on! 2010-11-30T16:06:12 GreenTea_: ? 2010-11-30T16:06:39 7 looses last 9 games! 2010-11-30T16:08:08 http://ai-contest.com/visualizer.php?game_id=9406501 2010-11-30T16:08:12 bah i hate these losses 2010-11-30T16:11:53 lol 2010-11-30T16:12:37 *** yasith_ has quit IRC (Ping timeout: 264 seconds) 2010-11-30T16:12:48 Error323: the ones where you get owned? ;) 2010-11-30T16:12:58 heh 2010-11-30T16:12:58 by a few ships :P 2010-11-30T16:12:59 ye 2010-11-30T16:13:16 14 ships, owned 2010-11-30T16:13:31 if (turn > 190 && hasntdonecrapfor100turns) { attacklikeamadman(); } 2010-11-30T16:13:47 yeah i admit, it was a really fancy way of capping neutrals by dmj 2010-11-30T16:13:56 but still :P hate losing ye ^_^ 2010-11-30T16:15:04 could you have taken that 64 neutral in the back? 2010-11-30T16:15:11 without being overrun? 2010-11-30T16:15:57 mmm perhaps 2010-11-30T16:17:33 if i'm still in the 80-90 by tomorrow morning i don't think i'll go up anymore at all 2010-11-30T16:17:38 I think that ought to be defined as a draw... this game has virtually no draws 2010-11-30T16:18:09 if you have equal growth and planets after 200 turns, draw 2010-11-30T16:21:04 *** rebelxt has left #aichallenge 2010-11-30T16:21:59 http://en.wikipedia.org/wiki/Line-plane_intersection 2010-11-30T16:22:02 *** antimatroid has joined #aichallenge 2010-11-30T16:22:41 err wrong window 2010-11-30T16:23:47 politik: I disagree 2010-11-30T16:23:48 *** bss03-AFK is now known as bss03 2010-11-30T16:23:54 *** bss03 is now known as bss03-AFK 2010-11-30T16:23:57 many maps simply aren't suited for attacking each other 2010-11-30T16:24:05 so it becomes a battle of who is the best at expanding 2010-11-30T16:24:49 *** Rubicon-|-Cross has joined #aichallenge 2010-11-30T16:25:08 *** yasith_ has joined #aichallenge 2010-11-30T16:27:42 *** narnach has joined #aichallenge 2010-11-30T16:27:59 Good evening everyone! 2010-11-30T16:38:34 http://zeroviz.us:8080/ not refreshing but games going 2010-11-30T16:41:34 Accoun: it's refreshing now 2010-11-30T16:41:43 also, use tcp.zeroviz.us 2010-11-30T16:43:55 McLeopold> setup 2nd final map set please 2010-11-30T16:44:00 8080 is official down 2010-11-30T16:44:09 Accoun: sure, give me a minute 2010-11-30T16:44:18 *** BaronTrozo has joined #aichallenge 2010-11-30T16:45:55 *** Naktibalda has joined #aichallenge 2010-11-30T16:46:02 Nobody writes, everybody is paying attenting to the ranking???? 2010-11-30T16:46:16 atteting->attetion 2010-11-30T16:46:31 attention, upppss sorry 2010-11-30T16:47:00 http://tcp.zeroviz.us/ 2010-11-30T16:48:34 Top 10 players: bocsimacko(3857), _iouri_(3655), jimrogerz(3623), _Astek_(3613), Slin-.-(3605), protocolocon(3602), GreenTea(3600), Accoun(3592), asavis(3585), george(3582) 2010-11-30T16:49:09 Hi Accoun, have you finally found what you where looking for in your youtube video??? 2010-11-30T16:49:51 *** meneky has joined #aichallenge 2010-11-30T16:50:16 Accoun: second set is on tcp.zeroviz.us. 2010-11-30T16:50:46 *** yasith_ has quit IRC (Ping timeout: 255 seconds) 2010-11-30T16:53:02 Привет Accoun, вы понимаете по-английски? 2010-11-30T16:54:44 *** amstan has joined #aichallenge 2010-11-30T16:54:44 *** ChanServ sets mode: +o amstan 2010-11-30T16:56:25 BaronTrozo> yes 2010-11-30T16:56:44 BaronTrozo> not found 2010-11-30T16:57:59 upss, continue searching I think the video is a good publicity 2010-11-30T17:00:31 BaronTrozo> it 2 years old 2010-11-30T17:01:13 * GreenTea preapare coard and soap.. 2010-11-30T17:01:37 * cord 2010-11-30T17:01:46 *** choas has joined #aichallenge 2010-11-30T17:02:00 GreenTea> why u cant win me on TCP ? 2010-11-30T17:02:44 coz I not test my last improvement on tcp 2010-11-30T17:02:57 with it i'l beet you ;P 2010-11-30T17:04:00 *** sigh has joined #aichallenge 2010-11-30T17:04:57 Accoun: One more question if this search was began two years ago, are you still looking for 16-18 or you have updated to 18-20??? 2010-11-30T17:05:21 GreenTea> you just got lucky couse that Canadians hate me 2010-11-30T17:05:36 *** yasith has joined #aichallenge 2010-11-30T17:05:50 *** mceier has quit IRC (Read error: Operation timed out) 2010-11-30T17:05:53 BaronTrozo> 19 and more to old 2010-11-30T17:06:03 BaronTrozo> u know russian? 2010-11-30T17:06:15 I have no its google transletor 2010-11-30T17:06:46 ups sorrry, No i don't know russian, it is from google transletor 2010-11-30T17:06:47 BaronTrozo> someone write it in text mode? 2010-11-30T17:07:00 yes google 2010-11-30T17:07:16 BaronTrozo> google? 2010-11-30T17:07:37 *** flag has joined #aichallenge 2010-11-30T17:07:37 BaronTrozo> gogle can translate sound? 2010-11-30T17:08:05 We have translate your video with google, also in 'rufes band' we have a russion guy that have translete the video 2010-11-30T17:09:18 *** mceier has joined #aichallenge 2010-11-30T17:09:23 Accoun: this is really you on the video? 2010-11-30T17:09:24 BaronTrozo> would be better if you had a girl from 16 to 18 years according to marry me 2010-11-30T17:09:34 GreenTea_> da 2010-11-30T17:09:39 )) 2010-11-30T17:09:55 We have shown this video to all femenine staff in our office, and they ask, how much rich are you?? 2010-11-30T17:10:39 BaronTrozo> enougth for have unworcked wawe 2010-11-30T17:11:08 BaronTrozo> from 14 to 18 need only 2010-11-30T17:11:34 BaronTrozo> from 16 to 18 need only 2010-11-30T17:11:58 Sorry in spain is only legal 18 and up 2010-11-30T17:12:09 *** Mistmanov has quit IRC (Quit: Page closed) 2010-11-30T17:12:21 BaronTrozo> 18 is good 2010-11-30T17:12:57 BaronTrozo> in russia 16 legal 2010-11-30T17:13:18 BaronTrozo> in ucraina aprox from 14 2010-11-30T17:15:25 OKey, it may be legal but not moral 2010-11-30T17:16:52 yeah guys... this is 1: offtopic, 2: possibly illegal 2010-11-30T17:16:59 so... please cease 2010-11-30T17:17:03 Lets change the conversation to planetwars 2010-11-30T17:17:27 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-11-30T17:17:27 BaronTrozo: so what company is it then? 2010-11-30T17:17:41 is it a big company? 2010-11-30T17:17:54 i'm talking about "rufes band" btw 2010-11-30T17:18:03 sorry i can't not say the company name for a lot of reasons 2010-11-30T17:18:17 aww, not even hints? 2010-11-30T17:18:22 BaronTrozo> stuped european slaves moral 2010-11-30T17:18:25 first some people of rufes band have been fired from this company 2010-11-30T17:18:34 wtf? 2010-11-30T17:18:40 why? for compeeting? 2010-11-30T17:18:55 we compete as a group of friends 2010-11-30T17:19:39 Protocolon started the contest and the rest of us think it was a nice way to spent our empty lifes 2010-11-30T17:20:04 *** medrimonia has quit IRC (Quit: Leaving.) 