2010-10-08T00:00:01 on the bot I just submitted... wait take it back!!! 2010-10-08T00:00:04 :) 2010-10-08T00:00:14 *** jesionaj has quit IRC (Ping timeout: 276 seconds) 2010-10-08T00:03:07 *** yasith has joined #aichallenge 2010-10-08T00:03:18 :) 2010-10-08T00:03:48 http://72.44.46.68/canvas?game_id=366910 2010-10-08T00:04:03 *** utyut has joined #aichallenge 2010-10-08T00:05:23 that is a nice solid win, too. 2010-10-08T00:10:42 *** delt0r_ has joined #aichallenge 2010-10-08T00:12:26 *** delt0r___ has quit IRC (Ping timeout: 272 seconds) 2010-10-08T00:42:20 *** oPless has quit IRC (Ping timeout: 240 seconds) 2010-10-08T00:59:55 Top 10 players: Hazard(3990), felixcoto(3902), rsergio(3857), cfaftw(3790), bix0r4ever(3779), ofrias(3767), savchenko(3744), davidjliu(3739), EBraun(3731), sequoh(3699) 2010-10-08T00:59:55 *** hornairs has joined #aichallenge 2010-10-08T01:03:16 *** Utkarsh has quit IRC (Read error: Connection reset by peer) 2010-10-08T01:10:29 *** Utkarsh has joined #aichallenge 2010-10-08T01:14:29 for OCaml, the main file should be MyBot.ml, right? 2010-10-08T01:17:32 *** optimum has joined #aichallenge 2010-10-08T01:30:00 *** Rybo has joined #aichallenge 2010-10-08T01:32:00 *** optimum has quit IRC (Ping timeout: 245 seconds) 2010-10-08T01:34:53 *** Rybo has quit IRC (Client Quit) 2010-10-08T01:38:52 *** optimum has joined #aichallenge 2010-10-08T01:43:09 *** Palmik has joined #aichallenge 2010-10-08T01:44:41 *** rebelxt_ has joined #aichallenge 2010-10-08T01:45:53 *** rebelxt_ has quit IRC (Client Quit) 2010-10-08T01:54:28 *** Itkovian has quit IRC (Quit: Itkovian) 2010-10-08T01:58:43 *** JCS^ has quit IRC (Quit: \x00) 2010-10-08T01:59:55 Top 10 players: Hazard(3990), felixcoto(3906), rsergio(3836), cfaftw(3766), ofrias(3765), EBraun(3741), davidjliu(3722), savchenko(3707), Arekku(3694), sequoh(3693) 2010-10-08T02:03:41 Zannick: I wonder after your compile_anything cleanup is merged if the language information could be separated out into its own module and the language specific command and main code file added to it and moved out of the database. That way all language information would be in one spot the database would just be used to enable which languages were active. 2010-10-08T02:03:57 yeah 2010-10-08T02:16:46 *** Itkovian has joined #aichallenge 2010-10-08T02:22:38 *** Utkarsh has quit IRC (Ping timeout: 255 seconds) 2010-10-08T02:25:39 *** Itkovian has quit IRC (Quit: Itkovian) 2010-10-08T02:27:48 *** airhorns has joined #aichallenge 2010-10-08T02:31:36 *** hornairs has quit IRC (Ping timeout: 245 seconds) 2010-10-08T02:33:57 *** hellman has joined #aichallenge 2010-10-08T02:37:55 *** hornairs has joined #aichallenge 2010-10-08T02:38:29 *** sigh has quit IRC () 2010-10-08T02:38:42 *** jesionaj has joined #aichallenge 2010-10-08T02:39:05 *** jesionaj` has quit IRC (Ping timeout: 260 seconds) 2010-10-08T02:41:34 *** airhorns has quit IRC (Ping timeout: 265 seconds) 2010-10-08T02:51:42 Janzert: sounds good, i was thinking of the exact thing a few days ago 2010-10-08T02:52:05 i really think i should get more involved in writing the core base of the contest next time 2010-10-08T02:59:56 Top 10 players: Hazard(3912), felixcoto(3859), rsergio(3791), ofrias(3734), cfaftw(3719), EBraun(3702), davidjliu(3701), sequoh(3671), savchenko(3666), Arekku(3654) 2010-10-08T03:00:54 *** wvdschel has joined #aichallenge 2010-10-08T03:02:04 dmj's resubmission loss to a rank 500+ bot seems like an argument against taking previous rank into account 2010-10-08T03:09:20 *** Mathnerd314 has quit IRC (Quit: ChatZilla 0.9.86-rdmsoft [XULRunner 1.9.2.8/20100722155716]) 2010-10-08T03:21:12 *** jesionaj` has joined #aichallenge 2010-10-08T03:24:02 *** jesionaj has quit IRC (Ping timeout: 240 seconds) 2010-10-08T03:25:39 *** aerique has joined #aichallenge 2010-10-08T03:26:00 *** jesionaj has joined #aichallenge 2010-10-08T03:28:05 *** jesionaj` has quit IRC (Ping timeout: 260 seconds) 2010-10-08T03:30:11 *** airhorns has joined #aichallenge 2010-10-08T03:34:45 *** hornairs has quit IRC (Ping timeout: 276 seconds) 2010-10-08T03:43:55 *** sigh has joined #aichallenge 2010-10-08T03:44:25 *** perror has joined #aichallenge 2010-10-08T03:47:06 *** criis_ has quit IRC (Ping timeout: 276 seconds) 2010-10-08T03:47:51 *** criis has joined #aichallenge 2010-10-08T03:54:16 *** jesionaj` has joined #aichallenge 2010-10-08T03:58:09 *** jesionaj has quit IRC (Ping timeout: 276 seconds) 2010-10-08T03:59:56 Top 10 players: Hazard(3902), felixcoto(3843), rsergio(3776), ofrias(3709), cfaftw(3703), davidjliu(3683), sequoh(3675), EBraun(3672), savchenko(3637), Arekku(3633) 2010-10-08T04:02:51 *** Itkovian has joined #aichallenge 2010-10-08T04:09:55 *** Florian has joined #aichallenge 2010-10-08T04:12:36 *** stevenheidel has joined #aichallenge 2010-10-08T04:13:38 *** jmcarthur has quit IRC (Ping timeout: 240 seconds) 2010-10-08T04:17:37 *** jesionaj has joined #aichallenge 2010-10-08T04:19:06 *** jesionaj` has quit IRC (Ping timeout: 245 seconds) 2010-10-08T04:21:15 *** edcba_ is now known as edcba 2010-10-08T04:21:24 *** Florian has quit IRC (Disconnected by services) 2010-10-08T04:21:29 *** jaspervdj has joined #aichallenge 2010-10-08T04:21:42 *** Florian95 has joined #aichallenge 2010-10-08T04:22:06 *** hellman has quit IRC (Remote host closed the connection) 2010-10-08T04:22:13 *** jmcarthur has joined #aichallenge 2010-10-08T04:24:23 *** mega1 has joined #aichallenge 2010-10-08T04:29:30 *** oPless has joined #aichallenge 2010-10-08T04:32:59 *** Florian95 has quit IRC (Quit: Leaving...) 2010-10-08T04:37:54 *** Florian95 has joined #aichallenge 2010-10-08T04:39:05 *** TomyMMX has joined #aichallenge 2010-10-08T04:39:18 Hi all 2010-10-08T04:39:34 I have a question about the server compiling C# entryes 2010-10-08T04:39:56 since yesterday none of my code gets compiled 2010-10-08T04:40:18 *** Naktibalda has joined #aichallenge 2010-10-08T04:40:34 sorry, on my way out but see http://code.google.com/p/ai-contest/issues/detail?id=185 2010-10-08T04:40:41 and it is always teh same error 2010-10-08T04:42:31 heh, ok.. then I'll work on my bot and hope this gets resolved:D 2010-10-08T04:44:51 *** HakanD has joined #aichallenge 2010-10-08T04:44:53 *** wh1teside has quit IRC (Ping timeout: 252 seconds) 2010-10-08T04:45:00 morning 2010-10-08T04:45:15 @rankings 2010-10-08T04:45:16 HakanD: Top 10 players: Hazard(3896), felixcoto(3824), rsergio(3765), ofrias(3704), cfaftw(3693), davidjliu(3677), sequoh(3670), EBraun(3662), savchenko(3639), Arekku(3629) 2010-10-08T04:47:15 evening 2010-10-08T04:47:20 @predict 2010-10-08T04:47:20 I predict that bss03 will win! 2010-10-08T04:48:13 morning 2010-10-08T04:51:22 *** Florian95 has joined #aichallenge 2010-10-08T04:59:57 Top 10 players: Hazard(3893), felixcoto(3822), rsergio(3762), ofrias(3702), cfaftw(3690), davidjliu(3669), sequoh(3668), EBraun(3660), savchenko(3636), Arekku(3626) 2010-10-08T05:02:48 * aerique found what made his bot time out, does a happy dance 2010-10-08T05:07:02 *** zl has joined #aichallenge 2010-10-08T05:08:02 *** nrub_ has joined #aichallenge 2010-10-08T05:08:41 *** fawek has joined #aichallenge 2010-10-08T05:16:34 aerique: did you see that i updated jbotmanager.jar 2010-10-08T05:17:12 delt0r_: yes, thanks! i'll be using it in a bit 2010-10-08T05:17:14 crap I just found a bug 2010-10-08T05:17:22 8-D 2010-10-08T05:17:40 guess i'll wait 2010-10-08T05:18:03 aerique: Its in the human player bit 2010-10-08T05:18:13 if you don't use that it won't affect you 2010-10-08T05:19:46 *** stevenheidel has quit IRC (Quit: Leaving) 2010-10-08T05:20:36 delt0r_: okey thanks, i never leave the replays tab :) 2010-10-08T05:20:57 well it should be more stable now 2010-10-08T05:21:09 let me know optherwise 2010-10-08T05:21:59 *** Meatkat has joined #aichallenge 2010-10-08T05:22:14 *** irchs has joined #aichallenge 2010-10-08T05:23:24 *** ichti has joined #aichallenge 2010-10-08T05:31:23 *** wvdschel has quit IRC (Quit: Leaving) 2010-10-08T05:33:03 *** espes has quit IRC (Ping timeout: 276 seconds) 2010-10-08T05:33:43 delt0r_: what's jbotmanager? 2010-10-08T05:34:15 replayer/game engine/human player/tcp player(with preview) in one 2010-10-08T05:35:19 url? 2010-10-08T05:35:33 I am just fixing a bug 2010-10-08T05:35:56 sounds good 2010-10-08T05:35:57 http://www.ai-contest.com/forum/viewtopic.php?f=18&t=751 2010-10-08T05:36:13 http://www.mabs.at/ewing/JBotManager.jar 2010-10-08T05:36:20 but give me a min or 2 2010-10-08T05:36:49 aw sigh, not top 10 anymore? been rewriting? 2010-10-08T05:37:23 Sylph: I haven't been in the top 10 since last week :P 2010-10-08T05:37:27 Ah 2010-10-08T05:37:31 I've been busy with UDK :| 2010-10-08T05:37:36 UDK? 2010-10-08T05:37:40 unreal development kit 2010-10-08T05:37:51 can't spend too much time on this atm. >.< 2010-10-08T05:37:55 ahhh 2010-10-08T05:38:49 but, I've been rewriting/refactoring quite a bit of my bot 2010-10-08T05:38:59 so I can try out more sophisticated ideas 2010-10-08T05:39:54 Ah :) 2010-10-08T05:39:54 fun 2010-10-08T05:40:03 Bet now it predict 3 times as much as my framework does >.< 2010-10-08T05:40:07 mm 2010-10-08T05:40:20 haha, not yet 2010-10-08T05:40:25 *** espes has joined #aichallenge 2010-10-08T05:41:04 :p 2010-10-08T05:42:03 mega1: Its now updated 2010-10-08T05:44:26 *** HakanD has quit IRC (Ping timeout: 240 seconds) 2010-10-08T05:44:29 thanks a lot 2010-10-08T05:44:39 mega1: when will you be uploading a bot? 2010-10-08T05:45:05 aerique: I uploaded one yesterday but it's as clever as I had time. 2010-10-08T05:45:08 very little 2010-10-08T05:45:36 the evaluation function is just the ship count :-o 2010-10-08T05:45:43 mega1: there's still ample time 2010-10-08T05:45:54 yes, in the calendar ... 2010-10-08T05:48:01 LOL 2010-10-08T05:48:12 nice answer, mega1 :p 2010-10-08T05:48:26 unfortunately same with me 2010-10-08T05:48:40 got all the time I need, but they're all in the calendar >.> 2010-10-08T05:59:12 Anyone know the story with the additional servers? Looks like they went offline a few hours ago... 2010-10-08T05:59:57 Top 10 players: Hazard(3879), felixcoto(3808), rsergio(3762), ofrias(3698), cfaftw(3669), sequoh(3669), davidjliu(3651), iMeller(3651), EBraun(3636), savchenko(3623) 2010-10-08T06:01:52 Meatkat: they hang now and then but also get automatically taken offline when the hourly price for an amazon instance goes over budget 2010-10-08T06:03:03 aerique: Ah, that makes sense, thanks! Here's hoping the price drops :) 2010-10-08T06:03:25 Meatkat: or the devs wake up :) 2010-10-08T06:04:30 aerique: Do you think they're hung up at the moment? I wasn't around to see if they just shut off or slowly died... 2010-10-08T06:04:57 Meatkat: i have no idea 2010-10-08T06:06:46 Well I'm out, lates! 2010-10-08T06:06:47 *** Meatkat has quit IRC () 2010-10-08T06:19:57 *** jimmerton has joined #aichallenge 2010-10-08T06:22:56 *** hellman has joined #aichallenge 2010-10-08T06:28:57 *** wh1teside has joined #aichallenge 2010-10-08T06:33:47 *** Utkarsh has joined #aichallenge 2010-10-08T06:35:36 *** Bobng has joined #aichallenge 2010-10-08T06:35:47 *** yasith has quit IRC (Ping timeout: 265 seconds) 2010-10-08T06:38:43 @predict 2010-10-08T06:38:43 I predict that fry_ will win! 2010-10-08T06:40:43 that'd be funny! 2010-10-08T06:42:12 *** jimmerton has quit IRC (Quit: jimmerton) 2010-10-08T06:49:25 glad I didn't claim that bot was much better than the old one :/ 2010-10-08T06:50:11 But, the one that was up for two weeks had the same bug in it that I think caused the loss on turn 5. 2010-10-08T06:54:10 *** yasith has joined #aichallenge 2010-10-08T06:59:58 Top 10 players: Hazard(3886), felixcoto(3812), rsergio(3765), ofrias(3707), cfaftw(3673), sequoh(3662), davidjliu(3656), iMeller(3654), EBraun(3653), savchenko(3629) 2010-10-08T07:01:46 *** ichti has quit IRC (Ping timeout: 264 seconds) 2010-10-08T07:06:48 *** jimmerton has joined #aichallenge 2010-10-08T07:08:57 *** jimmerton has quit IRC (Client Quit) 2010-10-08T07:10:58 dmj111: 1 loss isn't that bad... it's still working its way up 2010-10-08T07:12:52 hmmm... actually you completely failed at neutral stealing in that one... 2010-10-08T07:13:06 my current bot on the site has the same problem :D 2010-10-08T07:15:38 yeah, I never noticed that bug. I launch from one planet with 15 to steal the neutral, than with 45 from a closer planet to stage my ships. 2010-10-08T07:16:00 45 takes the neutral, then boom. There may be more reasons for that loss, but that one is bad. 2010-10-08T07:17:03 as janzert said though... this is certainly reason not to get seeded by previous results. But.... my previous bot would have done that too, so it is also a good reason to not only play people within +/- rankings. 2010-10-08T07:17:12 +/- 10 rankings i mean. 2010-10-08T07:17:59 dmj111: yup, I agree 2010-10-08T07:18:21 also you never notice in testing, because your other bots aren't that silly :P 2010-10-08T07:18:55 that's also why I didn't bother to upload a fixed version, when it gets up to the higher ranks that bug won't matter anymore 2010-10-08T07:19:57 sigh: famous last words.... "that bug won't matter......." 2010-10-08T07:20:21 the weird thing is my newer version doesn't do that, but I don't know why.... 2010-10-08T07:20:32 haha... delt0r_, I have fixed it locally :P 2010-10-08T07:21:21 dmj111: well, you've got a concrete position to test it on... shouldn't be hard to find out why 2010-10-08T07:23:37 *** fawek has quit IRC (Quit: No Ping reply in 180 seconds.) 2010-10-08T07:25:02 I think bots should be seeded at their previous rank, but play more games outside their rank range 2010-10-08T07:25:39 making people fight their way up from a random position doesn't benefit anyone 2010-10-08T07:26:20 sigh: re my concrete position... I plan to :) just trying to make sure it doesn't suck so I can try to resubmit. 2010-10-08T07:26:29 resubmit in a hurry --> probably even worse. 2010-10-08T07:26:41 yet it is one of those lessons i ignore. 2010-10-08T07:26:41 :) 2010-10-08T07:27:16 *** jesionaj has quit IRC (Ping timeout: 252 seconds) 2010-10-08T07:28:02 it takes so long to get back up to rank that I really can't see the benefit in hurrying 2010-10-08T07:28:18 unless you are desperate for your hourly ping from contestbot :P 2010-10-08T07:29:35 *** fawek has joined #aichallenge 2010-10-08T07:34:07 *** Accoun has quit IRC (Ping timeout: 245 seconds) 2010-10-08T07:39:24 ha. you brought me back down from the ledge. 2010-10-08T07:40:42 i just hate having such a bad bug in the wild. But, the whole point of submitting last night was to test this bot out... so that is what I will do :) 2010-10-08T07:43:30 *** jesionaj has joined #aichallenge 2010-10-08T07:44:53 back down from the ledge? 2010-10-08T07:46:53 !echo [rankings], $randomNick(3586) 2010-10-08T07:46:54 Top 10 players: Hazard(3891), felixcoto(3817), rsergio(3769), ofrias(3712), cfaftw(3678), sequoh(3667), davidjliu(3660), iMeller(3659), EBraun(3653), savchenko(3632) , aerique(3586) 2010-10-08T07:47:28 sorry, an americanism. 2010-10-08T07:47:53 goofy way to say you brought me back to my senses. 2010-10-08T07:48:23 me? what did I do? o_O 2010-10-08T07:59:58 Top 10 players: Hazard(3887), felixcoto(3813), rsergio(3765), ofrias(3709), cfaftw(3675), davidjliu(3665), sequoh(3656), iMeller(3651), EBraun(3650), savchenko(3630) 2010-10-08T08:01:24 sigh: what you said reminded me that I want to test my new bot... even with the known bug in it. So, I will let it play before putting in my fix, to see what else I can find. 2010-10-08T08:02:05 ah yup, fair enough 2010-10-08T08:10:47 *** delt0r___ has joined #aichallenge 2010-10-08T08:12:49 *** delt0r_ has quit IRC (Ping timeout: 276 seconds) 2010-10-08T08:20:02 *** krokokrusa has joined #aichallenge 2010-10-08T08:33:11 *** zserge has joined #aichallenge 2010-10-08T08:35:10 *** fawek has quit IRC (Ping timeout: 265 seconds) 2010-10-08T08:35:28 *** jesionaj has quit IRC (Ping timeout: 252 seconds) 2010-10-08T08:37:35 *** flez has joined #aichallenge 2010-10-08T08:37:49 hi all 2010-10-08T08:41:52 *** wvdschel has joined #aichallenge 2010-10-08T08:43:40 flez: hi 2010-10-08T08:45:56 *** JensTi has joined #aichallenge 2010-10-08T08:51:26 *** zserge has quit IRC (Remote host closed the connection) 2010-10-08T08:53:53 *** Arthur_ has quit IRC (Read error: Operation timed out) 2010-10-08T08:55:04 *** hellman has quit IRC (Ping timeout: 240 seconds) 2010-10-08T08:56:48 *** noid41900 has joined #aichallenge 2010-10-08T08:56:51 *** yellowseed has quit IRC (Ping timeout: 240 seconds) 2010-10-08T08:57:10 *** noid41900 is now known as Accoun 2010-10-08T08:57:20 *** Arthur_ has joined #aichallenge 2010-10-08T08:59:59 Top 10 players: Hazard(3891), felixcoto(3820), rsergio(3768), ofrias(3717), cfaftw(3676), davidjliu(3668), iMeller(3661), sequoh(3657), EBraun(3649), savchenko(3624) 2010-10-08T09:01:26 *** brx has joined #aichallenge 2010-10-08T09:03:38 *** sigh has quit IRC (Quit: Leaving...) 2010-10-08T09:14:07 *** mega1 has quit IRC (Ping timeout: 245 seconds) 2010-10-08T09:32:50 *** wvdschel has quit IRC (Read error: Connection reset by peer) 2010-10-08T09:34:30 *** Blkt has joined #aichallenge 2010-10-08T09:35:02 good day everyone 2010-10-08T09:37:08 *** Meatkat has joined #aichallenge 2010-10-08T09:39:09 *** Appleman1234 has quit IRC (Ping timeout: 250 seconds) 2010-10-08T09:44:58 *** Zeiris_ has quit IRC (Ping timeout: 264 seconds) 2010-10-08T09:50:02 *** BtbN has joined #aichallenge 2010-10-08T09:50:46 *** Blkt has quit IRC (Read error: Operation timed out) 2010-10-08T09:53:56 *** flez has quit IRC (Read error: Operation timed out) 2010-10-08T09:58:42 *** JensTi has quit IRC (Quit: Leaving.) 2010-10-08T09:58:59 *** irchs has quit IRC (Quit: irchs) 2010-10-08T09:59:59 Top 10 players: Hazard(3882), felixcoto(3812), rsergio(3756), ofrias(3704), cfaftw(3665), davidjliu(3662), iMeller(3652), sequoh(3647), EBraun(3628), savchenko(3613) 2010-10-08T10:11:07 *** rebelxt has joined #aichallenge 2010-10-08T10:27:36 Lol, my bot doesn't jump higher than necessary: http://72.44.46.68/canvas?game_id=373418 2010-10-08T10:30:08 LazyBot 2010-10-08T10:30:31 *** utyut has left #aichallenge 2010-10-08T10:45:01 *** flez has joined #aichallenge 2010-10-08T10:46:37 *** flez has left #aichallenge 2010-10-08T10:47:21 *** flavor has joined #aichallenge 2010-10-08T10:48:58 # 2010-10-08T10:50:50 *** JamesMG has joined #aichallenge 2010-10-08T10:53:55 *** mceier has joined #aichallenge 2010-10-08T10:57:55 *** Itkovian has quit IRC (Ping timeout: 276 seconds) 2010-10-08T11:00:00 Top 10 players: Hazard(3861), felixcoto(3815), rsergio(3765), ofrias(3714), cfaftw(3671), davidjliu(3663), sequoh(3656), EBraun(3641), savchenko(3619), _iouri_(3614) 2010-10-08T11:26:44 *** aerique has quit IRC (Quit: ...) 2010-10-08T11:27:28 *** hellman has joined #aichallenge 2010-10-08T11:27:40 wow that game was very much depending on the 200 turn limit 2010-10-08T11:27:47 Frontier: ^^ 2010-10-08T11:28:16 as in the strategy would have definitely had to be different without the turn limit 2010-10-08T11:32:02 *** flavor has quit IRC (Read error: Operation timed out) 2010-10-08T11:32:13 jmcarthur: Actually the bot wasn't aware of the turn limit ;) 2010-10-08T11:34:28 *** rebelxt has left #aichallenge 2010-10-08T11:36:24 ha! 2010-10-08T11:36:40 it sure looked like it 2010-10-08T11:36:56 *** cfaftw has joined #aichallenge 2010-10-08T11:37:22 *** Mathnerd314 has joined #aichallenge 2010-10-08T11:44:49 It is strange... it took me quite a long time to write a bot that would beat my old one. Now that it does it performs much worse against other peoples bots. 2010-10-08T11:47:16 You shouldn't do that. Focusing on beating your old bot, I mean. 2010-10-08T11:47:40 Focus on getting better ELO first, *then* worry about beating specific bots. 2010-10-08T11:48:02 Just my take, of course... 