2010-11-30T17:20:05 :( 2010-11-30T17:20:29 mmm 2010-11-30T17:20:35 also probably no one outside spain has hear of this company 2010-11-30T17:21:05 *** medrimonia has joined #aichallenge 2010-11-30T17:21:25 http://www.google.ru/search?sourceid=chrome&ie=UTF-8&q=rufes+band 2010-11-30T17:22:16 We can do another contest, try to guess the company name where at some time instant all of us work for it 2010-11-30T17:22:23 meeen the tension is killing me, the freakin contest has to stop i'm still #1 of my country :P 2010-11-30T17:22:31 BaronTrozo: lol 2010-11-30T17:23:05 BaronTrozo: so.. what does the name mean? 2010-11-30T17:23:39 Rufes is a spanish second name, and it is very comon 2010-11-30T17:24:21 ivan band 2010-11-30T17:24:25 a previous coworked who was fired, just something like a memorial 2010-11-30T17:24:38 his second name 2010-11-30T17:24:42 seems messed up 2010-11-30T17:24:47 indeed 2010-11-30T17:25:01 planetwars again please 2010-11-30T17:26:07 amstan: is there an exact time when the contest stops yet? 2010-11-30T17:26:26 want to ask the same question) 2010-11-30T17:26:34 the time is wed 2010-11-30T17:26:47 where the difference between the closing time and results time = d 2010-11-30T17:26:51 and d%24hours=0 2010-11-30T17:26:57 lol 2010-11-30T17:27:01 so that's a no then :P 2010-11-30T17:27:16 that will probably it 2010-11-30T17:27:24 so... at the same time that the contest closed 2010-11-30T17:27:38 the Pairing cutoff is going to stop at some point?? 2010-11-30T17:28:26 it must, because otherwise results will be biased and wrong 2010-11-30T17:29:52 is this point public?? 2010-11-30T17:30:22 *** yasith has joined #aichallenge 2010-11-30T17:30:57 I dont understand why organizers can't told exact time.. 2010-11-30T17:32:09 amsten: why this so big secret? 2010-11-30T17:33:13 it sounds like he's saying it exactly the same time as the submissions were closed, but on wednesday 2010-11-30T17:33:30 so 23:59 CST wed 2010-11-30T17:33:34 ? 2010-11-30T17:33:59 or 00:01 %) 2010-11-30T17:34:25 last day with cutting on 100????? 2010-11-30T17:34:35 nooo 2010-11-30T17:36:20 *** nn has joined #aichallenge 2010-11-30T17:36:36 *** Epona has joined #aichallenge 2010-11-30T17:36:48 and silence... 2010-11-30T17:36:54 ... 2010-11-30T17:37:11 like that days when servers are not working >:-| 2010-11-30T17:37:17 hahaha 2010-11-30T17:38:37 i'm not sure exactly.. ask on the forums 2010-11-30T17:39:59 *** yasith has quit IRC (Ping timeout: 265 seconds) 2010-11-30T17:40:33 *** choas has quit IRC (Quit: leaving) 2010-11-30T17:41:25 *** Epona has quit IRC (Ping timeout: 264 seconds) 2010-11-30T17:41:29 but amstam, aren't you one of the anfitrions?? 2010-11-30T17:42:10 organizer*, yes 2010-11-30T17:42:12 but i am unsure 2010-11-30T17:42:35 ????????????? 2010-11-30T17:42:41 i don't understand 2010-11-30T17:43:08 the organization still don't know when this is going to be closed??? 2010-11-30T17:43:53 is a question, not a reclaim 2010-11-30T17:44:47 i already told you what i think 2010-11-30T17:44:51 I prefer that the contest takes longer as at the moment i'm climing up on the rank 2010-11-30T17:45:01 i don't have any more info, ask janzert or jeff 2010-11-30T17:45:23 the intent i believe is to let it run until the rankings are fairly stable 2010-11-30T17:45:24 okey, thanks for all 2010-11-30T17:46:01 Zannick: but a total stabilzed maybe imposible, no?? 2010-11-30T17:46:37 BaronTrozo: they'll probably stabilize fairly well when the only games being played are between the top 100, maybe 2010-11-30T17:47:34 after all maps be played it what be next? 2010-11-30T17:47:36 I also think the same, but is because my bot take better result with good bots(i don't know the reason) ;) 2010-11-30T17:47:39 it stoped? 2010-11-30T17:47:40 i think people will care more about difference between 1 and 2nd than 1000th and 1001st 2010-11-30T17:47:53 true 2010-11-30T17:48:02 that's the reason we have the cutoffs 2010-11-30T17:48:18 true -> (1==1) 2010-11-30T17:48:35 Top 10 players: bocsimacko(3856), _iouri_(3654), jimrogerz(3616), _Astek_(3608), Slin-.-(3602), protocolocon(3595), GreenTea(3593), Accoun(3591), asavis(3582), george(3580) 2010-11-30T17:48:42 hmm? 2010-11-30T17:50:24 does this actual decrease of cuttof has a limit??? 2010-11-30T17:50:38 cutoff, sorry 2010-11-30T17:51:04 define limit, give me an example 2010-11-30T17:52:16 last cutoff stopped on 3020 2010-11-30T17:52:21 *** yasith has joined #aichallenge 2010-11-30T17:52:27 he's asking if the cutoff will decrease all the way to 2 or stop somewhere higher, like 100 2010-11-30T17:52:31 is this decrease going to stop on some limit 2010-11-30T17:52:56 Very sorry for my poor english 2010-11-30T17:53:34 haha...it would be funny if the cutoff went all the way to 2, even more funny if that resulted in #2 being demoted 2010-11-30T17:53:52 and then #3 suddenly has another fighting chance 2010-11-30T17:54:13 yup 2010-11-30T17:54:49 *** xxx has joined #aichallenge 2010-11-30T17:54:50 2194 at this moment 2010-11-30T17:55:14 every 12.9 second one guy is out 2010-11-30T17:58:32 that's about 280 people every hour 2010-11-30T17:59:08 *** eburnette has joined #aichallenge 2010-11-30T17:59:18 Is it soup yet? 2010-11-30T17:59:27 I will be gone in about 8 hours, oh no 2010-11-30T17:59:33 soup? 2010-11-30T17:59:58 settled, done, complete 2010-11-30T18:00:02 @ranking 2010-11-30T18:00:13 eburnette: Error: There was a problem accessing the interface to ai-contest.com 2010-11-30T18:00:17 no, still a while yet, I think 2010-11-30T18:00:33 @rankings 2010-11-30T18:00:35 sigh: Top 10 players: bocsimacko(3857), _iouri_(3654), jimrogerz(3617), _Astek_(3606), Slin-.-(3603), protocolocon(3593), Accoun(3592), GreenTea(3592), asavis(3583), george(3578) 2010-11-30T18:00:56 @predict 2010-11-30T18:00:57 I predict that superflit will win! 2010-11-30T18:01:10 @rankings 29 2010-11-30T18:01:13 eburnette: Top 29 players: bocsimacko(3857), _iouri_(3654), jimrogerz(3617), _Astek_(3606), Slin-.-(3603), protocolocon(3593), Accoun(3592), GreenTea(3592), asavis(3583), george(3578), bix0r4ever(3558), dmj111(3549), Raschi(3544), davidjliu(3543), BaronTrozo(3539), smloh1(3521), Mangara(3520), medrimonia(3520), wagstaff(3516), shangas(3511), thedreamer(3508), Neverstu(3503), jambachili(3502), (1 more message) 2010-11-30T18:01:29 YEah I should win 2010-11-30T18:01:33 my bot is stealth 2010-11-30T18:01:37 :P 2010-11-30T18:01:46 eburnette: why 29? 2010-11-30T18:02:01 I was trying to get rid of the "1 more message" message 2010-11-30T18:02:21 amstan: the thing is the cutoff cannot keep on going 2010-11-30T18:02:25 @next 2010-11-30T18:02:25 sigh: Error: You must be registered to use this command. If you are already registered, you must either identify (using the identify command) or add a hostmask matching your current hostmask (using the "hostmask add" command). 