2010-10-08T11:48:59 *** JCS^ has joined #aichallenge 2010-10-08T11:50:52 *** zl has quit IRC (Ping timeout: 265 seconds) 2010-10-08T11:55:13 *** bhasker has quit IRC (Quit: bhasker) 2010-10-08T12:00:01 Top 10 players: Hazard(3830), felixcoto(3804), rsergio(3763), ofrias(3712), cfaftw(3675), davidjliu(3656), sequoh(3646), EBraun(3625), savchenko(3621), _iouri_(3610) 2010-10-08T12:12:16 *** McLeopold has joined #aichallenge 2010-10-08T12:19:03 *** yasith_ has joined #aichallenge 2010-10-08T12:19:36 Has anyone played with the tcp server code? 2010-10-08T12:20:38 Feel free to say "no". 2010-10-08T12:20:51 no 2010-10-08T12:20:54 no 2010-10-08T12:20:56 :( 2010-10-08T12:21:00 *** yasith has quit IRC (Ping timeout: 252 seconds) 2010-10-08T12:21:08 I least I know someone is there. :) 2010-10-08T12:21:29 where 2010-10-08T12:21:29 Do you guys know c well? 2010-10-08T12:21:34 some 2010-10-08T12:22:24 I've got an error msg I need help with. 2010-10-08T12:22:39 shoot 2010-10-08T12:23:07 ... 2010-10-08T12:23:43 I'm generating it... 2010-10-08T12:23:57 *** bhasker has joined #aichallenge 2010-10-08T12:24:19 *** Blkt has joined #aichallenge 2010-10-08T12:30:36 rabidus: it's working now o_O 2010-10-08T12:34:03 *** optimum has quit IRC (Remote host closed the connection) 2010-10-08T12:38:30 *** optimum has joined #aichallenge 2010-10-08T12:44:00 how is it that on the official server the report says opponent timed out,but I lost the game??? 2010-10-08T12:44:06 http://ai-contest.com/visualizer.php?game_id=5611511 2010-10-08T12:44:15 both time out? 2010-10-08T12:44:36 http://ai-contest.com/visualizer.php?game_id=5609368 2010-10-08T12:45:01 i don't think i time out, been playing on tcp server and with much larger number of fleets in play my bot doesn't take over 200ms per turn 2010-10-08T12:45:08 these games are terminating with hardly any fleets in play 2010-10-08T12:45:19 if both time out shouldn't it be a draw? 2010-10-08T12:45:20 It looks like they got their logic backwards. 2010-10-08T12:45:42 That means you could write SurrenderBot and win the contest! 2010-10-08T12:46:07 or at least grant it to the person who times out last 2010-10-08T12:46:14 well actually nm 2010-10-08T12:46:18 they should add a ping check 2010-10-08T12:46:28 if a bot is alive it should respond to the ping 2010-10-08T12:46:37 if its going in a loop 2010-10-08T12:47:01 both bots need to send a go to stdout in time 2010-10-08T12:47:05 that should be good enough 2010-10-08T12:47:33 You'd better open an issue. There may be one already. 2010-10-08T12:47:46 btw, ScotBot6 is on tcp 2010-10-08T12:48:27 my current bot is doing terribly 2010-10-08T12:48:28 :( 2010-10-08T12:48:35 i added some crap 2010-10-08T12:48:38 and now its doing worse than before 2010-10-08T12:48:43 didn't you beat ScotBot5? 2010-10-08T12:48:45 yes 2010-10-08T12:48:54 That was my official version. 2010-10-08T12:49:00 on the official server that same one is crashing out 2010-10-08T12:49:03 on every other game 2010-10-08T12:49:07 so my rank is like 800 2010-10-08T12:49:08 :-\ 2010-10-08T12:49:14 :( 2010-10-08T12:49:53 Do you know how to get the results so you can replay it through your bot? 2010-10-08T12:49:59 yea 2010-10-08T12:50:03 will do that let me see 2010-10-08T12:50:56 *** JamesMG has quit IRC (Ping timeout: 260 seconds) 2010-10-08T12:54:52 *** irchs has joined #aichallenge 2010-10-08T12:55:14 just ran it through the bot and my last turn takes under 180ms on my box 2010-10-08T12:55:22 which is a core 2 duo laptop:- 2010-10-08T12:55:27 with full logging enabled 2010-10-08T12:55:33 on server i am not even logging 2010-10-08T12:55:33 *** JamesMG has joined #aichallenge 2010-10-08T12:55:36 so it should be even less:-\ 2010-10-08T12:55:38 if both time out, it will not be draw 2010-10-08T12:55:45 *** Titankiller has joined #aichallenge 2010-10-08T12:55:50 it should be shouldn't it? 2010-10-08T12:55:57 i think server does not check anymore next turn 2010-10-08T12:55:57 i mean neither sent you a go 2010-10-08T12:56:00 can you replay an official game by id using a command line tool? 2010-10-08T12:56:07 yes 2010-10-08T12:56:15 u can stream the game to your bot 2010-10-08T12:56:17 using a script 2010-10-08T12:56:31 which extracts the state out of the visualizer page 2010-10-08T12:56:36 it's possible the load on one of the servers is like 20 again, as it's happened before with some java bots that failed to get killed 2010-10-08T12:56:51 unfortunately i think danielvf is on vacation now 2010-10-08T12:56:54 yea 2010-10-08T12:57:23 http://ai-contest.com/visualizer.php?game_id=5611511 2010-10-08T12:58:03 Any admins here that are able to change affiliations etc? 2010-10-08T12:59:20 If so, can you guys sort this out? :D http://ai-contest.com/forum/viewtopic.php?f=19&t=900 2010-10-08T13:00:02 Top 10 players: felixcoto(3809), Hazard(3804), rsergio(3774), ofrias(3736), cfaftw(3689), davidjliu(3669), sequoh(3652), EBraun(3646), savchenko(3635), _iouri_(3619) 2010-10-08T13:00:25 there is a topic for it, but the updates haven't happened for a while (i'm still waiting) 2010-10-08T13:03:13 *** yasith_ has quit IRC (Ping timeout: 264 seconds) 2010-10-08T13:05:37 *** irchs has quit IRC (Ping timeout: 264 seconds) 2010-10-08T13:15:35 *** yasith_ has joined #aichallenge 2010-10-08T13:23:47 http://72.44.46.68/canvas?game_id=375237 2010-10-08T13:24:01 jdmetz is a good poacher/sniper 2010-10-08T13:24:16 Most of his wins involve waiting a turn and stealing things 2010-10-08T13:24:29 http://ai-contest.com/visualizer.php?game_id=5613381 <= baggy, no any moves no any crash, 1st bot win 2010-10-08T13:25:40 McLeopold: Yeah, I've been working on my counter-sniper code all morning because of that bot 2010-10-08T13:25:42 McLeopold> all top rated botes do it 8-) 2010-10-08T13:26:07 this game mostly: http://72.44.46.68/canvas?game_id=374567 2010-10-08T13:26:12 Accoun: but his waits at the first turn. This is different. 2010-10-08T13:27:03 JamesMG: he didn't get you with a second turn snipe? 2010-10-08T13:27:15 McLeopold> M.B. it result of best move calculation 8-) 2010-10-08T13:27:44 McLeopold> he alweys skeep 1st move? 2010-10-08T13:27:57 Not everything, but enough 2010-10-08T13:27:59 The growth 1 planet top center McLeopold, I think influenced the beginning of the game 2010-10-08T13:28:15 McLeopold> M.B. it result of best move calculation i think 2010-10-08T13:28:36 http://72.44.46.68/canvas?game_id=375139 2010-10-08T13:28:44 C# besl langiage for this contest 2010-10-08T13:28:49 C# best langiage for this contest 2010-10-08T13:28:58 oh really? 2010-10-08T13:29:03 yes 2010-10-08T13:29:14 Logical thinking best programmer for this contest 2010-10-08T13:29:16 i'm still holding out for a REXX compiler 2010-10-08T13:29:32 hmm i wanted to use SNOBOL 2010-10-08T13:29:33 best combination of execution spped an development speed 2010-10-08T13:29:38 Logo, fleets are like moving a turtle around... 2010-10-08T13:30:24 can you have multiple turtles? 2010-10-08T13:30:40 but C# isn't web scale 2010-10-08T13:31:26 the problem with the c# compiler might stand in the way 2010-10-08T13:32:38 I've been leaning heavily on TCP because of the compiler issue 2010-10-08T13:32:40 aany compiler may have a problem 2010-10-08T13:33:05 *** Titankiller has quit IRC (Quit: Nettalk6 - www.ntalk.de) 2010-10-08T13:33:21 java or C# 2010-10-08T13:33:30 C# beter 2010-10-08T13:33:50 in C hard to code and debug 2010-10-08T13:34:09 isn't it an old and slow version of compiler in server? 2010-10-08T13:34:13 izer have low execution speed 2010-10-08T13:34:34 I suppose so, I don't know much about it, I'm using this contest to learn C# 2010-10-08T13:34:42 almost regret that 2010-10-08T13:35:02 Talking about turtles, my bot is called turtle on the TCP server ^^ 2010-10-08T13:35:17 he use linux, M.B. it only have slow version of C# 2010-10-08T13:35:34 java the best in this cause 2010-10-08T13:37:33 Java takes rather long to load and doesn't execute as fast as C++, but I really like the language as such. 2010-10-08T13:37:57 Java needs some of the newer C# features. 2010-10-08T13:37:57 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]) 2010-10-08T13:38:17 What would that be? 2010-10-08T13:38:51 Lambdas and Closures for example 2010-10-08T13:39:24 Never heard of that. Do I need lambdas and closures? 2010-10-08T13:39:53 if you learned how they work, you don't want to write code without using them 2010-10-08T13:40:16 Linq is also a great C# feature/library 2010-10-08T13:40:17 they're pretty neat 2010-10-08T13:40:25 i love using them in python 2010-10-08T13:40:56 writing sql-like querys on collections is just great 2010-10-08T13:41:00 I have no idea what you talk about. I better stick with Delphi, C++ and Java :) 2010-10-08T13:41:26 they're a little more unwieldy in imperative languages than functional or scripting languages 2010-10-08T13:42:04 so 3 languages which don't implement any of the modern features... nice 2010-10-08T13:42:30 I'm a conservative -.- 2010-10-08T13:42:34 i'd like to see more support for esoteric languages, actually 2010-10-08T13:42:58 A brainfuck-AI in top10 would be awesome :D 2010-10-08T13:43:31 Their execution speed could become a problem. Or do native compilers exist? 2010-10-08T13:43:34 brainfuck, whitespace, malbolge 2010-10-08T13:43:36 i would be impressed with a malbolge entry 2010-10-08T13:43:51 well, noone will use malbolge 2010-10-08T13:44:00 that's why i'd be impressed 2010-10-08T13:44:02 :P 2010-10-08T13:44:18 is anyone using genetic algorithms? 2010-10-08T13:44:53 No, but I feel like this is another opportunity to ask: Will I need them? 2010-10-08T13:45:18 "need" is a tricky question in a competition like this 2010-10-08T13:45:19 An asm-programmer could also ask this question to a C programmer... 2010-10-08T13:45:24 Genetic Algorithms sounds obscure. Do they mutate? 2010-10-08T13:46:18 Frontier: it's a form of stochastic beam search 2010-10-08T13:46:22 *** flavor has joined #aichallenge 2010-10-08T13:48:44 beam search? i have absolutely no beams in my code. it works more like a thread from to to bottom. I could imagine that I will write a script that tunes my bots variables over time though. 2010-10-08T13:49:14 myself i like simulated annealing (another form of stochastic beam search), but i don't see offhand how to apply it 2010-10-08T13:50:15 *** jaspervdj has quit IRC (Quit: NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN) 2010-10-08T13:50:55 *** hellman has quit IRC (Remote host closed the connection) 2010-10-08T13:55:51 optimum: Interesting idea. To avoid local minima I should give my values little 'kicks' to see if I it takes me somewhere else. 2010-10-08T13:57:05 BtbN, java almost has closures with inner classes... almost 2010-10-08T13:57:19 personally i want continuations 2010-10-08T13:57:28 and double dispatch 2010-10-08T13:59:20 Yet 70% or so of my bot will always be the algorithm and not the tuning of its variables. And you can't apply anything there than your imagination, trial, error and math. 2010-10-08T13:59:43 C++ has inner classes, too 2010-10-08T14:00:02 Top 10 players: felixcoto(3818), Hazard(3791), rsergio(3782), ofrias(3734), cfaftw(3694), davidjliu(3662), sequoh(3661), EBraun(3648), savchenko(3633), _iouri_(3620) 2010-10-08T14:05:05 Frontier: grid searching is also an option 2010-10-08T14:05:22 *** McLeopold has quit IRC (Quit: Leaving) 2010-10-08T14:07:25 *** RainCT has joined #aichallenge 2010-10-08T14:09:39 delt0r: which programming language do you use with continuations and double dispatch? 2010-10-08T14:14:17 *** Florian95 has quit IRC (Remote host closed the connection) 2010-10-08T14:16:18 check out this sweet base-swap game: http://www.ai-contest.com/visualizer.php?game_id=5613366 2010-10-08T14:18:49 *** krokokrusa has quit IRC (Ping timeout: 264 seconds) 2010-10-08T14:21:58 *** Migi32 has joined #aichallenge 2010-10-08T14:26:00 *** Naktibalda has joined #aichallenge 2010-10-08T14:34:39 i'm not even sure what double dispatch is 2010-10-08T14:40:38 oh, dmj111 you resubmitted :P 2010-10-08T14:41:25 will it beat all the Java bots at the top and restore balance to the universe? 2010-10-08T14:44:13 migi32, his bot lost to a rank 500 bot... soooo we'll see 2010-10-08T14:44:50 that might have just been bad luck, though... 2010-10-08T14:45:30 no it doesn't seem like that 2010-10-08T14:45:31 http://ai-contest.com/visualizer.php?game_id=5596456 2010-10-08T14:45:38 turn 6 was a big mistake 2010-10-08T14:46:13 perhaps a bug? 2010-10-08T14:47:04 yeah, he said earlier that he has actually already corrected that bug on his local machine but is waiting for more games before resubmitting 2010-10-08T14:47:37 *** Arthur_ has quit IRC (Ping timeout: 264 seconds) 2010-10-08T14:48:42 ...why? 2010-10-08T14:49:16 *** JensT1 has joined #aichallenge 2010-10-08T14:49:23 *** Arthur_ has joined #aichallenge 2010-10-08T14:50:12 * Janzert shrugs 2010-10-08T14:50:31 just read it in the archives, it was about 8 hours ago if you want to read for yourself :) 2010-10-08T14:52:07 i like how player 2 stopped firing ships shortly before turn 200 2010-10-08T14:52:09 :P 2010-10-08T14:57:43 this contest is so long 2010-10-08T14:57:51 i think i might be burning out 2010-10-08T14:57:56 :( 2010-10-08T14:58:00 haha 2010-10-08T14:58:18 i think 2 months would have been ideal. 1 month was too short for the tron contest, IMO, but 3 is quite long 2010-10-08T14:58:32 2 months, definitely 2010-10-08T14:58:33 i still prefer 3 over 1 though 2010-10-08T14:58:40 *** irchs has joined #aichallenge 2010-10-08T14:58:44 or maybe i should have just taken more time before entering 2010-10-08T14:59:03 it's also nice that the extended deadline allows people to enter late 2010-10-08T14:59:13 true 2010-10-08T14:59:14 http://ai-contest.com/visualizer.php?game_id=5613381 <= baggy, no any moves no any crash, 1st bot win 2010-10-08T14:59:58 http://ai-contest.com/visualizer.php?game_id=5613381 <= buggy, no any moves no any crash, 1st bot win 2010-10-08T15:00:01 *** Accoun has quit IRC () 2010-10-08T15:00:02 Top 10 players: felixcoto(3831), Hazard(3814), rsergio(3786), ofrias(3736), cfaftw(3691), davidjliu(3675), sequoh(3667), EBraun(3639), savchenko(3635), _iouri_(3630) 2010-10-08T15:00:38 *** perror has quit IRC (Quit: Bye all !) 2010-10-08T15:02:01 *** mega1 has joined #aichallenge 2010-10-08T15:04:20 *** BtbN has quit IRC (Quit: Verlassend) 2010-10-08T15:04:43 http://ai-contest.com/visualizer.php?game_id=5591145 2010-10-08T15:05:00 petrrr jussssst barely takes the lead before turn 200 2010-10-08T15:05:35 *** Utkarsh has quit IRC (Ping timeout: 276 seconds) 2010-10-08T15:06:11 *** choas has joined #aichallenge 2010-10-08T15:06:40 *** agares has joined #aichallenge 2010-10-08T15:06:47 hello 2010-10-08T15:07:11 *** Meatkat has joined #aichallenge 2010-10-08T15:07:40 hello from colombia 2010-10-08T15:07:55 how clever 2010-10-08T15:08:28 hello 2010-10-08T15:09:55 *** Utkarsh has joined #aichallenge 2010-10-08T15:10:30 *** TomyMMX has quit IRC (Ping timeout: 265 seconds) 2010-10-08T15:11:42 sup meatkat 2010-10-08T15:12:35 meatkat, i just linked one of your games: http://ai-contest.com/visualizer.php?game_id=5591145 2010-10-08T15:13:26 *** noid41900 has joined #aichallenge 2010-10-08T15:13:43 *** noid41900 is now known as Accoun 2010-10-08T15:14:11 *** Itkovian has joined #aichallenge 2010-10-08T15:15:37 *** _Utkarsh <_Utkarsh!~admin@117.201.80.79> has joined #aichallenge 2010-10-08T15:18:33 *** nullkuhl__ has joined #aichallenge 2010-10-08T15:19:33 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-10-08T15:21:11 *** nullkuhl_ has quit IRC (Ping timeout: 265 seconds) 2010-10-08T15:22:24 *** TomyMMX has joined #aichallenge 2010-10-08T15:23:39 Any chance the c# compilation issues get resolved soon? 2010-10-08T15:29:40 *** decipher has joined #aichallenge 2010-10-08T15:30:01 hey could anyone help me out with my ai bot, come to the end of debugging and it still randomly crashes 2010-10-08T15:31:07 cfaftw: hey sorry was afk 2010-10-08T15:31:51 cfaftw: yeah that game was a heartbreaker, I lost by like 3 troops 2010-10-08T15:32:04 ridiculous :P 2010-10-08T15:32:17 hey have any of you programmed a bot for the contest 2010-10-08T15:32:18 ? 2010-10-08T15:32:27 most of us have 2010-10-08T15:33:01 Is there a verdict on what is causing the turn 1 losses without an invalid order/timeout message? 2010-10-08T15:33:01 my bot seems to be logically correct but it still crashes is sthere some rules i should know about like max planets you can own max fleets or some kinda over flop that would crash the game mid way 2010-10-08T15:33:02 http://ai-contest.com/visualizer.php?game_id=5618882 2010-10-08T15:33:08 or am i just making a bad move 2010-10-08T15:33:11 cfaftw: on the plus side I'm ranked in the top100 now! or at least I was a couple seconds ago 2010-10-08T15:33:24 meatkat, i saw that... congrats! 2010-10-08T15:33:32 decipher: do you get Invalid order errors in TCP server? 2010-10-08T15:33:36 my bot makes 20 sometimes 40 50 moves then the board dissapears 2010-10-08T15:33:47 just says player 2 time out 2010-10-08T15:33:59 decipher, do you ever say "go"? 2010-10-08T15:34:04 what go? 2010-10-08T15:34:06 lol 2010-10-08T15:34:23 should be an invalid order 2010-10-08T15:34:38 i say finish turn at the end of my doturn method just incase it doeing to many moves 2010-10-08T15:34:39 JamesMG: yeah it's frustrating, I haven't heard anything yet 2010-10-08T15:34:42 log all orders and check what you did wrong at the last turn 2010-10-08T15:34:51 JamesMG: perhaps the server doesn't have the error reporting code on it :/ 2010-10-08T15:34:54 cfaftw: thanks! 2010-10-08T15:35:06 does a bad move crash teh game? 2010-10-08T15:35:14 like senidng more ships then a player owns? 2010-10-08T15:35:14 decipher, in my bot, i modified the "issue_order" method to perform a legality check 2010-10-08T15:35:20 kk 2010-10-08T15:35:20 decipher: yes 2010-10-08T15:35:34 so your saying that 2010-10-08T15:35:35 decipher, a bad move is considered a forfeit (loss) for your bot 2010-10-08T15:35:39 ok 2010-10-08T15:35:46 the game will end immediately 2010-10-08T15:35:52 it's not a crash 2010-10-08T15:35:57 *** jesionaj has joined #aichallenge 2010-10-08T15:36:00 so lets say i do a game tree.. and by teh time my game tree ends mys tate changes and becasue the bot attacked me i didnt send teh wriet amount of ships 2010-10-08T15:36:04 ok 2010-10-08T15:36:06 sorry not a crash 2010-10-08T15:36:22 sorry for the awful typing 2010-10-08T15:36:37 so yeah my bot looses because i send to many ships because the game state changes? 2010-10-08T15:36:39 you make up for it by your use of "teh" 2010-10-08T15:36:46 lol 2010-10-08T15:36:46 decipher: the game state shouldn't be changing in your code unless you changed it on your side 2010-10-08T15:36:52 mid turn that is 2010-10-08T15:36:53 yeah 2010-10-08T15:37:17 i mean like the game state i receive and the actual game state playing, i make a move but stuff has changed 2010-10-08T15:37:43 like why would i make a wrong move i ensure my bot knows how many ships i have? 2010-10-08T15:37:45 stuff doesn't change until the next turn 2010-10-08T15:37:49 kk 2010-10-08T15:37:53 so why does it fail 2010-10-08T15:37:55 decipher: bug in your code :) 2010-10-08T15:37:58 lol 2010-10-08T15:38:04 so your saying i should 2010-10-08T15:38:06 you might be altering the current turn's ACTUAL game state 2010-10-08T15:38:09 just have a legality check 2010-10-08T15:38:09 it's allways a reason 2010-10-08T15:38:15 to ensure my bug doesnt complicate stuff 2010-10-08T15:38:22 so your bot thinks a planet has X ships, when really it has Y ships 2010-10-08T15:38:33 why would that happen 2010-10-08T15:38:41 that's for you to find out 2010-10-08T15:38:52 pass by reference vs by value errors maybe? What language? 