2010-11-30T18:02:31 @more 2010-11-30T18:02:31 sigh: Error: You haven't asked me a command; perhaps you want to see someone else's more. To do so, call this command with that person's nick. 2010-11-30T18:02:35 there are people not ranked higher because they keep losing from lower ones aswell 2010-11-30T18:02:49 if you let this cutoff go on the entire top 100 will change 2010-11-30T18:03:04 yeah, it shouldn't go past 1000 or so, I think 2010-11-30T18:03:19 just here for a second but the current cutoff change goes to rank 1000 at midnight server time 2010-11-30T18:03:27 you might aswell just reset 2010-11-30T18:03:32 Is the cutoff top N or ELO > X? 2010-11-30T18:03:45 top N 2010-11-30T18:04:00 is that the same time the finals end? 2010-11-30T18:04:08 *** amstan has quit IRC (Remote host closed the connection) 2010-11-30T18:04:26 sigh: no the finals end somewhere around 24 hours from now 2010-11-30T18:04:33 *** amstan has joined #aichallenge 2010-11-30T18:04:33 *** ChanServ sets mode: +o amstan 2010-11-30T18:04:40 ah, ok, thanks for the heads up 2010-11-30T18:05:02 janzert: will the cutoff keep going? because imho its a horrible mistake :S 2010-11-30T18:05:07 once Jeff gives me a firm time I'll post it in the finals information thread as well 2010-11-30T18:05:18 the cuttof must be 100 plus something more to let people have the oportunity of fighting for top100 2010-11-30T18:05:27 I think 1000 is fine? 2010-11-30T18:05:35 150 2010-11-30T18:05:40 even 1000 is already dangerous 2010-11-30T18:05:42 anyway I have to head out for a few hours 2010-11-30T18:05:49 :/ 2010-11-30T18:06:14 BaronTrozo: 150 is no good... people below 150 still get wins against top 100 or top 50 2010-11-30T18:06:21 indeed 2010-11-30T18:06:30 and the same happens all across the system 2010-11-30T18:06:41 hell, my bot is like 130 and it has bean bixor 2010-11-30T18:06:47 *beaten 2010-11-30T18:07:11 yeah but with this we stabilzed better the high positions, but this is only my oppinion 2010-11-30T18:07:21 BaronTrozo: no we don't 2010-11-30T18:07:41 there are very agressive strategies in play that win of top 100 bots a lot 2010-11-30T18:07:45 but also lose of < 200 2010-11-30T18:07:52 if you remove everything < 200 2010-11-30T18:07:55 they will climb 2010-11-30T18:08:02 okey, don't worry it will be waht organization decides, not what i decide 2010-11-30T18:08:04 which imo is unfair 2010-11-30T18:08:23 yes i know, but it just surprises me ppl don't realize this 2010-11-30T18:08:59 I'll be sad if my bot gets eliminated from the final competition 2010-11-30T18:09:06 even though it is well deserved 2010-11-30T18:09:52 hah back in top 50 2010-11-30T18:09:58 each one of us has his own oppinion, and his own reasons so it's better to wait for the organization 2010-11-30T18:11:02 my bot is like hanging at 50 by its teeth 2010-11-30T18:11:09 http://ai-contest.com/profile.php?user_id=5822 2010-11-30T18:11:32 hmmm... difference bw #100 and #1000 is 740 ELO 2010-11-30T18:11:35 *** amstan has quit IRC (Remote host closed the connection) 2010-11-30T18:11:43 I think that is large enough for a cutfoff 2010-11-30T18:12:09 bhasker: "Hobbyist Game Programmer" what have you worked on? 2010-11-30T18:12:27 yes true, but in the 1000 rankings ranks will change drastically 2010-11-30T18:12:29 *** amstan has joined #aichallenge 2010-11-30T18:12:29 *** ChanServ sets mode: +o amstan 2010-11-30T18:12:32 nothing much really i just keep tinkering and love to follow game dev 2010-11-30T18:12:47 ah 2010-11-30T18:12:52 i contributed a small demo to crystal space back in 2001 and have played around with ogre and panda3d etc 2010-11-30T18:12:59 but i always get stuck midway due to lack of artists 2010-11-30T18:13:03 i have grand ideas for games 2010-11-30T18:13:08 but i can never get someone to do some artwork! 2010-11-30T18:13:12 I write a basic 3d game engine every couple of years 2010-11-30T18:13:17 and then lose steam, haha 2010-11-30T18:13:22 lol 2010-11-30T18:13:23 ah i am not much an engine programmer 2010-11-30T18:13:31 too much math:o 2010-11-30T18:13:33 i wrote an ai for an open source rts game called spring 2010-11-30T18:13:40 ah TA clone 2010-11-30T18:13:41 which is doing pretty good :) 2010-11-30T18:13:41 !! 2010-11-30T18:13:42 TA: Spring? 2010-11-30T18:13:43 yes 2010-11-30T18:13:51 i should play that 2010-11-30T18:13:54 its awesome 2010-11-30T18:13:54 been busy with supreme commander 2010-11-30T18:13:59 i loved the original TA 2010-11-30T18:14:04 I liked Spring a bit 2010-11-30T18:14:04 <--big RTS fan 2010-11-30T18:14:10 but I played it last when it was very buggy 2010-11-30T18:14:11 * Error323 <<-- also 2010-11-30T18:14:11 been playing since warcraft 1 days 2010-11-30T18:14:20 actually before that 2010-11-30T18:14:26 but thats the one i remember 2010-11-30T18:14:40 wowa, i'm out of the 80 rankings 2010-11-30T18:14:44 79 \o/ 2010-11-30T18:14:45 grats 2010-11-30T18:14:52 :P 2010-11-30T18:14:59 78 2010-11-30T18:15:06 you are tied 2010-11-30T18:15:06 *** GreenTea_ has quit IRC (Quit: Page closed) 2010-11-30T18:15:08 with jwkam 2010-11-30T18:15:15 jwkvam rather 2010-11-30T18:15:20 and murrayr 2010-11-30T18:15:54 in other news my bot beat jambachilli and smloh1 :d 2010-11-30T18:16:08 i beat smloh quite a lot locally 2010-11-30T18:16:12 45% 2010-11-30T18:16:14 ah 2010-11-30T18:16:32 but yeh other ppl i don't :P 2010-11-30T18:17:44 if the cutoff reaches 1000 we'll get a game approx every 8 minutes 2010-11-30T18:17:56 that would be nice 2010-11-30T18:18:04 though games are taking longer to 2010-11-30T18:18:06 too* 2010-11-30T18:18:32 less people are nice enough to crash quickly now 2010-11-30T18:18:41 hehe ye 2010-11-30T18:18:41 *fewer 2010-11-30T18:18:43 :) 2010-11-30T18:18:43 JamesMG are you interested in some game dev :) 2010-11-30T18:18:56 sigh why did you give up midway into the contest? 2010-11-30T18:18:58 Yeah, when I have time 2010-11-30T18:19:13 i have a few ideas that i want to develop 2010-11-30T18:19:20 http://ai-contest.com/visualizer.php?game_id=9417096 lol 2010-11-30T18:19:29 though i am planning to use panda3d 2010-11-30T18:19:33 to save myself the pain of coding in c++ 2010-11-30T18:19:41 bhasker: lack of time, really 2010-11-30T18:19:45 never looked at panda3d, will have to 2010-11-30T18:20:18 panda3d is engine used by disney for their pirates game 2010-11-30T18:20:22 also, frustration at having a lot of stuff I did just not working :P 2010-11-30T18:20:22 its mostly c++ core 2010-11-30T18:20:26 ah 2010-11-30T18:20:30 with python wrapper apis 2010-11-30T18:20:34 i'm a hobbyist game programmer too :) 2010-11-30T18:20:36 which are very nice 2010-11-30T18:20:45 so makes for really quick development 2010-11-30T18:20:48 so am I 2010-11-30T18:20:51 well AI programmer 2010-11-30T18:20:59 so you layer game code as python on top of the engine? 2010-11-30T18:20:59 well my plan is to write an RTS 2010-11-30T18:21:06 yup 2010-11-30T18:21:16 they even have support for distributed objects 2010-11-30T18:21:21 I am planning to use the next AI contest to learn python I think 2010-11-30T18:21:29 since they use it for their MMO 2010-11-30T18:21:36 i might give lisp a shot 2010-11-30T18:21:41 so am I well www.1c.ru russian version of enterprise java programmer 2010-11-30T18:21:42 been wanting to learn that language 2010-11-30T18:21:43 Here in spain is to late, so good night, and good luck to all 2010-11-30T18:21:51 gnite BaronTrozo 2010-11-30T18:21:52 lisp frankly looks uninteresting the damn syntax is annoying 2010-11-30T18:22:02 i hate updating my .emacs 2010-11-30T18:22:04 its very powerfull and fast 2010-11-30T18:22:06 yes 2010-11-30T18:22:07 me vim user 2010-11-30T18:22:13 but the syntax just drives me nuts 2010-11-30T18:22:14 bhasker: we should talk more, might be fun to do an open source game 2010-11-30T18:22:19 sure 2010-11-30T18:22:20 I have to go for tonight though 2010-11-30T18:22:22 *** BaronTrozo has quit IRC (Quit: Page closed) 2010-11-30T18:22:26 drop me a note 2010-11-30T18:22:27 far the best 2010-11-30T18:22:31 hbhasker at yahoo dot com 2010-11-30T18:22:36 now on linux 2010-11-30T18:22:39 will do 2010-11-30T18:22:50 i have this grand idea for building an MMORTS 2010-11-30T18:23:06 but its too ambitious so i will tone it down for something simpler:) 2010-11-30T18:23:12 damn artwork kills me 2010-11-30T18:23:15 *** amstan has quit IRC (Ping timeout: 260 seconds) 2010-11-30T18:23:20 well if you need AI for it :P 2010-11-30T18:23:33 this IRC chan could write a hell of a game between us :) 2010-11-30T18:23:40 later folks 2010-11-30T18:23:42 haha oh yeah 2010-11-30T18:23:45 *** JamesMG has quit IRC (Quit: home) 2010-11-30T18:23:50 lots of power here :P 2010-11-30T18:24:12 i'm woking on a classic arcade remake in 3d 2010-11-30T18:24:45 wow cutoff drops fast 2010-11-30T18:25:25 ah that would be interesting meneky 2010-11-30T18:25:30 well, it has to make 1000 my midnight server time 2010-11-30T18:25:30 which engine are you usuing 2010-11-30T18:25:31 @time 2010-11-30T18:25:31 sigh: 06:25 PM, November 30, 2010 2010-11-30T18:25:39 from scratch (C++) 2010-11-30T18:25:39 using* 2010-11-30T18:25:46 oef 2010-11-30T18:25:50 that's a lot of work 2010-11-30T18:25:55 sure, the old version : https://groups.google.com/group/frikingshark3d?lnk=srg&hl=es&ie=UTF-8&safe=off&pli=1 2010-11-30T18:26:14 new is quite better :P 2010-11-30T18:26:40 cool 2010-11-30T18:26:50 *** jonoff has quit IRC (Ping timeout: 265 seconds) 2010-11-30T18:26:53 is it kind of like raptor - call of the shadows? 2010-11-30T18:26:58 I <333333333333 that game 2010-11-30T18:27:03 flying shark remake 2010-11-30T18:27:06 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-11-30T18:27:38 just like aero fighters and the like 2010-11-30T18:27:39 try using panda3d:) 2010-11-30T18:27:42 will save you a lot of time 2010-11-30T18:27:43 but simpler 2010-11-30T18:27:46 sure 2010-11-30T18:27:50 from scratch is painful 2010-11-30T18:27:52 but where is the fun then? 2010-11-30T18:27:58 writing all the math classes and transforms 2010-11-30T18:28:00 heh 2010-11-30T18:28:07 well depends what you are really trying to do 2010-11-30T18:28:09 write the game? 2010-11-30T18:28:13 or write the engine:) 2010-11-30T18:28:41 whenever i try to write from scratch i lose interest midway since writing all the boilerplate code is only fun so many times 2010-11-30T18:28:55 yes, thats true 2010-11-30T18:29:03 jep 2010-11-30T18:29:15 the engine is already done (but evolving) now i'm workint in entity and scenario editors 2010-11-30T18:29:18 lots of work 2010-11-30T18:29:50 yea well panda comes with a lot of that 2010-11-30T18:29:54 it has a scene editor too now 2010-11-30T18:29:56 i think* 2010-11-30T18:30:04 i've seen panda 2010-11-30T18:30:06 and being python almost everything is scriptable 2010-11-30T18:30:07 it looks pretty good 2010-11-30T18:30:16 the channel for panda is on this server 2010-11-30T18:30:18 #panda3d 2010-11-30T18:30:23 http://www.panda3d.org/ 2010-11-30T18:30:27 thats where most of the main developers hang out 2010-11-30T18:30:32 quite helpful lot 2010-11-30T18:31:12 the neat thing about the library also is its pstats support 2010-11-30T18:31:18 basicaly it has profiling built in 2010-11-30T18:31:23 instrumentation 2010-11-30T18:31:24 rather 2010-11-30T18:31:35 you can plot graphs of time spent in each function/class etc live 2010-11-30T18:31:36 *** tapwater has quit IRC (Quit: tapwater) 2010-11-30T18:31:41 as your program is running 2010-11-30T18:31:45 pretty neat instrumentation stuff 2010-11-30T18:31:51 cool 2010-11-30T18:34:25 the only downside is you can't do any threading 2010-11-30T18:34:33 at least not in 1.7 version 2010-11-30T18:34:36 next one will support some 2010-11-30T18:34:51 the python interpreter prevents threading from using multiple cores due to its GIL 2010-11-30T18:35:09 though you could just use the C++ api straight if you wanted threading 2010-11-30T18:35:38 btw do you guys read this guys page 2010-11-30T18:35:42 *** Palmik has quit IRC (Remote host closed the connection) 2010-11-30T18:35:49 http://www-cs-students.stanford.edu/~amitp/ 2010-11-30T18:36:14 read his game programming blog 2010-11-30T18:36:18 *** Appleman1234 has joined #aichallenge 2010-11-30T18:36:19 pretty interesting stuff 2010-11-30T18:36:27 and his A* pages are some of the best on the web 2010-11-30T18:36:34 rather pathfinding in general 2010-11-30T18:39:08 *** Appleman1234 has quit IRC (Excess Flood) 2010-11-30T18:39:35 *** Appleman1234 has joined #aichallenge 2010-11-30T18:40:24 interesting stuff 2010-11-30T18:41:21 *** yasith has joined #aichallenge 2010-11-30T18:43:20 surprise in turn 43 :P http://ai-contest.com/visualizer.php?game_id=9424389 2010-11-30T18:43:24 *** davidjliu has joined #aichallenge 2010-11-30T18:44:09 that kind of moves normally make me loss, but this time was the best move i can think of :) 2010-11-30T18:44:17 nice game 2010-11-30T18:47:28 interesting 2010-11-30T18:47:33 mega1s bot played iouri 3 times 2010-11-30T18:47:35 and lost all 3 games 2010-11-30T18:47:54 its pure luck or not enough cpu time 2010-11-30T18:48:02 locally here mega1 rapes iouri 2010-11-30T18:48:17 ah 2010-11-30T18:48:34 Top 10 players: bocsimacko(3850), _iouri_(3653), jimrogerz(3613), Slin-.