2010-10-08T15:39:01 java 2010-10-08T15:39:15 for example it happens, if you send 2 fleets from the same planet but don't update a number of ships on planet 2010-10-08T15:39:50 but each turn a pw is passed in and i parse it to a string array to give me a map state 2010-10-08T15:40:13 you won't solve this issue by guessing, log orders and compare last orders with number of ships in planets 2010-10-08T15:40:15 -_- 2010-10-08T15:40:19 test it in TCP server 2010-10-08T15:40:31 testing is a pain in teh ass lol 2010-10-08T15:40:35 i dont know what turn it ends on 2010-10-08T15:40:39 so i have to guess and hope 2010-10-08T15:40:44 log all turns 2010-10-08T15:40:47 that i catch the right state 2010-10-08T15:40:48 yeah 2010-10-08T15:40:51 yeah you can keep track of turns 2010-10-08T15:41:01 turn += 1 2010-10-08T15:41:03 i guess i need to add a counter to my file so it adds more files each time it gest that 2010-10-08T15:41:07 there* 2010-10-08T15:41:16 *** _Utkarsh <_Utkarsh!~admin@117.201.80.79> has quit IRC (Ping timeout: 260 seconds) 2010-10-08T15:41:17 i have a turn global counter 2010-10-08T15:42:23 any chance i can have ur new issue order method? 2010-10-08T15:42:31 actually nvm all i do is check ships right? 2010-10-08T15:43:29 yep 2010-10-08T15:43:33 umm 2010-10-08T15:43:37 so if your move is quicker then a second 2010-10-08T15:43:52 the game waits till the second to make both ur opponent and ur move so ur together? 2010-10-08T15:44:02 yes 2010-10-08T15:44:04 kk 2010-10-08T15:44:15 if you go over a second, your bot forfeits and the game ends immediately 2010-10-08T15:44:22 no 1 second is max 2010-10-08T15:44:31 man, even after you have picked targets, deciding how to create optimal orders for those targets can be tough. 2010-10-08T15:44:42 under one second, the game doesn't progress until both players have said "go" 2010-10-08T15:44:43 i store my score as i go 2010-10-08T15:44:48 if both bots makes orders faster, engine starts another turn imediatelly 2010-10-08T15:44:54 when do you say go? 2010-10-08T15:45:03 jmcarthur, yep. 2010-10-08T15:45:04 i didnt see that under pw 2010-10-08T15:45:05 decipher: if you and your opponent both say "go" before the 1 second is up, the engine goes ahead and advances to the next turn 2010-10-08T15:45:12 oh ok 2010-10-08T15:45:16 not necessary 2010-10-08T15:45:18 *** TomyMMX has quit IRC (Ping timeout: 265 seconds) 2010-10-08T15:45:20 :) 2010-10-08T15:45:29 if do turn is done or pw.finishturn is executed its teh same thing 2010-10-08T15:45:31 jmcarthur, have you submitted a bot yet? or do you play on the TCP server? 2010-10-08T15:46:14 cfaftw: still not yet 2010-10-08T15:46:16 decipher: finishturn sends go command 2010-10-08T15:46:24 kk 2010-10-08T15:46:30 jmcarthur, how long have you been working on your bot? 2010-10-08T15:46:40 a month :) 2010-10-08T15:46:50 cfaftw: for a long time, but it's not a large total hours spent on it 2010-10-08T15:47:08 *** Utkarsh has joined #aichallenge 2010-10-08T15:47:55 has anybody tried training a neural net for this 2010-10-08T15:48:01 can you read in values froma file 2010-10-08T15:48:33 decipher: i think you can 2010-10-08T15:48:37 you could train using the games on the official website 2010-10-08T15:49:25 you could create a supervised multi layer neural net that uses backpropagation to train but i dont have that kinda time ^^ lol 2010-10-08T15:50:04 and i don't have that kinda skill 2010-10-08T15:50:12 it take no time to make a move it already be trained but if a new map was presented i dont know hwo it handle that 2010-10-08T15:50:27 i guess thats where the 95 percent memorization comes from 2010-10-08T15:53:31 wenas 2010-10-08T15:53:35 one last question, where it issue the order in planetwars class it just system.out.prints then system.flush is that how it executes so if i just put that in a loop with if it will check it 2010-10-08T15:53:40 somebody speak spanish? 2010-10-08T15:54:48 *** jmpespxoreax has left #aichallenge 2010-10-08T15:55:23 agares: un poco 2010-10-08T15:56:20 hola mega1 2010-10-08T15:56:32 hola 2010-10-08T15:57:34 still crashes with this 2010-10-08T15:57:36 if(source.NumShips() > numShips){ System.out.println("" + source.PlanetID() + " " + dest.PlanetID() + " " + numShips); System.out.flush(); } } 2010-10-08T15:59:05 crashes or times out? 2010-10-08T15:59:08 did you implement logging? 2010-10-08T15:59:11 maybe 2010-10-08T15:59:15 invalid order :) 2010-10-08T15:59:17 LOL 2010-10-08T15:59:20 if it's timing out then you might not be sending "go" like you should 2010-10-08T15:59:28 "did you implement logging?" ... "maybe." 2010-10-08T15:59:35 im saying pw.finishTurn() at teh end of method 2010-10-08T15:59:38 ah 2010-10-08T16:00:03 Top 10 players: felixcoto(3844), Hazard(3822), rsergio(3777), ofrias(3741), cfaftw(3710), davidjliu(3680), sequoh(3672), savchenko(3651), _iouri_(3645), EBraun(3637) 2010-10-08T16:00:14 its impossible to be sending invalid move with that condition 2010-10-08T16:00:19 so it must be timing out 2010-10-08T16:00:25 grr 2010-10-08T16:00:26 it's still possible 2010-10-08T16:00:45 decipher, just time how long your turns take 2010-10-08T16:00:45 how? sending planets to ur self? 2010-10-08T16:00:50 i do usually 2010-10-08T16:00:54 standard issueOrder() method doesn't update numships 2010-10-08T16:00:55 print out the elapsed time at the end of the turn 2010-10-08T16:00:59 decipher: are you perhaps sending more than one fleet from the source planet? maybe the sum of those fleets is greater than the current ships at the planet? 2010-10-08T16:01:06 but i dont know what turn it ends on i need to log every turn and every tree branch state per turn 2010-10-08T16:01:12 decipher: if src == dst then it crashes too 2010-10-08T16:01:17 also... if you are printing this stuff to stdout, it's going think it's an invalid move 2010-10-08T16:01:21 decipher: yes, do that 2010-10-08T16:01:25 you need to debug to a file or something 2010-10-08T16:01:34 yeah 2010-10-08T16:02:14 i gotta head to work but tonight ill make a pure debug method this shouldnt be crashes this hard but maybe i'm checking to see if my fleet is less then my num of ships but im sending multiple fleets causing it still to crash? 2010-10-08T16:02:24 agares: eso ha sido todo? :-) 2010-10-08T16:03:14 hola mega1, es que estoy ocupado en el trabajo, pero queria probar el canal :D 2010-10-08T16:03:27 gracias por la respuesta... :D nos vemos en el concurso ;-) 2010-10-08T16:04:49 muy bien, esperare aqui haste el senyor regrese ;-) 2010-10-08T16:04:55 *hasta 2010-10-08T16:06:11 *** Palmik has quit IRC (Remote host closed the connection) 2010-10-08T16:07:03 *** FlagCapper has quit IRC (Ping timeout: 265 seconds) 2010-10-08T16:07:34 *** irchs has quit IRC (Quit: irchs) 2010-10-08T16:10:26 *** agares has quit IRC (Ping timeout: 265 seconds) 2010-10-08T16:11:16 *** delt0r_ has joined #aichallenge 2010-10-08T16:12:12 *** irchs has joined #aichallenge 2010-10-08T16:12:57 *** delt0r___ has quit IRC (Ping timeout: 255 seconds) 2010-10-08T16:16:24 *** irchs has quit IRC (Client Quit) 2010-10-08T16:17:31 *** agweber has joined #aichallenge 2010-10-08T16:18:26 *** nullkuhl__ has quit IRC (Remote host closed the connection) 2010-10-08T16:19:51 So I can't figure out why my code is now crashing. There's no errors in the code, and I can't find what else is wrong. I just added this chunk; http://pastebin.com/nU4c4fyc any ideas? It works fine if I comment it out. 2010-10-08T16:20:03 It just fails out on turn 1 2010-10-08T16:20:32 *** jmpespxoreax has joined #aichallenge 2010-10-08T16:23:24 what if you have no planets? 2010-10-08T16:23:31 then mine == null 2010-10-08T16:23:32 IssueOrder takes ints or Planets? 2010-10-08T16:23:54 both, I think 2010-10-08T16:24:00 there are two IssueOrder 2010-10-08T16:24:02 ah 2010-10-08T16:24:26 *** decipher has quit IRC (Quit: Page closed) 2010-10-08T16:24:59 Yeah, there are two 2010-10-08T16:25:09 This is for the very first turn, why wouldn't I have a planet? 2010-10-08T16:26:00 just speculating 2010-10-08T16:26:20 agweber: you might be sending too many ships 2010-10-08T16:26:40 Might I? Hmm.. 2010-10-08T16:27:10 yeah its pretty hard to figure out what's wrong with your bot in the current viewer so I made my own 2010-10-08T16:27:35 if some neutrals have the same number of ships, you increment count multiple times and go over 100 but don't do anything about that 2010-10-08T16:28:03 Ah... quite possibly 2010-10-08T16:28:13 (you also might increment count multiple times per planet, but that's probably unrelated) 2010-10-08T16:28:17 *** Palmik has joined #aichallenge 2010-10-08T16:29:41 oooh, you probably are sending multiple ships 2010-10-08T16:29:58 you're trying to calculate how many ships to send to the nearest planets 2010-10-08T16:30:08 Yeahh.. I need some equals instead of >'s and <'s 2010-10-08T16:30:18 then trying to conquer every neutral planet with less than i ships 2010-10-08T16:30:42 not just your closest neighbors 2010-10-08T16:30:53 s/multiple/too many/ 2010-10-08T16:31:33 I'm trying to find out how many neutral planets I can conquer within a certain radius for 100 ships, and then sending out enough ships to do that. 2010-10-08T16:31:45 *** TomyMMX has joined #aichallenge 2010-10-08T16:31:48 but you're ignoring the radius when you send those ships :) 2010-10-08T16:31:54 any news about the c# problems 2010-10-08T16:31:57 Oh my! 2010-10-08T16:32:05 I've overlooked so many factors 2010-10-08T16:35:27 I must save this small chitchat and get ready for work, thanks for the help Mr. Zan 2010-10-08T16:36:03 *** agweber has quit IRC (Quit: Page closed) 2010-10-08T16:36:32 *** TomyMMX has quit IRC (Ping timeout: 265 seconds) 2010-10-08T16:39:02 *** Zeiris_ has joined #aichallenge 2010-10-08T16:41:43 *** cfaftw has quit IRC (Quit: Leaving) 2010-10-08T16:55:39 *** Meatkat has quit IRC () 2010-10-08T17:00:04 Top 10 players: felixcoto(3835), rsergio(3799), Hazard(3794), ofrias(3746), cfaftw(3700), davidjliu(3689), sequoh(3676), _iouri_(3668), savchenko(3644), EBraun(3640) 2010-10-08T17:21:26 *** jimmerton has joined #aichallenge 2010-10-08T17:30:42 *** Zeiris_ has quit IRC (Ping timeout: 250 seconds) 2010-10-08T17:42:38 *** mega1 has quit IRC (Ping timeout: 252 seconds) 2010-10-08T17:43:21 *** RainCT has quit IRC (Remote host closed the connection) 2010-10-08T17:47:32 geez 2010-10-08T17:47:53 i didn't know that there are so much combinations on diffrent planets 2010-10-08T17:48:53 my 2 GHz computer calculates 16 planet combinations 0.7 seconds 2010-10-08T17:49:17 only lists them, without outputting them 2010-10-08T17:49:33 with what language? 2010-10-08T17:49:37 c++ 2010-10-08T17:50:31 and, intresting part is that php can do that same. 2010-10-08T17:50:40 only 20% slower 2010-10-08T17:50:49 16 combinations? what exactly do you mean? 2010-10-08T17:50:58 all subsets of the 16 planets? 2010-10-08T17:51:28 umm.. like with three planets you can have 6 combinations. 1,2,3,(1,2),(2,3),(1,2,3) 2010-10-08T17:51:39 four planets is twice as big as that 2010-10-08T17:51:44 okay, yeah, all subsets 2010-10-08T17:51:45 and so on. 2010-10-08T17:51:48 yeah 2010-10-08T17:52:08 yeah, that's 2^16 2010-10-08T17:52:12 +1 2010-10-08T17:52:21 also there are 8 combinations for 3 planets 2010-10-08T17:52:27 (2^n)+1 2010-10-08T17:52:27 none and (1,3) 2010-10-08T17:52:39 uh... 2010-10-08T17:52:52 it depends what are you calculating 2010-10-08T17:53:12 well, a set of size S has 2^S subsets. 2010-10-08T17:53:18 yeah 2010-10-08T17:53:21 but yes, it depends exactly what you're looking for 2010-10-08T17:53:58 yep. i'm trying to just check on my first turn that which planet set is best choise 2010-10-08T17:54:06 for ... productivity.. 2010-10-08T17:54:13 i don't know the word 2010-10-08T17:55:21 ie. 'how many ships you can gain in x turns from these y sets of planets' 2010-10-08T17:55:34 * Zannick nod 2010-10-08T17:55:42 nod? 2010-10-08T17:55:56 i understand what you mean 2010-10-08T17:56:01 k. 2010-10-08T17:56:31 i didn't know that that algorithm is that slow 2010-10-08T17:56:40 and that kinda sucks :D 2010-10-08T17:57:01 or i'm just fukken bad coder, which is so true 2010-10-08T17:57:03 you could get some good speedup by cutting out planets 2010-10-08T17:57:14 yep, that's true 2010-10-08T17:57:14 byt 2010-10-08T17:57:16 ut 2010-10-08T17:57:19 *but 2010-10-08T17:57:31 if this is your first turn, you should rule out planets closer to your opponent 2010-10-08T17:57:40 that's half 2010-10-08T17:58:16 *** Error323 has quit IRC (Read error: Connection reset by peer) 2010-10-08T17:58:44 true. still there might be at least 16 intresting planets, PLUS those what are on opponent's side 2010-10-08T17:58:49 *** fawek has joined #aichallenge 2010-10-08T17:58:52 and reducing the running time from 2^16 iterations to 2^8 is much much much much better than half 2010-10-08T17:59:04 no 2010-10-08T17:59:10 2^32 to 2^16 2010-10-08T17:59:31 i thinkg 32 is max 2010-10-08T17:59:33 either way 2010-10-08T17:59:37 i don't know tho 2010-10-08T17:59:39 atm there are 23 planets per map 2010-10-08T17:59:49 not fixed in stone, though 2010-10-08T17:59:53 but that function MUST procedure atleast 16 planets 2010-10-08T17:59:56 *** flag has joined #aichallenge 2010-10-08T18:00:03 prefer for 100 planets :( 2010-10-08T18:00:04 Top 10 players: felixcoto(3826), rsergio(3816), Hazard(3799), ofrias(3735), cfaftw(3703), davidjliu(3690), sequoh(3667), _iouri_(3654), EBraun(3640), adumlah(3631) 2010-10-08T18:00:35 cause that is not very special algorithm 2010-10-08T18:00:59 it just lists a set of planets 2010-10-08T18:01:01 nothing special 2010-10-08T18:01:10 calculations are just coming for those 2010-10-08T18:02:29 Are the ranks on the tcp server based on all the games you've played or just the x most recent ones? 2010-10-08T18:02:29 *** mceier has quit IRC (Quit: leaving) 2010-10-08T18:02:55 x most recent where x is like 2000 2010-10-08T18:03:08 it expires after some time interval i think 2010-10-08T18:03:11 er, games do 2010-10-08T18:03:37 someone said 24 hours but I don't know if that is accurate 2010-10-08T18:03:46 It seems like it's 24 hours 2010-10-08T18:03:53 that'd make sense 2010-10-08T18:04:09 I'm just wondering if I need to use a new name to see how good my new bot is or just wait a little while 2010-10-08T18:04:26 everybody's doing it 2010-10-08T18:04:41 true 2010-10-08T18:05:36 do you use planet combinations? 2010-10-08T18:06:05 or subsets, as Zannick use that word 2010-10-08T18:06:16 i actually don't know what is the correct word for that 2010-10-08T18:08:12 I think I need to write that soon, mine judges them individually still 2010-10-08T18:08:17 makes for some dumb choices sometimes 2010-10-08T18:10:23 *** optimum has quit IRC (Remote host closed the connection) 2010-10-08T18:10:49 *** optimum has joined #aichallenge 2010-10-08T18:13:35 *** fawek has quit IRC (Ping timeout: 252 seconds) 2010-10-08T18:18:32 *** sigh has joined #aichallenge 2010-10-08T18:19:15 morning 2010-10-08T18:19:28 afternoon 2010-10-08T18:19:30 00:11 here 2010-10-08T18:19:43 Blkt: that's "morning" :P 2010-10-08T18:19:51 :) 2010-10-08T18:20:00 :D 2010-10-08T18:23:27 really, 0.7 seconds? that still seems too long for only 65536 subsets 2010-10-08T18:24:21 computing the powerset of 16 elements on my machine happens too quickly for me to notice that it's not instantaneous 2010-10-08T18:28:43 *** yellowseed has joined #aichallenge 2010-10-08T18:29:53 he's probably doing something that takes on average 0.7/65536 seconds per subset 2010-10-08T18:31:16 ah 2010-10-08T18:31:20 probably true 2010-10-08T18:33:46 unusual graph http://72.44.46.68/canvas?game_id=373593 2010-10-08T18:36:38 what's unusual? 2010-10-08T18:37:25 *** Bobng has quit IRC (Ping timeout: 265 seconds) 2010-10-08T18:37:37 it curves 2010-10-08T18:37:46 most graphs go straight or have jaggies 2010-10-08T18:37:52 jdmetz has got a bug in attack code, sent enough ships to reduce bottom left to 0 units, but not enough to take over 2010-10-08T18:38:15 heh 2010-10-08T18:40:08 *** Appleman1234 has joined #aichallenge 2010-10-08T18:40:49 jmcarthur: how is your bot coming along? 2010-10-08T18:41:31 when can we see it :D 2010-10-08T18:47:43 that everybody keeps asking that makes me very afraid of having a crappy bot 2010-10-08T18:48:16 ha 2010-10-08T18:48:27 my strategy is to never talk about how my bot is coming along :P 2010-10-08T18:48:39 i mean 2010-10-08T18:48:42 i'm not writing a bot >.> 2010-10-08T18:53:07 haha yeah speaking of which.. 2010-10-08T18:53:13 not that i have anything to show for my efforts either 2010-10-08T18:53:34 well, i guess i do at least have what i threw together and it turns out to at least be somewhat respectable 2010-10-08T18:54:09 *** Zannick has quit IRC (Read error: Operation timed out) 2010-10-08T18:54:09 *** JensT1 has quit IRC (Ping timeout: 252 seconds) 2010-10-08T18:54:14 *** Zannick has joined #aichallenge 2010-10-08T18:56:09 *** jimmerton has quit IRC (Quit: jimmerton) 2010-10-08T18:58:01 *** JamesMG has quit IRC (Quit: home) 2010-10-08T18:58:46 *** amstan_ has joined #aichallenge 2010-10-08T18:58:46 *** ChanServ sets mode: +o amstan_ 2010-10-08T18:59:03 *** davidjliu has joined #aichallenge 2010-10-08T19:00:05 Top 10 players: felixcoto(3874), rsergio(3847), Hazard(3793), ofrias(3736), cfaftw(3722), davidjliu(3703), sequoh(3678), EBraun(3655), _iouri_(3650), savchenko(3645) 2010-10-08T19:00:48 *** Blkt has quit IRC (Ping timeout: 255 seconds) 2010-10-08T19:07:57 *** amstan_ has quit IRC (Remote host closed the connection) 2010-10-08T19:10:15 noobs 2010-10-08T19:10:16 *** JensT1 has joined #aichallenge 2010-10-08T19:15:09 who are noobs? 2010-10-08T19:20:20 *** McLeopold has joined #aichallenge 2010-10-08T19:20:29 CrowBot! http://72.44.46.68/canvas?game_id=379282 2010-10-08T19:20:55 everyone! 2010-10-08T19:21:00 but felixcoto 2010-10-08T19:21:27 ? 2010-10-08T19:21:55 what did i just watch 2010-10-08T19:22:19 um, a planet wars replay? 2010-10-08T19:23:20 what is crowbot's objective? i just saw a pretty well played standoff followed by an explosion of activity 2010-10-08T19:23:32 steal planets 2010-10-08T19:23:54 that last one was just cool, i'll find a better example 2010-10-08T19:24:03 aha! 2010-10-08T19:25:14 i wonder what McLeopold.RageBot.py would be like 2010-10-08T19:25:43 can someone look at http://72.44.46.68/canvas?game_id=379983 2010-10-08T19:25:54 at turn 35/36 to be precise 2010-10-08T19:25:55 http://72.44.46.68/canvas?game_id=379307 2010-10-08T19:26:48 actually never mind 2010-10-08T19:26:51 bhasker: defence fail? 2010-10-08T19:26:52 i launched the ship a turn late 2010-10-08T19:26:55 yea 2010-10-08T19:26:56 :-\ 2010-10-08T19:27:06 looks like i have a bug 2010-10-08T19:27:11 :-\ 2010-10-08T19:27:16 it looks like a planet got taken that shouldn't? 2010-10-08T19:27:31 yea i am not sure 2010-10-08T19:27:42 it looks like a 6 fleet of mine and the 14 hit the same turn 2010-10-08T19:27:52 or am i 1 turn behind 2010-10-08T19:27:57 McLeopold: not what i meant :P 2010-10-08T19:27:57 oh, one turn late. It's hard to tell distances with the viewer 2010-10-08T19:28:10 yea 1 turn late 2010-10-08T19:28:11 hmm 2010-10-08T19:28:19 miscalculation nm 2010-10-08T19:28:20 yeah, off by one error 2010-10-08T19:28:33 shouldn't be a hard fix :( 2010-10-08T19:28:34 :) 2010-10-08T19:28:41 hopefully 2010-10-08T19:29:13 okay, here it is http://72.44.46.68/canvas?game_id=379630 2010-10-08T19:29:26 CrowBot in action with what it was designed to do. 2010-10-08T19:29:56 yup, this is high-level ragebot 2010-10-08T19:29:58 very nice 2010-10-08T19:30:03 maybe only the first 10 moves are a good example... 