-(3603), _Astek_(3603), Accoun(3594), protocolocon(3591), GreenTea(3584), asavis(3578), george(3575) 2010-11-30T18:48:36 have you tested other bots against mega1 2010-11-30T18:48:41 yes 2010-11-30T18:48:45 its interesting 2010-11-30T18:48:56 mistmanov's bot for example wins 48 times iirc 2010-11-30T18:49:08 48 out of 150 thats quite good 2010-11-30T18:49:16 what about mine:) 2010-11-30T18:49:22 but also loses ~50 times from someone in the 500 ranks 2010-11-30T18:49:28 hehe 2010-11-30T18:49:32 he just has a very dangerous strategy 2010-11-30T18:49:34 oh, mega1 has released his code? 2010-11-30T18:49:37 yes 2010-11-30T18:49:43 bhasker: haven't tried urs yet :) 2010-11-30T18:49:47 i'll run it 2010-11-30T18:50:03 * bhasker waits impatiently to know how it fares against mega1 2010-11-30T18:50:18 quite well commented code at that sigh 2010-11-30T18:50:25 cool :) 2010-11-30T18:51:09 i've been optimizing my bot against a friend of mine 2010-11-30T18:51:14 which is extremely defensive 2010-11-30T18:51:15 :( 2010-11-30T18:51:19 i rape him 2010-11-30T18:51:29 but do much worse against oothers as a sideeffect lol 2010-11-30T18:52:18 lisp is a real mistery to me 2010-11-30T18:52:28 but it has this elegance to it 2010-11-30T18:54:01 *** amstan has joined #aichallenge 2010-11-30T18:54:01 *** ChanServ sets mode: +o amstan 2010-11-30T18:57:17 *** xxx has quit IRC (Ping timeout: 265 seconds) 2010-11-30T18:58:46 *** meneky__ has joined #aichallenge 2010-11-30T18:59:07 #seen janzert 2010-11-30T18:59:10 @seen janzert 2010-11-30T18:59:10 amstan: janzert was last seen in #aichallenge 53 minutes and 27 seconds ago: anyway I have to head out for a few hours 2010-11-30T18:59:13 darn 2010-11-30T19:00:40 *** meneky has quit IRC (Ping timeout: 265 seconds) 2010-11-30T19:02:38 *** yasith has quit IRC (Ping timeout: 265 seconds) 2010-11-30T19:02:38 bhasker: 2010-11-30T19:02:40 Number of player 1 wins: 5 2010-11-30T19:02:40 Number of player 2 wins: 67 2010-11-30T19:02:40 Number of draws: 3 2010-11-30T19:02:45 you being player 1 2010-11-30T19:02:58 lol 2010-11-30T19:03:04 i guessed as much 2010-11-30T19:03:14 thats sad :( 2010-11-30T19:04:15 well its similar to mine hehe 2010-11-30T19:04:25 he's got a real monster 2010-11-30T19:04:51 the way he's out there at the top with an elo difference > 200 is truely amazing 2010-11-30T19:05:36 My rank is still up there for all my timeouts. 2010-11-30T19:06:10 200 should mean he beats the next guy 3 out of 4 games. 2010-11-30T19:06:13 McLeopold: i've had at least 6 of them 2010-11-30T19:06:25 *** Azrathud has joined #aichallenge 2010-11-30T19:06:31 McLeopold: well if he had more cpu time 2010-11-30T19:06:37 The sad part is, my bot is prone to timeout when I'm winning. 2010-11-30T19:06:45 he'd beat him like 90% or sth 2010-11-30T19:06:54 oh that sux 2010-11-30T19:07:12 Yeah, it's cause it has so many "options" that it needs to sort through. 2010-11-30T19:07:56 Once the results are in, I'm going to get the raw pgn files, switch my timesouts to wins, and console myself with what could have been. 2010-11-30T19:08:17 ;_; 2010-11-30T19:08:44 heh you may not wanna do that 2010-11-30T19:09:16 Well, I figure my pretend rank will only go up. 2010-11-30T19:09:35 I would guess it at 20-15. 2010-11-30T19:09:55 well guys i'm off to bed 2010-11-30T19:09:58 cya 2010-11-30T19:13:00 well i can't even do that, haven't had any timeouts/crashes that i can lay blame on for my bad performance! 2010-11-30T19:16:57 *** yasith has joined #aichallenge 2010-11-30T19:23:50 *** Mathnerd314 has joined #aichallenge 2010-11-30T19:24:23 *** nn has quit IRC (Quit: Page closed) 2010-11-30T19:27:21 *** rebelxt has joined #aichallenge 2010-11-30T19:29:21 *** Epona has joined #aichallenge 2010-11-30T19:32:48 *** phreeza has joined #aichallenge 2010-11-30T19:33:46 *** Epona has quit IRC (Ping timeout: 240 seconds) 2010-11-30T19:39:52 =) I'm on a winning streak again 2010-11-30T19:42:58 *** Mathnerd314 has quit IRC (Disconnected by services) 2010-11-30T19:43:49 *** Mathnerd314_ has joined #aichallenge 2010-11-30T19:43:56 *** Mathnerd314_ is now known as Mathnerd314 2010-11-30T19:47:55 medrimonia: nice. opposite here ;) 2010-11-30T19:48:06 *** jonoff has joined #aichallenge 2010-11-30T19:48:35 Top 10 players: bocsimacko(3843), _iouri_(3649), jimrogerz(3610), Slin-.-(3606), _Astek_(3597), protocolocon(3588), Accoun(3580), GreenTea(3577), asavis(3577), george(3568) 2010-11-30T19:53:06 *** wh1teside has quit IRC (Read error: Connection reset by peer) 2010-11-30T19:55:36 *** medrimonia1 has joined #aichallenge 2010-11-30T19:56:44 *** rhinobot has joined #aichallenge 2010-11-30T19:57:05 *** wh1teside has joined #aichallenge 2010-11-30T19:59:20 *** medrimonia2 has joined #aichallenge 2010-11-30T20:00:00 McLeopold: yo 2010-11-30T20:00:09 *** medrimonia1 has quit IRC (Ping timeout: 265 seconds) 2010-11-30T20:00:27 *** medrimonia1 has joined #aichallenge 2010-11-30T20:01:30 *** medrimonia has quit IRC (Ping timeout: 245 seconds) 2010-11-30T20:03:03 woah.. I've moved up like 15 ranks since earlier today 2010-11-30T20:06:25 rebelxt: what's up 2010-11-30T20:07:10 my bot has put its roots down at 50 2010-11-30T20:07:12 and refuses to budge 2010-11-30T20:07:27 everyone should be finding their place by now 2010-11-30T20:07:32 leo: I bet your bot would awon this, if it went longer: http://ai-contest.com/visualizer.php?game_id=9428862 2010-11-30T20:08:38 hmm.. maybe 2010-11-30T20:08:54 If you bot is deterministic, we could find out... 2010-11-30T20:08:56 I had position+ship advantage.. but your bot is better lol 2010-11-30T20:09:08 it is 2010-11-30T20:11:32 *** smloh1 has joined #aichallenge 2010-11-30T20:11:38 nice one, medrimonia! 2010-11-30T20:11:47 so, I'm working on multiplayer planet wars, who's in? 2010-11-30T20:12:00 teams, or ffa? 2010-11-30T20:12:02 McLeopold: wow, both of us are having a rough patch... 2010-11-30T20:12:02 I've got the game engine. I'm wrapping up the tcp server. 2010-11-30T20:12:07 ffa 2010-11-30T20:12:25 but, I can't stop coordination 2010-11-30T20:12:26 not me.. I don't like the ffa strat 2010-11-30T20:12:50 I don't think anyone knows what that is right now 2010-11-30T20:13:11 Hopefully in a few days you can connect to zeroviz and play multiplayer games. 2010-11-30T20:13:23 turtle.. and preserve your population as you bounce between planets when attacked, until you can win 2010-11-30T20:14:06 I could do teams, but we'd all have to agree on some aspects of the game mechanics. 2010-11-30T20:14:31 McLeopold: i would be interested to see how our current bots would do in a multiplayer environment 2010-11-30T20:14:44 So would I 2010-11-30T20:14:45 probably most of them would freeze up 2010-11-30T20:14:50 yep :) 2010-11-30T20:14:52 mine would fail spectacularly 2010-11-30T20:14:59 if you make it appear as all enemy's are team2.. 