2010-10-08T19:30:19 how does it select planets? 2010-10-08T19:30:56 http://72.44.46.68/canvas?game_id=379933, kinda better 2010-10-08T19:31:51 I'm just not getting the close start maps. 2010-10-08T19:32:08 It's suppose to wait one turn, then take everything the opponent wants. 2010-10-08T19:32:20 Except it will build normal if it feels behind. 2010-10-08T19:32:31 Modified WormBot. 2010-10-08T19:32:55 *** Accoun_ has joined #aichallenge 2010-10-08T19:33:18 cool 2010-10-08T19:34:21 The weird thing is, it just sits there if it is ahead, not doing anything, and usually wins? 2010-10-08T19:34:24 o_O 2010-10-08T19:35:05 *** Accoun has quit IRC (Ping timeout: 260 seconds) 2010-10-08T19:36:03 like this: http://72.44.46.68/canvas?game_id=380191 2010-10-08T19:36:38 i'm amazed at how many bots (even top ones!) will resort to inaction when they are not winning 2010-10-08T19:37:23 jmcarthur: my bot doesn't know anything about who's ahead or behind :) 2010-10-08T19:37:34 neither did dmj111's or davidjliu's last time I spoke to them 2010-10-08T19:37:46 mine does nothing if it's winning 2010-10-08T19:38:13 Okay, here is the best example so far: http://72.44.46.68/canvas?game_id=380218 2010-10-08T19:38:23 it's "obvious" to everyone that the tactics should be different, but not so obvious how 2010-10-08T19:38:46 sigh: i'm sure it's not an explicit concept, but surely most reasonable heuristics are affected indirectly by the situation 2010-10-08T19:39:12 my official one doesn't. I think you're going to need somthing. 2010-10-08T19:39:46 McLeopold: in that last replay your bot does nothing when oldman2 takes the last neutral 2010-10-08T19:39:50 jmcarthur: sure, what I mean is that the bots don't use the ship count difference or production difference in the code at all 2010-10-08T19:39:59 it could have easily intercepted 2010-10-08T19:40:00 I know, it just doesn't care. 2010-10-08T19:40:52 sigh: yeah, i'm sure of that 2010-10-08T19:41:17 sigh: having special code for different scenarios is more work and ad hoc 2010-10-08T19:41:22 I tried a few meta-strategies that alter the bot's actions based on it's evaluation of the overall game, but haven't been able to get anything good 2010-10-08T19:41:43 sigh: i'm just surprised that so many heuristics aren't even showing any emergent behavior for such situations 2010-10-08T19:41:56 jmcarthur: also, many of the top bots err on the side of defence... which means inaction when outnumbered 2010-10-08T19:42:01 nrub_: It probably figured the 2 planets next to it might defend 2010-10-08T19:42:19 dmj111 is very defensive, for example 2010-10-08T19:42:24 yeah i've noticed 2010-10-08T19:42:33 jmcarthur: what types of behavior? 2010-10-08T19:43:02 McLeopold: something more aggressive when losing rather than just idling 2010-10-08T19:43:23 http://72.44.46.68/canvas?game_id=380277 2010-10-08T19:43:43 another good one, and by good, I mean something I wouldn't try and program, but worked. 2010-10-08T19:43:53 two things, 1) it is hard to figure out what to do, and 2) does it matter? are you too far behind to do anything in that situation? 2010-10-08T19:44:02 ^^ @jmcarthur 2010-10-08T19:44:23 sigh: i often see easy expansions that are not being taken 2010-10-08T19:44:26 sigh: it's not too hard. if you're behind, take a planet 2010-10-08T19:45:09 and i'm talking about with >100 turns left, plenty of time to make up for the losses of taking it 2010-10-08T19:45:18 figuring out if your behind is tricky in itself... if it is very obvious you are probabaly too far behind 2010-10-08T19:45:42 CrowBot changes everything in my mind. 2010-10-08T19:45:45 well, at least for me, I haven't had much success :) 2010-10-08T19:45:55 McLeopold: what does it change? 2010-10-08T19:45:57 I never thought it would work. 2010-10-08T19:46:06 you think bots should be more reactive? 2010-10-08T19:46:06 I thought it would be like MoldBot. 2010-10-08T19:46:22 what did moldbot do? 2010-10-08T19:46:22 It's not securing a border... 2010-10-08T19:46:29 It's not trying to grow faster... 2010-10-08T19:46:43 It's not forwarding to "correct" spots... 2010-10-08T19:47:00 Sometimes it doesn't forward... 2010-10-08T19:47:29 It feels like tit-for-tat 2010-10-08T19:47:47 interesting 2010-10-08T19:47:48 yeah, it's like copybot but smarter 2010-10-08T19:48:07 It's not trying to finish off the opponent... 2010-10-08T19:48:10 honestly, we never did figure out what makes ragebot so tough when you're starting out 2010-10-08T19:48:25 jmcarthur: easy, starter paks do no defence 2010-10-08T19:48:26 big fleets? 2010-10-08T19:48:31 *packs 2010-10-08T19:48:54 once you implement any sort of defence, ragebot is thwarted 2010-10-08T19:49:09 *** amstan has quit IRC (Ping timeout: 252 seconds) 2010-10-08T19:49:12 then it only does well on close maps, in which case you tighten up your defence code 2010-10-08T19:49:55 Okay, I'm going to try CrowBot2 2010-10-08T19:50:22 McLeopold: look forward to it :) 2010-10-08T19:50:22 dun dun dunnnn 2010-10-08T19:50:32 jmcarthur: does my reasoning make sense? 2010-10-08T19:50:40 sigh: yeah makes sense 2010-10-08T19:51:02 sigh: but why ragebot specifically? is it just because it's so aggressive? 2010-10-08T19:51:04 hey guys, i have a short break between projects, anyone need any AI help or feedback? 2010-10-08T19:51:09 compared to the other starters 2010-10-08T19:51:23 it always attacks the opponent, not neutrals? 2010-10-08T19:51:32 yeah, the other starters attack neutrals as well 2010-10-08T19:51:57 http://72.44.46.68/canvas?game_id=380388 2010-10-08T19:52:47 * Naktibalda needs a stronger opponent bot for local testing 2010-10-08T19:52:54 bhasker: woah, explosion 2010-10-08T19:53:01 yea his bot went nuts 2010-10-08T19:53:03 suddenly 2010-10-08T19:53:05 red was silly :P 2010-10-08T19:53:12 So, what you guys think about me releasing WormBot and CrowBot? 2010-10-08T19:53:28 i think people in the forums hated when someone released a top 50 c# bot 2010-10-08T19:53:36 since that makes it harder for newcomers 2010-10-08T19:53:37 There was already some backlash for someone else. 2010-10-08T19:53:41 yup 2010-10-08T19:53:45 bhasker: it mist have decided that taking lots of little planets was good because higher overall growth 2010-10-08T19:53:51 McLeopold: personally, I'm all for it 2010-10-08T19:53:51 yea 2010-10-08T19:54:12 man, i remember when Max was in first 2010-10-08T19:54:17 he's in 484 now 2010-10-08T19:54:25 Those bots couldn't win. But they help look for weaknesses against bot types. 2010-10-08T19:54:53 was that before or after a1k0n? 2010-10-08T19:54:58 before 2010-10-08T19:55:03 personally, I'm for anything that will raise the level of the competition 2010-10-08T19:55:12 but others don't agree, it seems :( 2010-10-08T19:55:17 i don't think releasing the code for a top 50 bot increases competition 2010-10-08T19:55:18 There was some good arguments against. 2010-10-08T19:55:38 all that does is raise the bar above where some people can continue to compete 2010-10-08T19:55:44 You don't want everyone just submitting your code. 2010-10-08T19:55:56 lamo 2010-10-08T19:55:58 need to fix this 2010-10-08T19:55:58 http://72.44.46.68/canvas?game_id=380437 2010-10-08T19:56:07 *** JensT1 has quit IRC (Quit: Leaving.) 2010-10-08T19:56:25 what mcleopold said 2010-10-08T19:56:27 yeah, if there are some groups that can't program a top 50 bot, they may still want to compete with each other 2010-10-08T19:56:30 discuss the strategy 2010-10-08T19:56:36 but don't release a fully functional bot 2010-10-08T19:56:36 it's like having a pole vaulting contest with 3000 people but assuming that everybody can manage at least 15 feet 2010-10-08T19:56:41 let someone work at the code at least 2010-10-08T19:56:46 If they just copy a bot they can't understand, what's the point? 2010-10-08T19:56:48 *** amstan has joined #aichallenge 2010-10-08T19:56:56 *** ChanServ sets mode: +o amstan 2010-10-08T19:56:56 it makes it pointless for all the people working to get a better bot 2010-10-08T19:57:15 knowing a strategy and implementing it correctly are two different things 2010-10-08T19:57:22 i am all for discussing strategy 2010-10-08T19:57:29 but not for releasing fully functional bots 2010-10-08T19:57:38 heck, knowing of a strategy and understanding it are different things 2010-10-08T19:57:45 hehe 2010-10-08T19:57:53 I want to release so people can test against it. I might just have to fallback to tcp server. 2010-10-08T19:57:54 Plus anyone who had to work to get to where your bot is currently has their work kind of invalidated 2010-10-08T19:57:58 But not everyone is using it. 2010-10-08T19:58:03 Because someone just joining is there already 2010-10-08T19:58:10 bhasker, releasing bots perpetuates understanding, which makes everyone better 2010-10-08T19:58:26 dude i totally heard about alpha beta minimax, so my bot should be at least that good just because i heard about it, right? 2010-10-08T19:58:29 release strategy perpetuates understanding 2010-10-08T19:58:32 not the source code 2010-10-08T19:58:38 so if your purpose is to help the greater good, release bots, if its to win, dont. 2010-10-08T19:58:48 the source code is not required 2010-10-08T19:58:52 to improve understanding 2010-10-08T19:58:58 bhasker, releasing source code does as well, for those who learn best by reading and understanding source code 2010-10-08T19:58:59 flag: if your work is invalidated by someone releasing source code, you haven't worked very hard... 2010-10-08T19:59:10 that can happen after the competition 2010-10-08T19:59:12 a lot of what I've learnt is not directly in my code 2010-10-08T19:59:18 bhasker, depends on the person, some people think in terms of code 2010-10-08T19:59:19 it would be like releasing the answersheet 2010-10-08T19:59:21 to get 95% 2010-10-08T19:59:23 in an exam 2010-10-08T19:59:27 if dhartmei would respond to my emails, I could run a McLeopold tcp server 2010-10-08T19:59:32 now you have to work for the remaining 5% 2010-10-08T19:59:37 makes it kinda pointless 2010-10-08T19:59:38 bhasker: but there is no 100% 2010-10-08T19:59:47 the bar gets raised 2010-10-08T19:59:56 to a point where only very few people can compete 2010-10-08T20:00:00 the competition will die 2010-10-08T20:00:02 bhasker, again, like i said, if your goal is to win, or to evaluate skill accuratly, sure.. but if its to perpetuate knowledge then no 2010-10-08T20:00:06 Top 10 players: felixcoto(3842), rsergio(3817), Hazard(3765), ofrias(3742), cfaftw(3732), davidjliu(3687), sequoh(3653), EBraun(3641), savchenko(3632), _iouri_(3631) 2010-10-08T20:00:09 why, can't they compete? 