2010-11-30T20:15:03 yea 2010-11-30T20:15:06 most would not crash 2010-11-30T20:15:10 but many would sit idle 2010-11-30T20:15:11 expand, then freeze up, otherwise you lose in FFA 2010-11-30T20:15:13 yup 2010-11-30T20:15:15 So, basically, the input from the server will have more than 2 owners 2010-11-30T20:15:28 yea well most bots will need some change 2010-11-30T20:15:30 You will still be player 1 2010-11-30T20:15:32 so handle that 2010-11-30T20:15:41 since it will need to calculate available ships per player etc 2010-11-30T20:15:50 and decide whether to attack user A or B etc 2010-11-30T20:16:02 and differentiate between enemy ships and enemy ships owned by A 2010-11-30T20:16:09 Yeah, just giving you fair warning if you want to try it out. 2010-11-30T20:16:16 sure i would be game for it 2010-11-30T20:16:23 would be interesting to see 2010-11-30T20:16:36 i will probably write it from scratch this time though 2010-11-30T20:16:58 McLeopold: need to try our bots in 3-player FFA. if A attacks B, C wins automatically i'm afraid. but i'm game to check if that's the case ;) 2010-11-30T20:17:04 I'm also going to create a tcp client that will port current bots to play the game. 2010-11-30T20:17:12 the only way I can see, to reasonably attempt multiplayer with existing bots: is to make the team-mate show up as neutral+100: teammates fleets show up as being your own: and all enemy's show up as team2 with population balanced, to be equal to the total of your controlled ships 2010-11-30T20:17:41 burny: that's a good idea. 2010-11-30T20:17:48 well A attacks B , b could attack C 2010-11-30T20:17:54 and then its not so certain 2010-11-30T20:18:00 At least we can see the current bots do something in multiplayer, if not good strategy. 2010-11-30T20:18:09 so if it's 2vs2.. each player thinks they have a single 100population planet, against 2 50population planets 2010-11-30T20:18:50 burny: that complication is why I'm going to do ffa only at first. 2010-11-30T20:18:54 and if team-mates attack each other.. they just give their population to their teammate 2010-11-30T20:19:07 I figure I'll just show all enemies as owner=2 2010-11-30T20:19:15 Why is the number of games played counting backwards at http://ai-contest.com/game_stats.php ? 2010-11-30T20:19:18 imagine our current bots in 3on3 match. both other players are one enemy with twice as many planets.. most current bots will either do nothing, or expand a bit and then do nothing ;-) 2010-11-30T20:19:21 leo, I'm sure that almsot all top 500 bots will sit nearly 100% idle in ffa 2010-11-30T20:19:31 eburnette: games played in the last 24 hours 2010-11-30T20:19:34 err not 3on3.. but 3-player FFA 2010-11-30T20:19:41 ah 2010-11-30T20:19:46 if the gpm 24 hours was greater than it is now, that number will go down 2010-11-30T20:20:01 I would imagine most bots, on most maps, would not expand atall 2010-11-30T20:20:13 and the nub bots, would dominate 2010-11-30T20:20:14 *** amstan has quit IRC (Remote host closed the connection) 2010-11-30T20:20:23 burny heh 2010-11-30T20:20:53 I guess games will take longer to play as cutoff shrinks 2010-11-30T20:21:48 *** smloh1 has quit IRC (Quit: Page closed) 2010-11-30T20:21:56 burny: turn 72 http://ai-contest.com/visualizer.php?game_id=9421041 2010-11-30T20:22:05 not all bots are passive when behind 2010-11-30T20:23:29 not all: but most 2010-11-30T20:24:37 bosco looks like it would do well ffa 2010-11-30T20:24:38 check out live 4-player games in galcon. they had to introduce self destruct rule (everyone loses if no planet changes hands in 30 seconds) to try and avoid stalemates. it barely helped. 2on2 on the other hand is awesome.. no stalemates 2010-11-30T20:25:10 =) I've just done my first win against bocsimacko! 2010-11-30T20:25:40 congrats 2010-11-30T20:25:41 grats, that's not easy 2010-11-30T20:25:52 *** McLeopold has quit IRC (Quit: Page closed) 2010-11-30T20:27:14 wow, gained 20ranks since noon today 2010-11-30T20:27:31 at this rate.. I'll be in top10 by the tiem it's over! 2010-11-30T20:27:49 burny, can you kick my bot in the nuts please? thanks 2010-11-30T20:27:51 tomorrow you'll be #-57 2010-11-30T20:28:06 well, since you asked nice.. 2010-11-30T20:28:12 lol eburn :) 2010-11-30T20:29:13 just need a few wins on boc's, and I could be top2 2010-11-30T20:29:34 easier said than done 2010-11-30T20:29:46 *** yasith has quit IRC (Ping timeout: 245 seconds) 2010-11-30T20:31:18 wonder how hard it'd be, to play bot vs bocs: with a handicap(extra ships).. to determine what handicap is required, to win 50% 2010-11-30T20:31:33 *** mceier has quit IRC (Quit: leaving) 2010-11-30T20:31:56 How I don't think you'll need a lot 2010-11-30T20:31:56 might be interesting to see how people's ELO's are, comapred to their 'vs bocs handicap' 2010-11-30T20:32:03 haha. that's a good idea 2010-11-30T20:32:16 even five ships of difference at the beginning is already a lot 2010-11-30T20:32:46 I bet a 5ship handicap, that bocs would still win more than 50%, against most bots, on most maps 2010-11-30T20:33:12 against most bots but not against the top 50 bots I think 2010-11-30T20:33:14 yep he would probably stay #1 with 1-2 ship handicap 2010-11-30T20:33:58 I think most people would need more than 5 2010-11-30T20:34:06 actually this is easy to test locally.. will do tonight 2010-11-30T20:34:07 same here eburn 2010-11-30T20:36:52 *** rebelxt has left #aichallenge 2010-11-30T20:41:08 awww... someone's gota feel bad about this one.. 2010-11-30T20:41:09 http://ai-contest.com/visualizer.php?game_id=9434366 2010-11-30T20:41:33 O.o 2010-11-30T20:42:43 *** yasith has joined #aichallenge 2010-11-30T20:48:04 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-30T20:48:36 Top 10 players: bocsimacko(3833), _iouri_(3638), Slin-.-(3599), jimrogerz(3598), _Astek_(3585), protocolocon(3577), GreenTea(3576), Accoun(3574), asavis(3570), george(3560) 2010-11-30T20:49:47 does anybody have a tarball for mega1's bocsimacko's bot? 2010-11-30T20:53:13 when's contest end? tomorrow night? 2010-11-30T20:59:34 *** meneky__ has quit IRC (Ping timeout: 265 seconds) 2010-11-30T21:00:15 *** bhasker has quit IRC (Quit: bhasker) 2010-11-30T21:02:08 *** superflit has quit IRC (Quit: superflit) 2010-11-30T21:02:14 *** McLeopold has joined #aichallenge 2010-11-30T21:03:17 I'm 2/2 on bix, and 2/3 on dmj 2010-11-30T21:05:07 which one is yours burny? 2010-11-30T21:05:13 nikiburny 2010-11-30T21:07:38 this game is my pain :( http://ai-contest.com/visualizer.php?game_id=9042759 2010-11-30T21:09:30 sending 86 instead of 56, otherwise I would have had advantage 2010-11-30T21:09:31 *** Mathnerd314 has quit IRC (Ping timeout: 255 seconds) 2010-11-30T21:12:32 I wish the system would give me more games against better rated bots, like http://ai-contest.com/visualizer.php?game_id=9428523 , http://ai-contest.com/visualizer.php?game_id=9424003 , http://ai-contest.com/visualizer.php?game_id=9408064 , http://ai-contest.com/visualizer.php?game_id=9406111 , http://ai-contest.com/visualizer.php?game_id=9404303 2010-11-30T21:12:47 *** xxx has joined #aichallenge 2010-11-30T21:14:26 I think you were already behind when you sent the 86 burny 2010-11-30T21:15:52 wasn't 2010-11-30T21:16:03 I woulda been ahead by 1ship 2010-11-30T21:16:27 it was an exact tie, till he attacked the 8s 2010-11-30T21:16:44 and if I sniped the one: woulda meant he spent 16, and grew 15 2010-11-30T21:16:56 prioer to when we each owned 1 2010-11-30T21:17:12 my bot woulda ideally sniped the other 8 also: but it wouldn't have been able to hold it 2010-11-30T21:18:01 at turn 7 he attacked the bottom 8 which would put him ahead 2010-11-30T21:18:19 he spent 8 to take it though 2010-11-30T21:18:22 I woulda spent 0 2010-11-30T21:18:37 *** yasith has quit IRC (Ping timeout: 264 seconds) 2010-11-30T21:19:02 you couldn't snipe it because he had the 4 sitting there 2010-11-30T21:19:21 coud have taken it for a few turns 2010-11-30T21:20:14 1 turn at most 2010-11-30T21:20:28 anything you sent could be countered 2010-11-30T21:20:47 *** rhinobot has quit IRC (Quit: Page closed) 2010-11-30T21:20:47 ya, true.. 1 turn 2010-11-30T21:21:10 still though.. 