2010-10-08T20:00:12 sigh: the netflix prize went on forever and they were only trying to improve by 10% 2010-10-08T20:00:14 It's not invalidated, but it gives everyone else a starting point that the previous group of people didn't have 2010-10-08T20:00:18 i don't really want to see this become that 2010-10-08T20:00:31 sigh: I don't want to raise the bar of bots, I want everyone to be able to analyze how mine plays. 2010-10-08T20:00:41 its unfair when it comes to "winning" but still the best way to perpetuate knowledge 2010-10-08T20:00:46 *** amstan has quit IRC (Remote host closed the connection) 2010-10-08T20:00:57 tty1: that's why we discuss it after the contest 2010-10-08T20:01:28 jmcarthur, thats fine by me, im not advocating one or the other, just pointing that out :) 2010-10-08T20:01:38 i know not everybody will stick around for that, but those people don't really care about the learning anyway :P 2010-10-08T20:01:48 *** Itkovian has quit IRC (Quit: Itkovian) 2010-10-08T20:02:15 jmcarthur, i like long running competitions like robocode and to see year after year how the top keeps changing 2010-10-08T20:02:27 jmcarthur, and often because the source code was released 2010-10-08T20:02:32 tty1: so do i. i just don't feel like that is what this is meant to be 2010-10-08T20:02:46 CrowBot: so far so good... 2010-10-08T20:02:54 9-9 is going down 2010-10-08T20:03:05 jmcarthur, well maybe not but after the competition i wouldnt be suprised if servers still ran and it became perpetual as well 2010-10-08T20:03:16 i think a couple months is a nice sweet spot for me. i don't want to dedicate my life to it, and there are already other places i can look just for learning 2010-10-08T20:03:30 tty1: the tron one didn't really continue that way for long 2010-10-08T20:03:36 although that wasn't open source 2010-10-08T20:03:53 wow: http://72.44.46.68/canvas?game_id=380474 2010-10-08T20:03:56 jmcarthur, well it is up to the shepards of said project to release it :) 2010-10-08T20:04:02 *** Migi32 has quit IRC (Quit: http://boinc.berkeley.edu) 2010-10-08T20:04:59 McLeopold: wait, crowbot is *simple* right? how is it managing all that? 2010-10-08T20:05:00 CrowBot is like advanced RageBot. 2010-10-08T20:05:13 It's build on top of my own framework. 2010-10-08T20:05:30 241 lines of code 2010-10-08T20:05:43 it's EPICRAGEbot 2010-10-08T20:05:52 >:D 2010-10-08T20:06:15 That makes me darth vader :) 2010-10-08T20:08:07 http://72.44.46.68/canvas?game_id=380419 2010-10-08T20:08:08 hmm crowbot owned my http://72.44.46.68/canvas?game_id=380612 2010-10-08T20:08:27 Another good game that shatters my preconceptions of this game. 2010-10-08T20:10:01 that retybok one is amazing 2010-10-08T20:10:39 yea he has a good bot 2010-10-08T20:10:47 it does no staging 2010-10-08T20:10:49 no forwarding 2010-10-08T20:10:56 just builds ships and sends them out in one sweep 2010-10-08T20:11:02 ohh, I list 2010-10-08T20:11:02 no funnelling at all 2010-10-08T20:11:03 lost 2010-10-08T20:11:38 hah http://72.44.46.68/canvas?game_id=380137 2010-10-08T20:11:40 I can make CrowBot better. 2010-10-08T20:12:05 bhasker: version 1 dude, I've moved on 2010-10-08T20:12:13 hehe 2010-10-08T20:14:02 here's some epic rage: http://72.44.46.68/canvas?game_id=380677 2010-10-08T20:17:50 *** choas_ has joined #aichallenge 2010-10-08T20:20:29 *** choas has quit IRC (Ping timeout: 240 seconds) 2010-10-08T20:21:53 *** Cyndre_ has joined #aichallenge 2010-10-08T20:23:11 crazy: http://72.44.46.68/canvas?game_id=380774 2010-10-08T20:35:32 This might be my favorite so far: http://72.44.46.68/canvas?game_id=380939 2010-10-08T20:36:23 *** McLeopold has quit IRC (Quit: Leaving) 2010-10-08T20:44:15 im trying to use the tcp client but it 2010-10-08T20:44:18 it's taking ages 2010-10-08T20:44:34 oh, connection timed out right now 2010-10-08T20:45:27 *** smhanov_ has joined #aichallenge 2010-10-08T20:46:24 *** smhanov__ has quit IRC (Read error: Operation timed out) 2010-10-08T20:48:15 *** choas_ has quit IRC (Quit: leaving) 2010-10-08T20:49:24 am I the only one who can't connect to the tcp server? :( 2010-10-08T20:51:11 Works for me 2010-10-08T20:51:56 this is weird I can telnet to the server on port 995 2010-10-08T20:52:11 but when I try to use the tcp client it wont work 2010-10-08T21:00:11 Top 10 players: felixcoto(3818), rsergio(3811), Hazard(3771), ofrias(3721), cfaftw(3706), davidjliu(3681), sequoh(3665), EBraun(3627), adumlah(3625), _iouri_(3611) 2010-10-08T21:00:29 man the top 10 is changing quick 2010-10-08T21:00:33 what's going on 2010-10-08T21:11:43 nice to see some mixed nationalities up there 2010-10-08T21:24:13 *** Zeiris_ has joined #aichallenge 2010-10-08T21:25:34 *** Bobng has joined #aichallenge 2010-10-08T21:29:54 Hey, I've been away from this contest for a while and I'm wondering if there's an equally easy alternative to ShowGame.jar 2010-10-08T21:33:17 javagamer: I use the canvas visualiser: http://github.com/DanielVF/Planet-Wars-Canvas-Visualizer 2010-10-08T21:33:29 there are other ones that people have made as well, if you look in the forums 2010-10-08T21:39:20 sigh, how easy is it to use the canvas visualizer from the commandline? 2010-10-08T21:39:58 very easy, there is a python script called visualise_locally.py, just put that in the place of ShowGame.jar 2010-10-08T21:40:14 Perfect! Just what I was looking for, thanks :) 2010-10-08T21:47:21 Has PlayGame.jar ever been updated? I'm still using the original one and I'm assuming by now the fleets disappearing into a planet thing has been fixed. 2010-10-08T21:47:33 Should I just download a new starter pack and replace that file? 2010-10-08T21:57:45 you could try although I don't know of the starter packages ever having been rebuilt. I do think there are some more recent versions stickied in the forum though 2010-10-08T22:00:06 Top 10 players: rsergio(3810), felixcoto(3804), Hazard(3782), ofrias(3725), cfaftw(3688), davidjliu(3673), sequoh(3664), adumlah(3627), _iouri_(3611), EBraun(3610) 2010-10-08T22:04:40 i believe they have been fixed, yes 2010-10-08T22:05:39 *** Bobng has quit IRC (Ping timeout: 264 seconds) 2010-10-08T22:05:41 * Janzert is wishing more and more that turn limit games were ties 2010-10-08T22:05:48 http://ai-contest.com/visualizer.php?game_id=5625918 2010-10-08T22:06:23 *** Bobng has joined #aichallenge 2010-10-08T22:07:27 *** Mathnerd314 has quit IRC (Ping timeout: 264 seconds) 2010-10-08T22:09:08 *** Bobng has quit IRC (Client Quit) 2010-10-08T22:12:55 *** optimum has quit IRC (Ping timeout: 245 seconds) 2010-10-08T22:15:24 hmm, it doesn't seem the api_get_matchup.php in svn is likely the one running on the server either 2010-10-08T22:16:08 *** optimum has joined #aichallenge 2010-10-08T22:32:39 *** bhasker has quit IRC (Ping timeout: 264 seconds) 2010-10-08T22:55:35 There is a version 1.2 of the Java Starter-Kit JAR file. But it still doesn't show more than just the game (ie. production, ships on each side) 2010-10-08T23:00:06 Top 10 players: felixcoto(3805), rsergio(3798), Hazard(3795), ofrias(3736), cfaftw(3691), davidjliu(3663), sequoh(3658), f0rbidik(3623), adumlah(3613), EBraun(3611) 2010-10-08T23:03:45 *** yasith_ has quit IRC (Ping timeout: 265 seconds) 2010-10-08T23:15:09 i find that one of the most annoying things about the java visualizer that comes with the starter packages is the fact that the aspect ratio is all fucked up 2010-10-08T23:15:30 and that it does not display planet growth info 2010-10-08T23:15:57 I mean how are you gonna tell if one planet is further from another if it 2010-10-08T23:16:01 *** yasith_ has joined #aichallenge 2010-10-08T23:16:04 it's all stretched out 2010-10-08T23:16:18 its like 640x480 or some bullshit like that 2010-10-08T23:17:08 *** krokokrusa has joined #aichallenge 2010-10-08T23:20:36 *** rwa has joined #aichallenge 2010-10-08T23:27:07 *** Tobu has quit IRC (Ping timeout: 272 seconds) 2010-10-08T23:28:32 http://72.44.46.68/canvas?game_id=382984 2010-10-08T23:31:47 i just realized that over time i've whittled my planet wars IO and parser code down to a 62 loc module 2010-10-08T23:33:46 http://hpaste.org/40463/planet_wars_microstarter 2010-10-08T23:34:30 it doesn't have the main loop or any niceties, so wouldn't be fair to compare to a real starter, but i'm still pleased at the tininess 2010-10-08T23:40:52 nrub_: awesome match name! "ObligatoryShampoo v.s. nutrub" 2010-10-08T23:41:12 *** hellman has joined #aichallenge 2010-10-08T23:44:51 *** amstan has joined #aichallenge 2010-10-08T23:44:51 *** ChanServ sets mode: +o amstan 2010-10-08T23:51:33 McLeopold is doing animal experiments again. 2010-10-08T23:52:08 After his worm bot became the plague he invented the crow bot: http://72.44.46.68/canvas?game_id=383292 2010-10-08T23:53:17 he's excited about that one, too 2010-10-08T23:53:54 it works better than he imagined it would 2010-10-08T23:57:11 It akts like worm bot with two attacks instead of one at a atime 2010-10-08T23:59:31 But I guess all of his bots can rely on good heuristics and calculations so the attack plan plays a minor role.