5ships, for nothing 2010-11-30T21:22:49 3 maybe 2010-11-30T21:23:12 minus lost opportunity 2010-11-30T21:24:02 1 turn: he could counter it from his main base, the next turn(resnipe) 2010-11-30T21:25:01 actually, it woulda given me a 7ship advantage, not 1.. those planets were only 3grow 2010-11-30T21:25:14 so he spent 8ships to gain 3grow 2010-11-30T21:25:33 he woulda won anyway though, I'm sure.. 2010-11-30T21:26:20 but the whole, leaving my base undefended thing, is the cause of about half my losses 2010-11-30T21:27:38 Yeah I had that same problem with many versions 2010-11-30T21:27:59 I just didn't have time to finish it.. missing a major feature 2010-11-30T21:28:23 So finish it and play on zeroviz, I'm playing Accoun there now 2010-11-30T21:28:35 my bot doesn't ever, ever, try to "re-enforce the front lines", or "make sure it has enough ships to defend" 2010-11-30T21:30:40 meh, nto worth the effort anymore 2010-11-30T21:30:59 you've been chatting in irc for days, could have been programming :) 2010-11-30T21:31:14 I have been :P 2010-11-30T21:31:17 or playing games 2010-11-30T21:32:24 *** sanddbox has joined #aichallenge 2010-11-30T21:42:59 What's the next cutoff for the tourny? 2010-11-30T21:43:20 cuttoff dropping every few seconds, down to 1000 at midnight 2010-11-30T21:43:31 cool 2010-11-30T21:43:36 Ok, thanks 2010-11-30T21:43:43 I still have a while longer to live then :) 2010-11-30T21:45:43 See http://ai-contest.com/game_stats.php 2010-11-30T21:48:36 Top 10 players: bocsimacko(3830), _iouri_(3630), Slin-.-(3592), jimrogerz(3587), _Astek_(3583), Accoun(3565), asavis(3564), GreenTea(3563), protocolocon(3561), george(3552) 2010-11-30T21:53:04 candidate for fastest top100 game. http://ai-contest.com/profile.php?user_id=13271 2010-11-30T21:53:55 weird 2010-11-30T21:54:18 ya.. 2010-11-30T21:54:29 it's the kind of play you'd expect to see in a chess match 2010-11-30T21:55:19 *** politik has quit IRC (Read error: Operation timed out) 2010-11-30T21:57:35 *** rebelxt_ has joined #aichallenge 2010-11-30T21:58:07 *** politik has joined #aichallenge 2010-11-30T21:59:59 *** McLeopold has quit IRC (Ping timeout: 265 seconds) 2010-11-30T22:06:58 for nearly the first 24hours, I didn't lose to anyone that isn't in the top100 2010-11-30T22:09:20 for nearly the first 24hours, I didn't... beat anyone in the top 100 2010-11-30T22:09:50 ok, slight exaggeration. 12 hours 2010-11-30T22:13:48 *** Rubicon-|-Cross has joined #aichallenge 2010-11-30T22:14:11 *** Naktibalda has quit IRC (Remote host closed the connection) 2010-11-30T22:14:32 *** superflit has joined #aichallenge 2010-11-30T22:30:27 Check out this game using a random move generator: http://tcp.zeroviz.us/canvas.cgi?game_id=36379 2010-11-30T22:34:21 eburnette: seriously? 2010-11-30T22:34:25 wait 2010-11-30T22:34:27 wuht? 2010-11-30T22:35:11 yeah, sends a bunch of random ships to random destinations, does that a bunch of times and picks the best one 2010-11-30T22:36:57 jeez, I better not lose to it... 2010-11-30T22:37:46 eburnette: running it right now? wanna see if i would manage to lose to it :) 2010-11-30T22:38:07 trying to fix a timeout problem, gimme a few 2010-11-30T22:38:48 you can play against me and accoun in the meantime :) 2010-11-30T22:39:21 eburnette: are you talking about UCT or something? 2010-11-30T22:39:33 davidjliu: fine! ;) 2010-11-30T22:39:57 my bot's actually been beating accoun... probably not for long 2010-11-30T22:40:27 no not as fancy as uct 2010-11-30T22:41:08 heh, ok 2010-11-30T22:42:53 yey ;-) 2010-11-30T22:43:54 ah, I forgot that rebelxt_ actually kills my bot 2010-11-30T22:44:11 heh, I just lost 130 elo 2010-11-30T22:44:36 okay, maybe that wasn't killing. whatever 2010-11-30T22:45:22 *** Appleman1234 has quit IRC (Ping timeout: 255 seconds) 2010-11-30T22:45:29 heh, a win by a single ship 2010-11-30T22:45:30 *** bss03-AFK is now known as bss03 2010-11-30T22:47:33 the mere presence of rebelxt has made my bot lose over 200 elo :) 2010-11-30T22:48:37 Top 10 players: bocsimacko(3824), _iouri_(3627), Slin-.-(3590), jimrogerz(3583), _Astek_(3578), Accoun(3561), GreenTea(3553), asavis(3553), protocolocon(3552), george(3542) 2010-11-30T22:49:04 davidjliu: heh you lost that 2nd game by 7 ships. sorry :( 2010-11-30T22:49:56 why are we getting these draw maps? my bots is optimized for fast expansion, but it will lose to you on "normal" maps 2010-11-30T22:51:41 draw maps are unavoidable 2010-11-30T22:51:55 well, without removing possible interesting maps 2010-11-30T22:52:55 antimatroid: do you think you were timing out in this game? http://ai-contest.com/visualizer.php?game_id=9399841 2010-11-30T22:53:19 *** phreeza has quit IRC (Quit: Computer has gone to sleep) 2010-11-30T22:54:25 or some sort of overflow error? 2010-11-30T22:54:34 *** superflit_ has joined #aichallenge 2010-11-30T22:54:40 no, it just made a really bad first move 2010-11-30T22:54:50 and 2nd through 10th or so 2010-11-30T22:55:33 antimatroid: weird it did not capture that 1 for a number of turns 2010-11-30T22:55:38 the tree isn't working properly, i'm honestly really surprised it's ranked that high 2010-11-30T22:55:54 eburnette: that's cool, I wanted to do that 2010-11-30T22:56:48 *** superflit has quit IRC (Ping timeout: 250 seconds) 2010-11-30T22:56:48 *** superflit_ is now known as superflit 2010-11-30T23:00:56 *** bss03 is now known as bss03-AFK 2010-11-30T23:01:29 3rd place is still in flux apparently 2010-11-30T23:05:22 Random bot is running now 2010-11-30T23:07:20 cool, hopefully i'll get through accoun's bots ;) 2010-11-30T23:10:39 *** Appleman1234 has joined #aichallenge 2010-11-30T23:10:49 second has stablized? 2010-11-30T23:11:36 well, not sure, but iouri has been holding 2nd for 24 hours if i'm not mistaken 2010-11-30T23:11:47 hmm, http://tcp.zeroviz.us/canvas.cgi?game_id=36446 looked good for random bot for a while 2010-11-30T23:12:04 yeah, i don't see second place changing 2010-11-30T23:12:23 i wouldn't mind my bot jumping up a couple of places 2010-11-30T23:12:29 it's at 31 2010-11-30T23:13:31 eburnette: yeah i had to dig deep ;-) 2010-11-30T23:13:54 i would love to see someone implement a minimax tree with them as the returning player and trying to exploit minimax 2010-11-30T23:14:07 i think it'd be more likely to show sucess with something like tron though 2010-11-30T23:14:27 antimatroid: you've been pretty stable though which is nice. i've been sliding all day today. -13 2010-11-30T23:14:34 ouch 2010-11-30T23:14:42 i peaked in the 20s i think 2010-11-30T23:14:50 and haven't really dropped back to 40 since going above 2010-11-30T23:14:51 I wonder how well this bot would work against top ranked bots.. 2010-11-30T23:15:05 figure out the best move you can make, and the best counter to it, etc.. with minimax.. 2010-11-30T23:15:14 i've stopped thinking about this problem and moved onto things i've been neglecting 2010-11-30T23:15:19 and then for your move: play the opponents counter-move 2010-11-30T23:15:33 i'm not going to try and crank an arimaa bot out in a month either, i think i might take a year and try to throw a good one together 2010-11-30T23:16:17 oldman is surging late 2010-11-30T23:16:26 he was hanging around 30 till about a day ago, now he's 21 2010-11-30T23:16:38 smloh dropped 2010-11-30T23:16:40 yeah he is. randomness ftw 2010-11-30T23:16:49 learn lisp first antimatroid 2010-11-30T23:17:30 *** yasith has joined #aichallenge 2010-11-30T23:17:42 eburnette: nah, i'm going to learn haskell 2010-11-30T23:17:49 but c++ <3 for ai bots 2010-11-30T23:17:55 if i'm using a tree, i'm using c++ 2010-11-30T23:18:35 I'd pick c over c++ but prolly cause i'm more familiar with it 2010-11-30T23:18:50 i just treat c++ like c with classes 2010-11-30T23:19:11 i'm slowly starting to learn some more c++, but atm most of my code just uses pieces of stl and not much else 2010-11-30T23:19:28 * eburnette imagines rebeltxt yelling at his bot to do something 2010-11-30T23:20:08 rebelxt I mean, dunno why I keep putting a t in there 2010-11-30T23:20:17 i'm still quite surprised by the lack of participation from asian nations 2010-11-30T23:20:22 i would have thought they'd love this shit 2010-11-30T23:20:38 no offense to them, that's more of a compliment from me :) 2010-11-30T23:20:59 eburnette: heh everyone misspells it 2010-11-30T23:21:04 I couldn't get anybody at work interested in it :( 2010-11-30T23:21:23 yeah, i didn't convince anyone where i live to give a decent attempt 2010-11-30T23:21:41 possibly a wise move on their part 2010-11-30T23:21:48 *** smloh1 has joined #aichallenge 2010-11-30T23:21:58 eburnette: i got "lucky" and got 7 people involved.. 2 out of 7 got addicted 2010-11-30T23:21:59 i think i should nickname my bot 'Streaky" 2010-11-30T23:21:59 smloh1: i finally got one against you :) 2010-11-30T23:22:15 antimatroid: i think you got two :( 2010-11-30T23:22:25 :P 2010-11-30T23:22:38 Streaky seems to like dinnertime... 2010-11-30T23:23:21 ohh, bhasker needs to jump 4 places before the end too 2010-11-30T23:23:25 so he can make top 50 2010-11-30T23:23:52 smiley has climbed, but nowhere near enough to catch me :) 2010-11-30T23:24:23 smiley has been going up like crazy today 2010-11-30T23:25:37 I think I need a better evaluation function than result = (double) node.PredictedNumShips(movesLeft, 1) / node.PredictedNumShips(movesLeft, 2); 2010-11-30T23:28:05 eburnette: why is it / not -? 2010-11-30T23:28:53 *** rebelxt_ has quit IRC (Quit: Page closed) 2010-11-30T23:28:54 Trying to get a number which does not increase simply because of the progression of the game 2010-11-30T23:29:22 ah i see 2010-11-30T23:30:08 well it wouldn't if both were about even 2010-11-30T23:31:28 I'll try it 2010-11-30T23:33:16 I need to learn how to use compilers properly some day :( 2010-11-30T23:35:56 i want to write my own operating system, programs and programming language, with a full keyboard containing the greek alphabet and maths notation 2010-11-30T23:36:01 unfortunately i don't think that'd feasible 2010-11-30T23:36:23 nobody ever makes things how i want them :P 2010-11-30T23:36:49 start with an old APL keyboard 2010-11-30T23:37:45 darn, I got tricked just like dmj... 2010-11-30T23:37:45 http://ai-contest.com/visualizer.php?game_id=9456606 2010-11-30T23:37:51 that's beautiful 2010-11-30T23:37:55 why don't they do that anymore! 2010-11-30T23:38:33 i'd want more possible operator/relation symbols though 2010-11-30T23:39:14 i need to have a stern word with keyboard maufacturers 2010-11-30T23:39:25 evil, evil move by bocsi 2010-11-30T23:39:25 i have searched and can't find a single custom keyboard giving anything like what i would want 2010-11-30T23:39:42 get one of those oled keyboards and draw your own 2010-11-30T23:39:48 get oen of the ones with programmable LED keytops 2010-11-30T23:40:06 burny: that's probably a good idea 2010-11-30T23:40:25 the next thing would be, do source files have to be using characters from a particular ascii set? 2010-11-30T23:40:38 no 2010-11-30T23:40:44 <3 2010-11-30T23:40:49 that's now one of my summer projects 2010-11-30T23:43:01 if you attacked it 1turn earlier, you woulda won? or wouldn't gain any population from it, and would have to do that on teh turn before? 2010-11-30T23:43:13 burny: any idea who makes such keyboardS? 2010-11-30T23:43:19 i'm looking and not finding 2010-11-30T23:43:37 @google programmable led keyboard 2010-11-30T23:43:38 burny: Optimus Maximus keyboard: ; Luxeed U5 White LED Illuminated and Individually Programmable ...: ; 40x6 inch Scrolling Programmable LED Sign Wireless Keyboard ...: them? 2010-11-30T23:43:44 there's a new lang that uses funny chars, fortress maybe? 2010-11-30T23:45:09 *** virdo has joined #aichallenge 2010-11-30T23:45:38 http://www.youtube.com/watch?v=K4YyonPeoYQ 2010-11-30T23:45:53 damn that optimus keyboard is expensive 2010-11-30T23:46:01 that's like the price of my computer aha 2010-11-30T23:48:38 Top 10 players: bocsimacko(3826), _iouri_(3625), Slin-.-(3584), jimrogerz(3576), _Astek_(3573), Accoun(3559), asavis(3549), GreenTea(3548), protocolocon(3546), george(3538) 2010-11-30T23:48:57 same price as my monitor lol.. 2010-11-30T23:49:08 and I look at my monitor a LOT more than I look at my keyboard 2010-11-30T23:49:50 *** smloh1 has quit IRC (Quit: Page closed) 2010-11-30T23:49:51 my monitor is my keyboard - Sent from my iPhone 2010-11-30T23:50:22 * sigh doesn't actually own an iPhone 2010-11-30T23:50:25 hmmm, i want a cheap rip off of that, or else i might have to resort to just reprogramming keys 2010-11-30T23:50:39 Looks like I need to do a whole lot more moves per second to pull off the random bot 2010-11-30T23:50:52 just key a normal keyboard and write your own labels 2010-11-30T23:50:57 *just get 2010-11-30T23:51:23 *** Azrathud has quit IRC (Ping timeout: 265 seconds) 2010-11-30T23:52:46 * eburnette yawns 2010-11-30T23:52:52 l8r 2010-11-30T23:52:57 *** eburnette has quit IRC (Quit: Page closed) 2010-11-30T23:53:20 ouch, i lost to 142